You are on page 1of 275

The Rise and Erosion of the South

Eastern Canadian Cordillera


-
Cenozoic Cooling, Exhumation and Elevation of the
Columbia Mountains and Southern Rocky Mountains
in Western Canada from Low-Temperature
Thermochronology

Thesis submitted for the degree of


Doctor of Philosophy (PhD) at the
University of Leicester

By

Annika Szameitat

Diplom (MSc) Geoscientist


(University of Tübingen, Germany), 2011

Department of Geology, College of Science and Engineering


University of Leicester

December 2015
Abstract

The North American Cordillera influences climate on a local and global


scale, by forming a distinct barrier to Pacific moisture reaching the continental
interior. The extent to which this climatic pattern existed in the past is uncertain,
so improving our understanding of the elevation history of the Cordillera is critical
to determining the controls on climate change within the Northern Hemisphere
[e.g. Foster et al. 2010]. In western Canada, the Cordillera comprises two parallel
mountain chains separated by a high elevation (~1100 m) intermontane plateau.
The Cenozoic exhumation history of the western range, the Coast Mountains, has
been well studied [e.g. Parrish, 1983; O'Sullivan and Parrish, 1995; Farley et al.,
2001], while the Cenozoic history of the eastern Cordillera remains poorly
constrained.

This study presents a comprehensive new apatite (U-Th)/He, apatite fission


track and zircon (U-Th)/He dataset of the south-eastern Canadian Cordillera and
constrains the Cenozoic cooling, exhumation and elevation history of the area.
Cooling ages show rapid cooling (>10°C/Ma) from peak metamorphic
temperatures (>500°C) to below 120°C during the Cretaceous to Eocene (75-40
Ma) followed by a period of slow cooling (<1°C/m.y.) and a later phase of rapid
cooling (>10°C/m.y.) to below 70°C since the Miocene (10-0 Ma). Corresponding
exhumation phases modelled using age-elevation relationships of numerous
vertical profiles show 1-5 km of erosion between 70-35 Ma and up to ~2.5 km in
current valleys since ~10 Ma. Paleo-surface reconstructions and Paleo-mean
elevations estimated from isostasy indicate a high elevation (~2.5 km) but low
relief plateau at the end of orogeny (~45 Ma), which gradually lowered in mean
elevation by ~1.5 km until ~10 Ma. A later Neogene increase lifted the peaks a
further ~2.5 km to their current height, while incising up to 2.5 km of relief. The
main causes for both exhumation phases are found to be a combination of
lithosphere delamination, asthenosphere upwelling and thermal expansion, while
the last phase of incision and surface uplift was further enhanced by glacial
incision and isostasy.

2|Page
‘There are no greater examples

of confusion in nature than mountains,

singly or jointly considered.’

-John Ray-

3|Page
Acknowledgements

First, I would like to thank my supervisors, Randy Parrish, Andy Carter, Fin Stuart and
Stewart Fishwick. Thank you for your support, for keeping me on track, for sharing
your experience and opening your labs for me to use. Thanks for always having
thoughtful advice when I couldn’t see the light at the end of the tunnel.

Thank you to the College of Science and Engineering, the Department of Geology and
NIGL for funding this research.

Thank you very much to the technicians at NIGL and SUERC: Adrian Wood and Luigia
Da Nicola, for their help and the wealth of knowledge and experience they have
shared with me during my lab-work.

Thank you to Andrew Smye and Ann-Kathrin Schatz, for accompanying me on my field
work. A special thanks to Andy as well for his help with the forward modelling and for
adapting his code for me to use in my research.

I would like to thank Jim Crawley and Gavin Foster for providing me with additional
samples and data. The same goes for Bernard Guest and his students Andrew
Halverston and Naomi Miles, who are thanked for sharing their data with me. I hope
that our cooperation will go on in the future and that my results will be equally useful
to your research.

I would like to thank my parents Uwe and Edith with all my heart, as well as my
grandparents for their continuing moral and financial support and for always believing
in me.

Last but not least thanks to all the friends I have made during my time here in
Leicester. Leah, Laura, Nina, Nicola, Rowan, Rob, Howard, Alistair, Vince, Tom(s), Phil
and Adeyinka; the three Sarahs (D, Lee and G) for always supporting me; and
everybody else I forgot. Thank you for your friendship, advice, moral support and the
occasional distraction during these trying last couple of years. Particularly Leah and
Rob, thank you so much for keeping me sane, giving me support and shelter when I
needed it and for putting up with my thesis brain towards the end.

4|Page
Table of Contents

Abstract …………………………………………………………………………………………………………..2

Acknowledgements ……………………………………………………………………………………………….4

Table of Figures ……………………………………………………………………………………………..10

Table of Tables ……………………………………………………………………………………………..13

List of Abbreviations ……………………………………………………………………………………………..14

Chapter 1: Introduction

1.1 Project Background and Research Questions ……………………………………………….15

1.2 Geology of the Canadian Cordillera …………………………………………………………..17

1.2.1 Physiographic Location and Division of the Study Area …………….17

1.2.1.1 Overview of the Geological History and Geological Units …………….19

1.2.1.2 Recent Extrusive Volcanic Rocks ……………………………………………….27

1.2.1.3 Cenozoic Plate Tectonics and Heat Flow ……………………………………30

1.3 Paleo-elevation Evidence and Existing Thermochronology ………………………..36

1.3.1 General Physiography, Paleoclimate and Paleoecology …………….36

1.3.2 Erosional Remnants of Flood Basalts ……………………………………………….37

1.3.3 Existing Thermochronology Data ……………………………………………….38

1.3.4 Sedimentary Basins ………………………………………………………………………43

Chapter 2: Methods of Thermochronology

2.1 Underlying Principles ………………………………………………………………………………….47

2.1.1 Closure Temperature and Annealing Behaviour ……………………….48

2.1.2 Combining Methods and Sampling Vertical Transects ……………53

2.2 General Sample preparation ……………………………………………………………………..54

2.3 (U-Th)/Helium Dating …………………………………………………………………………………55

2.3.1 Age Calculation and Temperature Sensitivity …………………………………..55

5|Page
2.3.2 Alpha Recoil Correction ………………………………………………………….57

2.3.3 Standards …………………………………………………………………………………60

2.3.4 Apatite Grain Selection ………………………………………………………….60

2.3.5 Apatite Analysis ……………………………………………………………………..62

2.3.6 Zircon Grain Selection ……………………………………………………………………..64

2.3.7 Zircon Analysis ……………………………………………………………………..65

2.3.8 Caveats …………………………………………………………………………………66

2.3.8.1 Age Dispersion and Effective Uranium ………………………………….68

2.4 Fission Track Dating …………………………………………………………………………………71

2.4.1 Fission Track Formation …………………………………………………………71

2.4.2 Sample Mount Preparation …………………………………………………………72

2.4.3 External Detector Method …………………………………………………………73

2.4.4 Age Calculation …………………………………………………………………….75

2.4.4.1 Zeta Calibration …………………………………………………………………….76

2.4.4.2 Counting Procedure …………………………………………………………………….78

2.4.4.3 Age Dispersion and Central Age ……………………………………………..79

2.4.5 Track Length Measurements …………………………………………………………81

2.4.6 Mineral Chemistry …………………………………………………………………….83

2.4.7 Caveats ………………………………………………………………………………..84

2.5 Thermal modelling of Thermochronology Data ………………………………….85

2.5.1 HeFty ……………………………………………………………………………………………85

2.5.1.1 Forward modelling …………………………………………………………………….85

2.5.1.2 Inverse modelling …………………………………………………………………….87

2.5.1.3 Caveats ………………………………………………………………………………..91

2.5.2 QTQt ……………………………………………………………………………………………93

2.6 Chapter Summary and Conclusions …………………………………………………………95

6|Page
Chapter 3: Thermochronology of the Columbia Mountains

3.1 Introduction ……………………………………………………………………………………………97

3.2 Results …………………………………………………………………………………………………….99

3.2.1 Trench Transect …………………………………………………………………….99

3.2.1.1 Vertical Profile E …………………………………………………………………..103

3.2.2 Mica Creek Transect …………………………………………………………………..105

3.2.2.1 Vertical Profile F …………………………………………………………………..106

3.2.2.2 Vertical Profile G …………………………………………………………………..109

3.2.2.3 Vertical Profile B …………………………………………………………………..109

3.2.2.4 Vertical Profile A …………………………………………………………………..112

3.2.2.5 Vertical Profile D …………………………………………………………………..115

3.2.3 Selkirk-Shuswap Transect ……………………………………………………….118

3.2.4 Southern Transect …………………………………………………………………..120

3.2.4.1 Vertical Profile C …………………………………………………………………..122

3.2.5 Data Discussion …………………………………………………………………..124

3.2.5.1 (U-Th)/He Data …………………………………………………………………..124

3.2.5.2 Fission Track Data …………………………………………………………………..128

3.3 Discussion and Conclusions …………………………………………………………………..142

Chapter 4: Thermochronology of the Southern Rocky Mountains

4.1 Introduction ………………………………………………………………………………………….156

4.2 Results …………………………………………………………………………………………………..157

4.2.1 The North ………………………………………………………………………………158

4.2.1.1 Jasper Vertical Profile …………………………………………………………………..160

4.2.2 The South ………………………………………………………………………………163

4.3 Data Discussion ………………………………………………………………………………164

4.4 Discussion, Summary and Conclusions …………………………………………...173

7|Page
Chapter 5: Cenozoic Exhumation, Elevation and Uplift

5.1. Introduction ………………………………………………………………………………………….176

5.2. Considerations for Interpreting Thermochronology Data …………………….177

5.2.1. Calculation of Modern Geothermal Gradients from Surface Heat Flow ….179

5.2.2. The Influence of Topography on Closure Depth: Draping Correction ……..182

5.2.3. Changes in Geothermal Gradient with Time …………………………………184

5.3. Obtaining Exhumation Histories from Profile Data …………………………………185

5.3.1. Forward Modelling Vertical Profile Thermal Histories and Exhumation Rates

……..……………………………………………………………………………………………………………186

5.3.2. Quantifying Exhumation …………………………….…………………………..203

5.4. Paleo-Relief and Paleo-Elevation …………………………………………………………210

5.4.1. Paleo-Relief from Back-stacking Exhumation Estimates ……….….211

5.4.2. Paleo-Valley Inferences …………………………………………………………214

5.4.3. Former Mean-Elevation Estimates from Isostasy ………………………215

5.4.4. Neogene Summit Uplift and Valley Incision ………………………………….219

5.5. Causes of Exhumation and Uplift …………………………………………………………221

5.5.1. Orogenic Events and Exhumation 70-45 Ma ………………………………….222

5.5.2. Crustal and Back-arc Mantle Processes from 45 Ma to the Present .222

5.5.2.1. Lithosphere Delamination …………………………………………………………223

5.5.2.2. Erosion of Residual Crustal Thickness …………………………………………….225

5.5.3. Neogene Creation of Modern Topography: 10 – 0 Ma …………..226

5.5.3.1. Neogene Back-arc Configuration ……………………………………………..228

5.5.3.2. Possible Mountain Uplift from Asthenosphere Upwelling …………..230

5.5.3.3. Local Maximum Pattern: Relation to Slab Window? ………………………236

5.5.3.4. Glacial Erosion, Isostasy and Summit Uplift ………………………………….238

8|Page
5.6. Conclusions …………………………………………………………………….………………………239

Chapter 6: Synthesis and Conclusions

6.1. Introduction: Aim of this Work ………………………………………………………….242

6.2. Synthesis of Results and Main Conclusions ………………………………………………243

6.3. Future Work ……………………………………………………………………………………………..245

Appendices

Appendix 1: Sample list ………………………………………………………………………………….248

Appendix 2: Sample images ………………………………………………………………………………….252

Appendix 3: Data ……………………………………………………………………………………………..252

A3.1 eU plots ……………………………………………………………………………………………..252

A3.1.1 Chapter 3 ……………………………………………………………………………………………..252

A3.1.2 Chapter 4 ……………………………………………………………………………………………..258

A3.2 Radial plots ……………………………………………………………………………………………..260

A3.2.1 Radial plots of bedrock AFT data …………………………………………………………..260

A3.2.2 Radial plots of modern detrital AFT data, chapter 3 and chapter 4 ………....262

References ………………………………………………………………………………………………………………265

9|Page
Table of Figures
Chapter 1 Page

Figure 1.1-1 Location map of the Canadian Cordillera 16


Figure 1.2-1 Physiographic map of the eastern Canadian Cordillera 18
Figure 1.2-2 Geological map 20
Figure 1.2-3 Field photos: Foothills, front ranges, Rocky Mountains 22
Figure 1.2-4 Field area cross section E to W at 50°N 24
Figure 1.2-5 Field photos: Purcell Mountains, Trench and Monashee complex 25
Figure 1.2-6 Plate tectonic crustal stress evolution 26
Figure 1.2-7 Field Photos: Bugaboo Granite and Chilcotin Basalts 29
Figure 1.2-8 Field Photos: Wells -Gray-Clearwater basalts 30
Figure 1.2-9 Cenozoic plate movements 31
Figure 1.2-10 Present day plate tectonic situation 32
Figure 1.2-11 Back-arc mantle cross section through the Canadian Cordillera 32
Figure 1.2-12 S-wave velocities and elastic thickness of the back-arc 34
Figure 1.2-13 Heat flow cross section of the back-arc 35
Figure 1.3-1 Sedimentary basin map of western Canada 45

Chapter 2

Figure 2.1-1 Temperature ranges of common thermochronometers 47


Figure 2.1-2 Closure temperature concept 49
Figure 2.1-3 PAZ/PRZ temperature ranges 51
Figure 2.1-4 Holding time dependency of measured helium ages 51
Figure 2.1-5 Cooling history dependency of measured helium ages 52
Figure 2.3-1 Helium production from decay inside crystal 55
Figure 2.3-2 Alpha ejection, retention and implantation 58
Figure 2.3-3 Ft correction dependency on grain size 59
Figure 2.3-4 Examples of natural apatites (photographs) 63
Figure 2.3-5 Helium-extraction line setup at SUERC, East Kilbride 64
Figure 2.3-6 Dependency of closure temperature range on effective uranium 69
Figure 2.4-1 Fission track formation inside crystal 72
Figure 2.4-2 External detector method workflow 74
Figure 2.4-3 Zeta calibration of the analyst (Annika Szameitat) 77
Figure 2.4-4 Radial plot principle for statistical analysis of Ft age measurement 80
Figure 2.4-5 Radial plot examples for two real samples 81
Figure 2.5-1 HeFty user interface screenshot 87

10 | P a g e
Figure 2.5-2 Forward modelling of thermal histories 92

Chapter 3

Figure 3.1-1 Physiographic map of the study area 98


Figure 3.2-1 Columbia Mountain samples overview map 100
Figure 3.2-2 Trench transect sample location map 101
Figure 3.2-3 Elevation profile of trench transect 102
Figure 3.2-4 Vertical profile E sample locations, elevation profile and age-elevation plot 104
Figure 3.2-5 Mica creek transect sample location map 105
Figure 3.2-6 Elevation profile of Mica creek transect 106
Figure 3.2-7 Vertical profile F sample locations, elevation profile and age-elevation plot 107
Figure 3.2-8 HeFty thermal history model results of vertical profile F 108
Figure 3.2-9 Vertical profile G sample locations, elevation profile and age-elevation plot 110
Figure 3.2-10 Vertical profile B sample locations, elevation profile and age-elevation plot 111
Figure 3.2-11 HeFty thermal history model results of vertical profile B 112
Figure 3.2-12 Vertical profile A sample locations, elevation profile and age-elevation plot 113
Figure 3.2-13 HeFty thermal history model results of vertical profile A 114
Figure 3.2-14 Vertical profile D sample locations, elevation profile and age-elevation plot 116
Figure 3.2-15 HeFty thermal history model results of vertical profile D 117
Figure 3.2-16 Selkirk-Shuswap transect sample location map 118
Figure 3.2-17 Elevation profile of Selkirk-Shuswap transect 119
Figure 3.2-18 Southern transect sample location map 120
Figure 3.2-19 Elevation profile of southern transect 121
Figure 3.2-20 HeFty thermal history model results of sample PCA-58-84 122
Figure 3.2-21 Vertical profile C sample locations, elevation profile and age-elevation plot 123
Figure 3.2-22 HeFty thermal history model results of vertical profile C 124
Figure 3.2-23 eU correlations for samples presented in this chapter 127
Figure 3.3-1 Summary map of measured ages 143
Figure 3.3-2 Age elevation and -location trend analysis of all samples 145
Figure 3.3-3 Age contour map: all data below 800 m a.s.l. 147
Figure 3.3-4 Age contour maps: all data between 800 - 1200 m and 1200 - 1800 m a.s.l. 148
Figure 3.3-5 Age contour maps: all data between 1800 - 2000 m and >2000 m a.s.l. 149
Figure 3.3-6 All thermal histories in overview: observed cooling phases 151

11 | P a g e
Chapter 4

Figure 4.1-1 Physiographic map of the study area 157


Figure 4.2-1 Map of northern Rocky Mountains results 158
Figure 4.2-2 Age trends and elevation profile E-W across northern Rocky Mountains 160
Figure 4.2-3 Jasper vertical profile samples, elevation profile and age-elevation plot 161
Figure 4.2-4 HeFty thermal history model results of Jasper vertical profile 162
Figure 4.2-5 Map of southern Rocky Mountains results 163
Figure 4.3-1 eU correlations for samples presented in this chapter 166

Chapter 5

Figure 5.2-1 Common closure depth assumptions for vertical profiles 178
Figure 5.2-2 Geothermal gradients calculated from different heat flow values 180
Figure 5.2-3 Principle of warping correction of near surface low-T isotherms 183
Figure 5.2-4 DEM of study area: high resolution and 10km average filtered 183
Figure 5.2-5 Geothermal response to long term exhumation rates 185
Figure 5.3-1 Vertical profile location map 187
Figure 5.3-2 Suitable exhumation histories for vertical profile D 188
Figure 5.3-3 Suitable exhumation histories for vertical profiles A and B 190
Figure 5.3-4 Suitable exhumation histories for vertical profiles F and G 192
Figure 5.3-5 Suitable exhumation histories for vertical profiles C and Jasper 193
Figure 5.3-6 Summary figure of exhumation phases and rates shown in Table 5.3-1 197
Figure 5.3-7 Erosion rate vs geothermal gradient for all profile results 198
Figure 5.3-8 QTQt forward model results for vertical profile D 200
Figure 5.3-9 QTQt forward model results for vertical profile C 201
Figure 5.3-10 QTQt forward model results for Jasper vertical profile 202
Figure 5.3-11 Map of estimated exhumation since ~10 Ma 210
Figure 5.4-1 Back-stacked paleo-surfaces and mean elevations for vertical profiles 213
Figure 5.4-2 Creation of relief and peak surface uplift 220
Figure 5.5-1 Temperatures, gravity and crustal/lithosphere thickness across back-arc/craton 224
transition
Figure 5.5-2 Surface uplift evidence 227
Figure 5.5-3 Back-arc mantle cross section 229
Figure 5.5-4 Plate tectonic and volcanic framework of centre of exhumation 237

12 | P a g e
Table of Tables

Chapter 1
Page
Table 1.3-1 Existing thermochronology data of the south eastern Cordillera 40

Chapter 2
Page
Table 2.5-1 Overview and comparison of currently relevant 90
annealing/diffusion models available in HeFTy and QTqt

Chapter 3

Table 3.2-1 Chemical analysis results of selected apatite samples 130


Table 3.2-2 Apatite (U-Th)/Helium replicate analyses and mean ages 132
Table 3.2-3 Zircon (U-Th)/Helium replicate analyses and mean ages 137
Table 3.2-4 Bedrock apatite fission track results 139
Table 3.2-5 Bedrock apatite fission track results of unpublished data 140
Table 3.2-6 Modern river sediment apatite FT results of unpublished data 141

Chapter 4

Table 4.3-1 Apatite (U-Th)/Helium replicate analyses of bedrock samples 167


Table 4.3-2 Zircon (U-Th)/Helium replicate analyses of bedrock samples 170
Table 4.3-3 Modern Detrital Apatite Fission Track Data 172

Chapter 5

Table 5.2-1 Modern geothermal gradients for profile locations from heat 181
flow
Table 5.3-1 Acceptable range of exhumation histories for vertical profiles 196
Table 5.3-2 Exhumation estimates for vertical profile samples 205
Table 5.3-3 Exhumation estimates for all other samples 209
Table 5.4-1 Paleo-mean elevation from profile exhumation isostasy 218
Table 5.4-2 Peak surface uplift for vertical profiles 220

13 | P a g e
List of Abbreviations

10 ME Mean elevation at 10 Ma

45 ME Mean elevation at 45 Ma

AFt Apatite fission track

AHe Apatite (U-Th)/He

a.s.l. Above sea level (modern sea level, unless otherwise specified)

Ft Alpha ejection correction factor (see chapter 2)

GOF Goodness of fit

He Helium

HeFty Computer program to model thermal histories (Ketcham, 2005)

HF Hydrofluoric acid

Ma Million years (age in geological time)

MME Modern Mean Elevation

m.y. Million years (duration)

NHG Northern Hemisphere Glaciation

ppm Parts per million

RMT Rocky Mountain Trench

Sm Samarium

Th Thorium

U Uranium

ZFt Zircon fission track

ZHe Zircon (U-Th)/He

14 | P a g e
Introduction Chapter 1

Chapter 1:
Introduction
1.1 Project Background and Research Questions

The North American Cordillera runs from Alaska to Mexico and includes all
coastal ranges along the way. It is a major N-S striking mountain range and as such
influences the northern hemisphere climate by providing a barrier for circumventing
wind systems and the distribution of moisture [Foster et al., 2010]. In order to assess
how far into the past such climatic patterns existed, the topographic history of
mountain ranges has to be determined. The southern Canadian part of the North
American Cordillera is characterized by two distinct parallel NW-SE striking mountain
belts with elevations reaching 4000m, the Coast Mountains in the west and the
Eastern Cordillera, including the Columbia and Rocky Mountains, in the east. These
mountain belts are separated by an elevated intermontane plateau (Figure 1.1-1). The
elevation and uplift history of the western side is well studied, while the topographic
history of the Eastern Cordillera remains poorly known.

Numerous thermochronology studies (dating of cooling through specific


temperatures) conducted in the Coast Mountains have unlocked their thermal history
and thus the Cenozoic uplift and exhumation history [e.g. Parrish, 1983; O'Sullivan and
Parrish, 1995; Rohr and Currie, 1997; Farley et al., 2001] . These studies concluded that
the Coast Mountains first cooled through late- and post-orogenic uplift between 60
and 45 Ma and subsequently experienced a hiatus before being slowly and steadily
exhumed from 10 Ma until a final phase of rapid exhumation at 4 Ma. The Eastern
Cordillera is generally assumed not to have experienced any significant increase in
exhumation since the end of the initial orogeny at approximately 45 Ma [Osborn et al.,
2006]. There is, however, some evidence for younger surface uplift, such as high-
elevation erosional remnants of tilted Neogene flood basalts which were originally
deposited horizontally at lower elevations [Peslier et al., 2002; Mathews, 1989].

15 | P a g e
Introduction Chapter 1

Figure 1.1-1: Location of the Eastern Canadian Cordillera and the Coast Mountains, within the north American
Cordillera. The study area of this project is focussed on the south eastern Cordillera as marked by the blue box.
(Background image: Google Earth, 2015)

In a nutshell, it is unclear whether the steep and glaciated topography in the east seen
today was caused by a recent (<45 Ma) tectonic event or is instead an artefact of
glaciation and glacial erosion which incised deep valleys into an older (>45 Ma) orogen
over the past 10 M.y.. This study provides a new, large-scale, low temperature
thermochronology dataset from the south eastern Canadian Cordillera, with the
objective of investigating the post-orogenic exhumation and uplift history. Apatite
fission track (sensitive to ~120°C) and (U-Th)/He dating of apatite (70°C) and zircon
(180°C) have been applied to numerous samples from the study area [e.g. Reiners et
al., 2005]. Cooling histories of individual samples are assessed through kinetic thermal
modelling using HeFTy [Ketcham, 2005]. Finally, forward modelling of varying
exhumation histories and changing geothermal states of the crust is undertaken to
constrain Cenozoic stages and spatial distribution of exhumation within the study area
which will lead to interpretation of potential geological causes for these observations.
The following research questions are addressed through the thesis:

- Is there a systematic pattern of cooling ages that records the Cenozoic


cooling history of the eastern Canadian Cordillera?

16 | P a g e
Introduction Chapter 1

- What is the spatial and temporal distribution of exhumation across the


eastern Cordillera?
- To what degree can these patterns be related to increases in paleo-relief
and potential surface uplift?
- What are the underlying geological mechanisms causing the exhumation
and surface uplift?

Before the new data is presented and these questions addressed, this chapter
introduces the regional geology of the study area as well as the wider geological and
plate tectonic context, and reviews existing evidence related to post-orogenic uplift
and cooling, including existing thermochronological data, to describe the context of
the study.

1.2 Geology of the Canadian Cordillera

1.2.1 Physiographic Location and Division of the Study Area

The term ‘Eastern Canadian Cordillera’ describes the mountain ranges in


eastern British Columbia and western Alberta which are separated from the Coast
Mountains by the Interior Plateau (as shown in Figure 1.1-1). The eastern Cordillera is
divided by a major valley, which is clearly visible on satellite imagery (e.g. Figure 1.2-1),
the Rocky Mountain Trench (for the remainder of this thesis it is abbreviated to RMT).
The precise structural origin of this valley is uncertain, changing from an interpreted
normal fault near the US border to a dextral strike slip fault in northern British
Columbia. Its position roughly marks the buried edge of normal thickness Precambrian
continental crust of North America, with its thinned equivalent to the west at depth. In
the western part of the Eastern Canadian Cordillera, the geology includes accreted
terranes and island arcs thrust upon the North American continental margin rocks
[Wheeler et al., 1991]. The trench also marks a boundary between two lithologically

17 | P a g e
Introduction Chapter 1

Figure 1.2-1: Physiographic map of the eastern Canadian Cordillera, redrawn after [Mathews, 1986]. Inset:
Generalized range names and extent of study area (pink). (Background image: Google Earth, 2015)

different areas: the largely sedimentary and un-metamorphosed fold and thrust
ranges in the east and the metamorphic and igneous rocks in the west (to be
described in detail in sections 1.2.1.1 and 1.2.1.2).

18 | P a g e
Introduction Chapter 1

All ranges to the east of the RMT are considered the Canadian Rocky
Mountains (Figure 1.2-1), while the western side consists of the Omineca and Cassiar
Mountains in the north and the Columbia Mountains in the south. These are
themselves subdivided into six smaller areas, the Quesnel Highland, the Cariboo
Mountains, the Monashee Mountains, the Shuswap Highland and the Selkirk- and
Purcell Mountains.

The study area of this project extends to the southern Rocky Mountains and
the Columbia Mountains. The elevation of the southern half of the Eastern Cordillera
generally is highest near the Rocky Mountain trench and also generally decreases from
north to south. The highest peak in the northern Continental Ranges is Mt Robson
(near Jasper) at 3954 m, while the highest peaks of the north eastern Columbia
Mountains still reach well over 3500 m and tend to decrease to the south west.

1.2.1.1 Overview of the Geological History and Geological Units

The Canadian Cordillera consists of numerous accreted terranes which collided


with the North American continent as a result of north-eastward subduction of the
intermittent oceanic crust from the Middle Jurassic until the Palaeocene (165 - 60 Ma).
The early stage of this phase of contraction is often referred to as the Laramide
Orogeny. It is defined by Jura-Cretaceous subduction of several oceanic plates beneath
island arcs positioned to the west of the North American continent and gradual
amalgamation of these island terranes to the continent edge. This process caused
obduction of back arc sediments adjacent North America onto the Jurassic continental
margin just west of the present day RMT. Subsequent further collisions farther west
and continued shortening led to the formation of the extensive fold and thrust belt
which makes up the Rocky Mountains and the Rocky Mountain Foothills [Dahlstrom,
1970; Price, 1981b; Wheeler et al., 1991]. The contractional faulting and thrust faulting
began in the west and progressed eastward, bringing older and thicker deep water
sediments onto younger blocks. This is evident from the countless thrust faults

19 | P a g e
Introduction Chapter 1

Figure 1.2-2: Simplified geological map of the south eastern Canadian Cordillera, including the southern Rocky Mountains, the
Columbia Mountains and the Interior Plateau, redrawn after [Wheeler et al., 1991]. Numbers indicate significant Volcanic and
basement complexes (see key for details).

20 | P a g e
Introduction Chapter 1

oriented parallel to the RMT (NW-SE) through Proterozoic to Oligocene rocks (green,
light blue and brown colours) (Figure 1.2-2). The boundary between the foothills and
the Rocky Mountains is defined by the first Cambrian limestone thrust upon younger
strata, which can be seen e.g. near Banff, where the McConnell thrust has placed thick
bedded Cambrian limestone onto Cretaceous shale, now exposed as gently rolling
foothills (Figure 1.2-3). The latest thrust movements have been recorded on the
McConnell- and Lewis-thrusts during the earliest Eocene (52Ma) [Van der Pluijm, B. A.
et al., 2006], marking the end of the orogeny for the mountains east of the RMT.

Unlike the brittle-deformed and mostly un-metamorphosed strata east of the


RMT, the Proterozoic North American continental margin rocks west of the trench are
largely deformed by smaller scale folding and altered by greenschist to amphibolite
metamorphism [Wheeler et al., 1991]. The collision of terranes to the west generated
nappes which buried the continental margin sediments transforming them in many
areas from sandstones, siltstones and limestone (as found east of the RMT) into mica-
rich phyllites and schists. These rocks now form the Cariboo and Purcell Mountains as
well as the northern parts of the Selkirk and Monashee Mountains.

These Proterozoic to Cambrian deposits together with the Proterozoic to


Triassic meta-volcanics and magmatic arc rocks are generally referred to as the
Kootenay terrane. The word ‘terrane’ has to be used tentatively, because although it is
widely labelled as such, the allochthonous origin of the rocks comprising it, so whether
the eastern units were at any point actually detached from the continent or whether
they always remained in a similar position since deposition, has been debated for
decades [Wheeler, 1966; Price, 1981a; Brown and Read, 1983]. The most recent
studies suggest that it is pericratonic in origin and that it can be stratigraphically linked
– despite being overturned - to the North American shelf strata underneath [Smith
and Gehrels, 1992a; Smith and Gehrels, 1992b; Colpron and Price, 1995].
Nevertheless, the term ‘Kootenay terrane’ is still widely used, most often describes all
stratified rocks of Neoproterozoic to early Mesozoic age west of the Purcell thrust and
geographically includes almost all of the Columbia Mountains, thus comprising most of
the study area of this thesis.

21 | P a g e
A C

Figure 1.2-3: Field photographs from south eastern Canadian Cordillera ʹ east of the Rocky Mountain Trench. A: Westward view of the foothills and front ranges when approaching on
highway 1 from Calgary. B: The front ranges begin where the first Cambrian strata is thrust upon the softer Triassic and Jurassic rocks which make up the Foothills, as shown here for the
McConnell thrust near Banff (view to the North). C: Typical view in the Canadian Rocky Mountains, looking towards the North over Peyto Lake. The images shows the Cambrian and
Devonian limestones of the main ranges tilting westward (left) and being separated by a valley which marks a thrust fault from the next thrust stack to the east.

22 | P a g e
Introduction Chapter 1

The rocks to the east of the Monashee Complex, i.e. the Selkirk and Purcell
Mountains (see Figure 1.2-2), - although chemically very similar - differ significantly in
metamorphic grade and extent of structural deformation to those from the western
part. This complicates the definition of the Kootenay terrane boundary and has caused
the correlations to be questioned [Fyles and Eastwood, 1962; Wheeler, 1966; Smith
and Gehrels, 1992a; Colpron and Price, 1995]. The general consensus is that this
eastern unit, now called the ‘Selkirk Allochthon’, was thrusted over the underlying
units via the Purcell thrust and folded and deformed in the process. The subsequent
rise of the Monashee complex caused the thrust surface to be reactivated in a normal
fashion and the Selkirk Allochthon to be downthrown in relation to the core complex
along the Columbia River normal fault. A schematic cross section created during field
work along 50°50’N (Figure 1.2-4) and field photographs (Figure 1.2-5) illustrate this
divided appearance.

Further to the west the crust of the Interior Plateau and the Coast Mountains
comprise of the Intermontane Superterrane (which in southern BC includes sub-units
such as Slide Mountain Group, the Quesnel terrane (dark green), the Cache Creek
terrane (dark grey) and Stikinia terrane (light green)) and the Insular Superterrane
(sub-units include Gravinia, Iluzotin, Alexander and Wrangellia) (Figure 1.2 2 and
Figure 1.2 4). Both superterranes amalgamated from smaller terranes during the
upper Triassic to upper Jurassic time and subsequently collided with the North
American continent during the Jurassic to Cretaceous due to the ongoing convergence
[Monger et al., 1982; Rubin et al., 1990; Woodsworth et al., 1991; Unterschutz et al.,
2002; Trop et al., 2002]. Because the study area of this project does not extend to
include them directly, their geology will not be discussed in more detail at this point.

Following northeast-southwest convergence in the Jurassic to Cretaceous (165


to 65 Ma), the stress field changed leading to a phase of northwest-southeast
extension from 65-45 Ma [Tempelman-Kluit and Parkinson, 1986; Price and
Carmichael, 1986; Parrish et al., 1988]. This rotation initiated NW-SE oriented dextral

23 | P a g e
Figure 1.2-4͗^ĐŚĞŵĂƚŝĐĐƌŽƐƐƐĞĐƚŝŽŶĂƐĚƌĂǁŶĚƵƌŝŶŐĨŝĞůĚǁŽƌŬĂůŽŶŐϱϬΣϱϬ͛E to illustrate the general structure of the Canadian Cordillera at this latitude. Cross sections drawn further
north or south might include other terrane segments or basement complexes which are not shown here. From east to west it shows how the flat lying sedimentary deposits of the Alberta
Plains are stacked eastward in the Fold and Thrust Belt, which constitutes the Rocky Mountain Foothills, Front Ranges and Main Ranges. To the west of the RMT and the Purcell thrust lies
the Kootenay terrane with the highly metamorphic, folded and in parts overturned Selkirk Allochthon thrust over the Monashee complex in the east and the stacked continental shelf
metasediments and arc magmatic terrane rocks (including the mafic Tsalkom ophiolite) in the west. Following a change of scale this cross section also includes a rough illustration of the
Intermontane Superterrane and the beginning of the insular Superterrane which make up the Interior Plateau, Coast Mountains and Insular Belt (not depicted) to the west. Note that this
section only illustrates the underlying geology and not the current topography in addition to omitting the Neogene flood basalt cover as well. Furthermore the thrusts faults are shown as
highly exagerrated and oversteepened to illustrate the general stacking order.

24 | P a g e
A B

Figure 1.2-5: Field Photographs from south eastern Canadian Cordillera. A: Westward view towards the northern Purcell Mountains from inside the wide, flat valley of the Rocky
Mountain Trench. B: Northward facing view from the top of Mt Revelstoke, showing both the snow covered peaks of the Monashee Complex (left) and the Selkirk Allochthon (right). The
valley separating the two marks the location of the Columbia River Normal fault, which was responsible for overthrusting the Selkirk Allochthon to the east and got reactivated when the
Monashee complex rose, reverting the thrust motion into a downward slipping normal fault.

25 | P a g e
Introduction Chapter 1

strike slip movement along the Fraser River-Straight Creek Fault system and the
Tintina-Northern Rocky Mountain Trench system (Figure 1.2-6) and thus led to the
formation of numerous SW-NE striking normal faults in the Interior Plateau. This
thinning of the crustal stack in combination with lithospheric thinning and increased
buoyancy facilitated the rise of several metamorphic ‘core complexes’ through the
thin accreted crust west of the RMT [Wanless and Reesor, 1975; Brown and Read,
1983; Parrish and Armstrong, 1983; Evenchick et al., 1984; Ranalli et al., 1989;
Wheeler et al., 1991]. These include the early Proterozoic (1.8 - 2.2 Ga) basement
gneisses of the Monashee Complex and the smaller Malton Gneiss (Figure 1.2-2) as
well as the craton related lower level granitoids, ortho- and paragneisses of the
Paleocene-Cretaceous Valhalla complex [Wheeler et al., 1991; Spear and Parrish, 1996;
Spear, 2004]. In addition, numerous back-arc flood basalts erupted during the
Paleogene and Neogene.

Figure 1.2-6: Evolution of the crustal stress field from upper Cretaceous northeast-southwest compression (thick
black arrows labelled as 75-59Ma) to Eocene northwest-southeast extension (black arrows labelled as 58-42Ma)
and the resulting structural framework. The strike slip movement resulting from extension is translated from the
Fraser River ʹ Straight Creek fault system into the Tintina ʹ Northern Rocky Mountain Trench system by
numerous accomodating SW-NE striking normal faults which caused stratigrahically lower metamorphic and
granitic rocks to rise to the surface as core complexes (Figure from [Price, 1994] after [Price and Carmichael,
1986]).

26 | P a g e
Introduction Chapter 1

Climatic conditions cooled during the Miocene period resulting in the Northern
Hemisphere glaciation, and the Cordilleran Ice Sheet periodically covered most of the
Canadian Cordillera. Mountain glaciation is thought to have begun in the region as
early as the late Miocene (~10Ma) [Clague, 1991a]. This is similar in time to the onset
of widespread Miocene glaciation in southern Alaska [Denton and Armstrong, 1969;
Clague, 1991b; Clague, 2009]. At its maximum extent the ice sheet covered all of the
Coast Mountains and Eastern Cordillera, including the Interior Plateau. During
frequent interglacial warm periods the ice retreated to the mountainous areas. Major
glacial episodes with such large ice sheets have been suggested to date as far back as
3-1 Ma [Souther et al., 1984; Naeser et al., 1982; Mathews and Rouse, 1986], yet very
few glacial deposits still exist to support this and thus the glacial history of the early
Pleistocene remains poorly known. Abundant evidence for the last two major
glaciations during the past 40 ka (early and late Wisconsinian) exists in the form of
widespread glacial till deposits, in some areas up to 100 m thick, which often obscure
the underlying bedrock geology.

1.2.1.2 Recent Extrusive Volcanic Rocks

Aside from the numerous phases of syn-orogenic intrusive/extrusive volcanism


which are mostly associated with the subduction of oceanic crust and terrane collision
(Figure 1.2-2), there are several phases of post-orogenic volcanism in the Eastern
Canadian Cordillera. Because the rise of melts through the crust may influence the
thermal history of rock masses and melt events may be linked to rock and surface
uplift, these will now be described in more detail.

During the earliest Neogene re-intensification of subduction beneath the Coast


Mountains led to the establishment of the current back-arc setting of the Interior
Plateau and resulted in back-arc volcanism. Most notable for this study are the
Neogene flood basalts of the Chilcotin group, which cover 50,000 km2, almost the
entire Interior Plateau [Tipper, 1978]. These non-marine back-arc basalts are

27 | P a g e
Introduction Chapter 1

compositionally transitional between alkaline and tholeiitic and have an average


exposed thickness of only 70 m (max. thickness is 140m) [Bevier, 1983]. They typically
appear as jointed basalt columns in outcrop (Figure 1.2-7). Although the Chilcotin
flows range in age from the Oligocene (minor deposits as old as 31.1 Ma) to
Quaternary (0.7 Ma) they display three major phases of eruptive episodes: 16 -14 Ma,
10 - 6 Ma and 3 - 2 Ma [Mathews, 1989; Mathews, 1991].

During the Quaternary there have been two notable smaller volcanic fields. The
Anahim volcanic belt is thought to be a hot spot trace which stretches from the coast
all the way into the Interior Plateau along 52°40’N and decreases is age from east to
west from 14.5 Ma to 7ka [Bevier, 1989] (the easternmost cones on the geological
map (labelled ②)). The Wells Gray-Clearwater volcanic field is a small scale valley-
filling volume of flood basalts (①in Figure 1.2-2, also Figure 1.2-8) that erupted
between 3.5 Ma and the late Holocene [Hickson, 1987]. Initially they were thought to
be the easternmost extension of the Anahim hot spot trace [Rogers and Souther,
1983; Hickson and Souther, 1984; Bevier, 1989]. However, unlike for the western
expressions of the hot spot trace which line up perfectly along a gradient in both space
and time, the oldest Wells Gray flows are considerably older than the youngest
Anahim flows further west. Furthermore the trace of the hot spot does not line up
with the position of the Wells Gray field and thus the plate movement would have to
have changed abruptly to accommodate this. This has not been observed by the latest
plate tectonic reconstructions [Madsen et al., 2006] and thus calls the hot spot origin
for these basalts into question. An alternative interpretation suggests that the two
fields are independent and that the Wells Gray-Clearwater field was caused by
asthenospheric upwelling, potentially related to the Nootka-Fault slab window (to be
discussed in next section) and production of melts which rose facilitated by deep
crustal structures [Madsen et al., 2006].

28 | P a g e
A C

Figure 1.2-7: Field Photographs of igneous rocks from the Columbia Mountains and Interior Plateau. A: West-facing view of the spires created by increased resistance to glacial erosion of
the Creatceous Bugaboo granodiorite batholith (Purcell Mtns) in comparison to the surrounding Proterozoic siltstones and quartzites. B and C: Chilcotin flood basalts of the Interior
Plateau near 100 Mile House. Vertical and horizontal columnar joints can be found here At the same locality within 100m of each other.

29 | P a g e
Introduction Chapter 1

Figure 1.2-8: Field photographs of the valley-filling flood basalts of the Wells-Gray-Clearwater volcanic field. The
National Park contains 39 waterfalls which leads to great exposure of the columnar joints lining the valley walls.
Left: Several layers of columnar joint generations are clearly stacked on both sides of the Cougar falls (water drop
20 m). Right: The Helmcken waterfall (Murtle River drops 141 m) has caved out most of the valley filling basalt,
however at the top ledge layers of cooling joints are still visible.

1.2.1.3 Cenozoic Plate Tectonics and Heat Flow

Not only crustal processes such as collision of plates and thrusting can cause
surface uplift. Deeper processes such as lithospheric thinning or delamination as well
as upwelling of mantle material in the asthenosphere can also cause uplift of the land
surface due to increased buoyancy of the crust caused by load reduction or an
increase of hot material near the Moho. In order to interpret patterns of uplift and
exhumation using thermochronology data it is important to be aware of the past and
modern plate tectonics. The Cenozoic evolution of the plate tectonic framework which
succeeded the formation of the terrane assemblage according to Madsen et al. (2006)
is shown in Figure 1.2-9. The continuous and rapid northward movement and
subduction of the Resurrection- and Kula Plates during the lower Eocene eventually
caused the Pacific Plate to reach the Alaska subduction zone around 40 Ma ago. The

30 | P a g e
Introduction Chapter 1

Figure 1.2-9: Reconstruction of plate tectonic movement and subducting slab positions from the lower Eocene
(53 Ma) to Miocene (10Ma) by [Madsen et al., 2006]. Continuous north-east movement of the Kula-,
Resurrection- and Pacific Plate caused the Kula- and Resurrection plates to be completely subducted by the
upper Eocene (39Ma), at which point eastward subduction ceased and the Pacific plate strike-slip movement was
established parallel to the Canadian Coast. The Farallon Plate continued subduction to the east on the US coast,
slowly moving northward. The northern part of the Farallon Plate separated at 28 Ma and became known as the
Juan de Fuca Plate while its subducting slab was positioned underneath the southern Canadian Cordillera.

31 | P a g e
Introduction Chapter 1

Figure 1.2-10: Present day plate tectonic situation and Cenozoic volcanics of the southern Canadian Cordillera,
redrawn after [Madsen et al., 2006] with plate velocities from [Riddihough, 1984] and slab angle from [Govers
and Meijer, 2001]. The extent of the study area is shown by the dashed line. The Nootka transform fault formed
approximately 4 Ma ago when conflicting plate motions caused part of the Juan de Fuca plate to become
partially coupled with the Pacific Plate and separate, forming the Explorer Plate.

Figure 1.2-11: Schematic cross section through the crust and lithosphere of the Canadian Cordillera, redrawn
after [Hyndman et al., 2005; Currie and Hyndman, 2006]. This section follows approximately the profile A to B
shown in Figure 1.2-13.

movement of this plate relative to the Canadian continent is to the north-west.


Consequently, subduction ceased and the current coast-parallel dextral strike slip

32 | P a g e
Introduction Chapter 1

motion was established. During the Oligocene the northward components of both the
Farallon- and Pacific Plate motion caused the spreading ridge between them to move
to the north and the Farallon Plate subduction zone reached the Canadian coast.

At 28 Ma the northern segment of the Farallon Plate separated, (becoming the


Juan de Fuca Plate) and continued to move northward, increasing the length of the
subduction zone along the Canadian coast. The slab edge reached the sub-Rocky
Mountain asthenosphere in the upper Miocene and continued to subduct, until partial
coupling of the northern section with the Pacific Plate caused the plate to tear,
resulting in the small Explorer Plate and forming the Nootka Transform Fault (Figure
1.2 10) [Madsen et al., 2006].

The present day plate configuration (Figure 1.2-10) shows how the trace of the
Nootka transform fault lines up with the location of the Wells Gray-Clearwater
volcanic field which began to erupt at 3.8Ma (~200ka after the formation of the
Nootka Fault). This led Madsen et al. (2006) to the conclusion that the difference in
subduction angle between the shallow Explorer Plate and the steeper (~45-65° dip,
after [Govers and Meijer, 2001]) Juan de Fuca Plate opened up a vertical ‘slab window’
of unknown aperture [Thorkelson, 1996], facilitating asthenospheric upwelling and
melt production which led to the eruption of the volcanic field. This slab window has
migrated approximately 50km northward since its formation, and is currently located
beneath the study area.

The crustal and lithospheric structure of the overlying continent is illustrated


by the cross section in Figure 1.2-11. S-wave tomography and gravity data (Figure
1.2-12) have established that the crust and lithosphere composed by the accreted
terranes west of the RMT are thin and can be classified as a classic ‘mobile belt’
[Hyndman and Lewis, 1999; Hyndman et al., 2005; Currie and Hyndman, 2006].
Despite the shallow depth of the Moho (~ 35 km) and a lithosphere-asthenosphere
boundary depth of ~60 km the elevation of the Interior Plateau and the Columbia
mountains is fairly high (~100m – 1300 m mean elevation). The lack of a crustal root is
compensated for by thermal buoyancy, which lifts the back arc by ~ 1600 m due to the

33 | P a g e
Introduction Chapter 1

high temperatures (Figure 1.2-12) [Hyndman et al., 2009]. The Interior Plateau is
currently in a back-arc position relative to the Coast Mountains volcanic belt and thus
asthenospheric upwelling around the subducting slab causes higher temperatures
underneath the Cordillera in comparison to the craton (Figure 1.2-11 and Figure
1.2-12a). The Rocky Mountain Trench marks the edge of the ancient continent and
thus the thick cratonic crust and thicker, cooler lithosphere underneath (Figure 1.2-11
and Figure 1.2-12b).

Figure 1.2-12: Lithospheric architecture from [Currie and Hyndman, 2006]. a) S-wave velocities at a depth of
100km, data from [Van Der Lee and Frederiksen, 2005]. b) The effective elastic lithospheric thickness from
topography and gravity data [Lowry and Smith, 1995; Flück et al., 2003].

This difference in temperature and lithospheric thickness across the Cordillera


and adjacent craton generates a wide variation and complicated pattern of heat flow
through the crust. Currie and Hyndman (2006) have compiled a dataset of heat flow
measurements (Figure 1.2-13a) and established that it increases from approximately
40 mW/m2 in the forearc to an average of 75 ± 15 mW/m2 in the mobile belt and
decreases again to 42 ± 10 mW/m2 on the craton, east of the RMT (Figure 1.2-13b).
The average value given for the backarc is composed of various datasets including
Hyndman and Lewis (1999), who provided a more detailed distinction of heat flow

34 | P a g e
Introduction Chapter 1

bands and found that values increase from 73 mW/m2 in the Interior Plateau to 86
mW/m2 for the Omineca Belt (Kootenay terrane) and even 104 mW/m2 for the Selkirk
Allochthon.

Figure 1.2-13: a) Heat flow data for the Canadian Cordillera subduction zone from [Currie and Hyndman, 2006].
Black line from A to B indicates the heat flow profile location and all data within the dotted lines was included. b)
Heat flow profile A to B across the Canadian Cordllera. White circles show raw data while black circles have been
corrected for near surface heat generation (refer to Currie & Hyndman, 2006 for details). c) Geothermal gradients
for the backarc (Interior Plateau and Columbia Mountains) and the craton (Rocky Mountains and Plains)
calculated from the surface heat flow and other constrainst such as seismic velocities, xenoliths and lithoprobe
imagery.

However, while these detailed distinctions can be made for the Columbia
Mountains and the data density between the fore- and backarc also allows for a
gradient of the heat flow transition to be observed (see Figure 1.2-13b), no such
dataset exists for the transition from mobile belt to craton (i.e. across the RMT). The
same problem occurs for the slope of the lithosphere-asthenosphere boundary
underneath the RMT, which can only be inferred from sparse seismic data and thus
the precise rate of change is debated [Zelt and White, 1995; Hyndman and Lewis,
1999; Burianyk et al., 1997]. What is generally documented are upper mantle
temperatures in the back arc ~500 degrees higher than on the craton for the same
depth, which significantly influences estimated geothermal gradients [Hyndman and
Lewis, 1999].

35 | P a g e
Introduction Chapter 1

The cause of this increased thermal state is thought to be delamination of the


lithosphere in the back-arc [Currie et al., 2008]. However, it is not well known when
this occurred and therefore how far this pattern of elevated heat flow and hot
Asthenosphere extended into the past. The most likely explanation based on the
regional geology is that the increase in temperature is related to the onset of
increased subduction of the Farrallon plate ~55 Ma ago [Currie et al., 2008; Hardebol
et al., 2012]. This issue will be further discussed in chapter 5, when making
interpretations regarding the exhumation and uplift histories resulting from our
thermochronology data.

1.3 Paleo-elevation Evidence and Existing Thermochronology

1.3.1 General Physiography, Paleoclimate and Paleoecology

The appearance of the Eastern Cordillera with its glaciated, rugged peaks and
steep, clearly visible thrust sheets at first glance looks like a relatively young and
recently tectonically active orogen rather than an ancient, erosional remnant of a
much larger range. Nevertheless, a recent study has provided support for a prevailing
view that the end of the collisional Laramide orogeny at 60 Ma marked the end of the
‘constructional process’ and the topography observed today is the result of continuous
erosion since the Palaeocene [Osborn et al., 2006]. Based on stratigraphic and critical-
taper reconstructions they concluded that particularly the ranges to the east of the
RMT experienced several kilometres of erosion which reduced a post-collisional hilly,
high-elevation plateau to the current rugged terrane of the Rockies and Foothills.

While it may be feasible that the current surface-topography was shaped by erosion
rather than reactivated compression, several more climatically focussed studies
indicate that overall uplift of the land surface could have occurred which facilitated
this increased erosion [Kohn et al., 2002; Takeuchi and Larson, 2005; Kent-Corson et

36 | P a g e
Introduction Chapter 1

al., 2006; Rouse and Mathews, 1979]. Oxygen isotope based paleoaltimetry from
paleosols and fossil equid teeth from the Coast Ranges and also the Eastern Cordillera
in the USA (just south of the border and thus just south of the study area) indicate a
shift in δ18O and thus an increase in elevation over the past 10-15 m.y. [Kohn et al.,
2002; Takeuchi and Larson, 2005; Kent-Corson et al., 2006]. Furthermore, palynology
and paleoecology data from the Quesnel area in the Interior Plateau near the western
flank of the Eastern Cordillera showed that the pollen and organic material washed
into lakes which are fed by rivers from the mountain flanks indicate a significant shift
from paratropical vegetation to a more temperate and drier climate between the
Eocene and Miocene [Rouse and Mathews, 1979]. This observation is based on a
sharp increase in coniferous trees and on its own could simply be attributed to the
overall cooling of the Cenozoic climate and the development of a significant rain
shadow in response to the emergence of the Coast Mountains. However, it could also -
particularly in combination with other factors described in this section - indicate an
increase in elevation of the Eastern Cordillera sometime between 50 and 5 Ma.

1.3.2 Erosional Remnants of Flood Basalts

As described in section 1.2.1.2, the Chilcotin group back-arc flood basalts were
erupted in three major phases in the Neogene to Quaternary (16 -14Ma, 10 - 6 Ma and
3 - 2 Ma) [Mathews, 1989; Mathews, 1991]. Mathews (1989) was also amongst the
first to realize that these post-orogenic and originally more or less horizontally
deposited flows are today widely warped and tilted, especially along the eastern flank
of the Coast Mountains where they are found at up to 2000m a.s.l. while the average
elevation of the basalt flows in the plateau is only approximately 1000m a.s.l.. This is
consistent with up to 4 km of uplift that the Coast Ranges have experienced over the
past 10 M.y. [Parrish, 1983]. However, erosional remnants of Chilcotin basalts are also
found at high elevations on the edge of the Columbia Mountains. For example on the
Okanagan Highland tilted caps of 11-5 Ma Chilcotin Group basalts are located at the

37 | P a g e
Introduction Chapter 1

tops of peaks well over 1750 m.a.s.l., indicated as ‘Chilcotin outliers’ in Figure 1.2-10.
These rocks provide some of the strongest evidence that the end of the orogenic
phase in the upper Eocene did not mark the end of uplift for at least the southern
Columbia Mountains.

1.3.3 Existing Thermochronology Data

Thermochronology is the study of rock cooling, typically using mineral specific-


techniques to date samples using radiogenic isotops. Commonly used minerals include
zircon, apatite, micas and feldspars which are dated using methodologies such as e.g.
(U-Th)/He, fission track, U-Pb or Ar-Ar dating. Currently, temperature ranges from as
low as 40°C (apatite U-Th/He) up to 900°C (zircon U-Pb) are routinely investigated this
way [Reiners et al., 2005]. For the investigation of the most recent rock cooling
histories and surface processes such as erosion which shaped a modern mountain
range, low temperature thermochronology methods such as apatite and zircon (U-
Th)/He and apatite fission track dating (sensitive to 40-70°C, ~180°C and 80-120°C
respectively) are most useful, since they record cooling through the uppermost 1-5 km
of the crust. A more detailed discussion is given in Chapter 2: Methods of
Thermochronology.

The Cenozoic exhumation of the Coast Mountains is well studied. Post orogenic
exhumation from 60-45 Ma followed by a slwoing and then increased exhumation,
creation of relief and resulting surface (peak) uplift over the past 10 M.y. [Parrish,
1983; O'Sullivan and Parrish, 1995; Rohr and Currie, 1997; Farley et al., 2001].

For the Eastern Cordillera research has focussed mainly on peak metamorphism and
emplacement ages of the numerous igneous rocks and basement complexes of the
Kootenay terrane, while the lower temperature cooling history remains poorly known.
Numerous U-Pb and Ar-Ar dating studies have shown that rocks cooled through 300 °C
from the upper Cretaceous to the upper Eocene (100-45 Ma), consistent with the

38 | P a g e
Introduction Chapter 1

timing of syn-orogenic peak metamorphism, emplacement of intrusions and later the


rising of the basement complexes during NW-SE extension ([Parrish, 1995; Colpron et
al., 2000; Vanderhaeghe et al., 2003] and references therein). For a comprehensive list
of existing thermochronology data from the study area see Table 1.3-1.

Only three published works exist which used lower temperature methodologies.
Lorencak et al. (2001) investigated the cooling of the Shuswap metamorphic complex
via zircon and apatite fission track dating (240-80°C) and found Oligocene to Eocene
(28-54 Ma) cooling ages. A study by Osadetz et al. (2004) used apatite fission track
dating to determine exhumation caused by the Lewis thrust fault and found similar
Eocene to upper Cretaceous cooling ages (45-85 Ma). However, both these studies are
localized to smaller areas and are not necessarily indicative of the overall cooling
history of the entire range. Mc Donough et al. (1995) studied the rage of AFt ages
along one transect across the Rocky Mountains along the Athabasca River from Hinton
to the Malton Gneiss to investigate the potential of hydrocarbon storage in these
rocks. They found that the AFt ages decrease from 52-73 Ma in the Foothills to 30-46
Ma in the Rocky Mountains and are as young as 14-25 Ma in the vicinity of the RMT.
Because thermal histories meant that the rocks were deemed unprospective for
hydrocarbons, no larger scale study was undertaken. Unpublished apatite fission track
data of bedrock ages for the Valhalla and Monashee Complex showed Oligocene to
Eocene (21-44 Ma) cooling below 120°C and also unpublished modern detrital river
sediments, which provide a distribution of ages across the entire river catchment,
dated by the same method yielded a similar Oligocene to Eocene age range (28-48 Ma)
for numerous locations across the Columbia and Rocky Mountains (All unpublished
data were provided by R. Parrish and G. Foster (analysed by A. Carter) and are
included in the results of this study).

39 | P a g e
Table 1.3-1: Collection of previously available thermochronology data from the south eastern Canadian Cordillera

Max.
Method temperature Age [Ma] Location Comments Reference
[°C]

Mt Copeland 740 ± 36, Syenite Lake 670 ±


Igneous (formation ages) Monashee Complex, [Parrish, 1995] and references
U-Pb zircon 800 96, Northern Monashees 541 ± 11, West
540-740 Selkirk Allochthon therein
flank Monashees 1.5 Ga

Mount Revelstoke [Parrish, 1992; Crowley and


U-Pb zircon 800 71 Intrusion age
Leucogranite Brown, 1994]

Monashees mainly 58 Ma (North 90-136


Monashee Complex, [Parrish, 1995] and references
U-Pb monazite 800 58-90 Ma, North West 55-95 Ma), Valhalla 60-90
Selkirk Allochthon therein
Ma, Cariboos 125-154 Ma

Monashees mainly 52-59 Ma (Northern


flank 57-137 Ma and West side 55-73 Ma) ,
Metamorphic ages mainly Monashee Complex, [Parrish, 1995] and references
U-Pb zircon 800 Valhalla stages from 52-110 Ma, Selkirk
52-73 Selkirk Allochthon therein
Allochthon 162-174 Ma, Cariboos 125-154
Ma

Monashee Complex,
Selkirk 174 Ma, Northwest Monashee 54-74 [Parrish, 1995] and references
U-Pb titanite 580 54-74 (Allochthon 174) Valhalla Complex,
Ma, Valhalla 60 Ma therein
Selkirk Allochthon

Most ages around 60-70 Ma, including Mt


Ar/Ar Clachnacudainn Revelstoke, Mica creek some older ones [Colpron et al., 2000] and
500 64-149
hornblende Complex (127-149 Ma) and also Albert Peak (towards references therein
the east, around 106 Ma)

40 | P a g e
Max.
Method temperature Age [Ma] Location Comments Reference
[°C]

Cooling from 500 to 280 from 60-45 Ma at


U-Pb titanite, K-
24 ± 6 degrees per km' (U-Pb titanite 60 Ma, [Parrish, 1995] and references
Ar hbl, musc, 500 60 - 45 Valhalla Complex
K-Ar hbl 58 Ma, Rb-Sr musc 55 Ma, K-Ar therein
biot, Rb-Sr musc
musc 48 Ma, bt 47 Ma)

Monashee Complex, Valhalla 58 Ma, northern Monashee 70-82


Ar-Ar [Parrish, 1995] and references
450 58-82 (Allochthon 164) Valhalla Complex, Ma, west Monahsee 61 Ma, sellkirk 164
hornblende therein
Selkirk Allochthon Ma,

Ar-Ar
450 50-56 Grand Forks Complex - [Cubley et al., 2013]
hornblende

Ar-Ar
450 48-59 Shuswap Complex - [Vanderhaeghe et al., 2003]
hornblende

Monashee Complex north west Monashee 53 Ma, southern [Parrish, 1995] and references
U-Pb rutile 400 53 and 70-80
and Cariboo Mountains Cariboos 70-80 Ma, therein

Ar-Ar muscovite 350 49 Shuswap Complex age given is average, see paper for detail [Vanderhaeghe et al., 2003]

Clachnacudainn Mt revelstoke 48-55 ma, south of TCH older [Colpron et al., 2000] and
Ar-Ar muscovite 350 48-91
Complex 69-91Ma references therein

Monashee Complex, Valhalla 47-48Ma, southern cariboo 55-


[Parrish, 1995] and references
K-Ar mica 350-300 47-66 (Allochthon 166) Valhalla Complex, 66Ma, northern Monashee 48-58Ma, west
therein
Selkirk Allochthon Monashee 50Ma, Selkirk 158-166Ma

41 | P a g e
Max.
Method temperature Age [Ma] Location Comments Reference
[°C]

Ar-Ar biotite 300 51 Grand Forks Complex age given is average, see paper for detail [Cubley et al., 2013]

Ar-Ar biotite 300 49 Shuswap Complex age given is average, see paper for detail [Vanderhaeghe et al., 2003]

Mt Revelstoke and eastward. South of TCH


Clachnacudainn slightly older (Albert Stock etc 70-80Ma) [Colpron et al., 2000] and
Ar/Ar biotite 300 47-83 (main cluster 50-55)
Complex while Mt Revelstoke area and Mica Creek references therein
ages are 46-59Ma

zircon FT 220 38-54 Shuswap Complex [Lorencak et al., 2001]

apatite FT 120 45-85 Lewis Thrust footwall 45-80, hanging wall 65-80 [Osadetz et al., 2004]

apatite FT 120 28-49 Shuswap Complex [Lorencak et al., 2001]

Valhalla Complex, Valhalla Complex 26-44Ma, Monashee unpublished data, R. Parrish.


apatite FT 120 21-44
Monashee Complex north 45 ±3 Ma, west 21-27Ma (Included in this study)

Foothills 52-73 Ma, Front Ranges 30-46


apatite FT 120 14-73 Rocky Mountains Ma, Trench and Purcell thrust area: 14-25 [McDonough et al., 1995]
Ma, western Selkirk Allochthon: 42-48 Ma

modern detrital river


sediments sampled
Ages in/near RMT Oligocene to upper unpublished data, G. Foster, A.
form numerous
apatite FT 120 28-48 Eocene, Rocky Mountains and Foothills as Carter and R. Parrish. (Included in
locations across the
well as Interior Plateau lower Eocene this study)
Columbia and Rocky
Mountains

Abbreviations used: hbl = hornblende, musc = muscovite, biot or bt = biotite, FT = fission track

42 | P a g e
Introduction Chapter 1

1.3.4 Sedimentary Basins

Before the development of the thermochronology technique to estimate


exhumation and removal of rock cover, one of the best ways to estimate paleo-
elevations was to reconstruct the removed rock volume by looking at the volume of
sediment in foreland basins. The main foreland basin for the eastern Canadian
Cordillera is the Western Canadian Sedimentary Basin (WCSB), which extends
eastward from the Rocky Mountain foothills in Alberta for 500-1000 km into the plains
of Saskatchewan and Manitoba [Wright et al., 1994]. The abundance of hydrocarbon
resources means this basin is one of the best studied foreland basins in the world, with
close to 8000 boreholes providing excellent 3D resolution of its composition and
thermal structure [Wright et al., 1994]. Borehole analysis has shown that here the
depth of the base of the Jurassic, and thus the start of the ‘foreland’ nature of this
basin because of the onset of the orogeny, lies at just over 4000 m (Wright et al., 1994
and references therein). Three areas where the total sedimentary fill exceeds 3000 m
are often described as important sub-basins: The Liard Basin (L) at the northern end of
the north eastern Cordillera (Northern Canadian Rockies), the Alberta Basin (A), in
older publications also referred to as the Rocky Mountain Trough and further to the
south east the Williston Basin (W), the main extent of which lies south of the US
border.

The depositional ages of the sedimentary succession found today in the WCSB
ranges from Cambrian to upper Cretaceous, and only locally few Oligocene deposits.
This absence of younger sediments in the foreland basin at first glance contradicts the
theory of younger uplift, as this would most certainly have led to increased erosion
and thus deposition of more post-Cretaceous Sediments. However, thermal history
models based on borehole vitrinite reflectance and fission track data from inside the
WCSB show that a considerable amount of sedimentary cover was removed since the
end of the Palaeocene and the basin thermally exhumed. While the overall Williston
basin data infers only 200 m of average removed cover, the boreholes along the

43 | P a g e
Introduction Chapter 1

foothills show between 1 – 1.3 km of erosion since the upper Palaeocene [Wright et
al., 1994]. A location near the RMT at 55°N (just south of the Peace River Arch
basement dome) was even interpreted to have had up to 5.5 km of cover removed
over the past 65 M.y. [Nurkowski, 1985; Kalkreuth and McMechan, 1988; Issler et al.,
1990; Osadetz et al., 1990; Morrow et al., 1993]. These studies also inferred that for all
locations except the one near the RMT the sedimentary cover at some point must
have contained up to 1.2 km of Palaeocene sediments which were subsequently
eroded away. This would have been caused either by increasing the surface elevation
(i.e. uplift) or lowering the local base level, both of which would have also prevented
the deposition of younger sediments.

To the other side of the eastern Cordillera lies the Interior Plateau, which with
an average elevation of approximately 1100 m and thus almost 3000 m lower than the
highest peaks of the range, should also provide an excellent basin platform where the
westward draining rivers could deposit sediment during and after the orogeny. Yet
there are no major sedimentary basins containing rocks older than Cretaceous in age
present here either, and more importantly no known deposits of Eastern Cordilleran
origin exist in the entire area. This can be explained by a major drop in the fluvial base
level in the Interior Plateau, which locally dropped up to 1000m during the Palaeocene
to Oligocene. Between 60 and 35 Ma and thus likely related to the phase of tectonic
extension and rise of core complexes, this dramatic drop caused incision of deep
canyons into the late to early Eocene surface and resulted in almost no deposition. A
slight increase in base level post-Eocene caused Oligocene and Miocene sediments to
be deposited locally in smaller quantities and partially fill the Eocene canyons [Tribe,
2005], yet no new large sedimentary basin was established which could be used as
evidence for a younger tectonic event in the hinterland.

Finding potentially younger (<45 Ma) deposits to undermine a hypothesis of


increased upper Cenozoic exhumation and uplift is furthermore inhibited by the fact
that paleo-sediment transport directions cannot be directly inferred from modern
rivers. Currently all rivers of the Interior Plateau, the eastern Coast Mountains and the
northern Columbia Mountains drain southward towards the Fraser Delta and the

44 | P a g e
Introduction Chapter 1

Figure 1.3-1: Map of the Western Canadian Sedimentary Basin, redrawn after [Wright et al., 1994]. Contour lines
show the thickness of the Phanerozoic cover from the present day surface to the basement in 1000m intervals.
Areas with depth greater than 3000m appear darker and are separated into sub-basins: L: Liard Basin. A: Alberta
Basin and W: Williston Basin. Black markers and numbers indicate thickness of removed sediment cover (in km)
since the end of the Paleocene, based on borehole vitrinite reflectance and fission track data (as compiled by
Wright et al. 1994). The major rivers feeding these basins are annotated in blue, while the red dashed line shows
the trace of the Rocky Mountain Trench (RMT) and the green dashed line marks the extent of our study area.

Pacific via the Fraser River drainage, while the Southern Columbia Mountains (Selkirk
and Purcell Ranges) drain southward to the US via the Columbia River. However, it has
been suspected for over a century that the Fraser River used to drain to the north
[Dawson and McConnell, 1895; Lay, 1940] and recent studies using modern dating
techniques have confirmed that over the last 8 M.y. local drainage directions gradually
reversed in the Interior Plateau from Northward to Southward [Tribe, 2005; Andrews
et al., 2011; Andrews et al., 2012]. Although locally dams created by basalt flows and
southward tilting of the existing land surface have been found to have caused this
reversal , the exact cause for the overall directional change or the complete ancient
path of the Fraser River remain unknown [Andrews et al., 2012]. Yet any sediment
which was transported into the Plateau before the reversal would have to have
followed said path and thus finding potential larger scale Oligocene and Miocene
deposits from the Eastern Cordillera (if they exist) is very difficult.

45 | P a g e
Introduction Chapter 1

Despite the apparent lack of Cenozoic sedimentary deposits found in


surrounding basins, the steep modern physiography of the south eastern Canadian
Cordillera, existing paleoecology data along with tilted Neogene basalt remnants as
well as previously unpublished post-Eocene low-T thermochronology data presented
in this chapter clearly indicate that the recent exhumation history of these mountains
warrants further investigation. This study will present a new and comprehensive low
temperature thermochronology dataset to shed light on the Cenozoic cooling ad
exhumation of these ranges. The following chapters will first introduce the
thermochronology methods applied in more detail (Chapter 2) and then present the
observed apatite and zircon (U-Th)/He and apatite fission track results for the south
eastern Canadian Cordillera (Chapters 3 and 4). This new large scale thermochronology
dataset will subsequently be used to investigate feasible scenarios of geothermal
gradients and erosion rates which might have caused the observed exhumation
(Chapter 5) before conclusion regarding potential uplift and the most likely geological
causes can be drawn and a summary of the Cenozoic exhumation history of the south
eastern Canadian Cordillera can be given (Chapter 6).

46 | P a g e
Methods of Thermochronology Chapter 2

Chapter 2:
Methods of Thermochronology

2.1 Underlying Principles

The term thermochronology describes the science of dating the thermal history
of a rock. It uses products of radioactive e decay from naturally occurring isotopes
where the retention of daughter isotopes or preservation of lattice damage is
thermally controlled. It was first introduced in the 1960s [Hurley, 1955; Mason, 1961;
Hurley et al., 1962; Hanson and Gast, 1967] and has since evolved and expanded, with
intensive research deepening our understanding of the underlying physical principles
and potential caveats of those methods and the introduction of new minerals to
existing techniques.

Figure 2.1-1: Sensitivity temperature range of retention/annealing zones for available thermochronometers
which are currently used. The three low temperature methods applied in this study are highlighted in grey.
Redrawn after [Gehrels et al., 2002]

47 | P a g e
Methods of Thermochronology Chapter 2

Several thermochronological methods are routinely used (Figure 2.1-1), which


span a range of measurable paleo-temperatures from 40°C for the lowest temperature
technique (U-Th)/He dating of apatite up to 900 °C for U-Pb dating of zircon [Reiners et
al., 2005]. The two main low-temperature thermochronometers are applied in this
work: (U-Th)/Helium Dating and Fission Track Dating. Both methods will be described
in this chapter.

2.1.1 Closure Temperature and Annealing Behaviour

Before delving into the detailed methodologies and the minerals used for this
study, a few basic principles which all thermochronological methods have in common
have to be established. The earliest work explored the concept of mineral- and decay -
process -specific closure temperature (Tc) [Dodson, 1973]. It was shown that at high
temperatures the decay products, such as 4He produced through α-decay or fission
tracks induced by spontaneous fission are constantly lost. Due to thermal energy in
the system which allows for 4He diffusion out of the mineral or healing of fission tracks
through vibrations in the atomic bonds of the crystals. These vibrations allow the
displaced atoms to jump back into their original position, thus ‘healing’ the damage. As
the temperature decreases, these processes start to be inhibited by a lack of energy
and the rate of diffusion or annealing decreases, causing an increase in the rate of
retention of daughter isotopes or preservation of tracks. This process continues until
isotopes are in effect quantitatively retained by the mineral, and tracks are perfectly
preserved in the crystal lattice. At these lowest temperatures the rate at which the
daughter products are formed is the same as the rate of production of daughter
isotopes/tracks in relation to the mineral and process specific decay constant [Reiners
et al., 2005; Zeitler et al., 1987]. This means that by calculating the time, t(0), where
the preservation of daughter products began, one can calculate the time since cooling
below the closure temperature. According to Dodson, the closure temperature Tc can

48 | P a g e
Methods of Thermochronology Chapter 2

be described as the temperature where quantitative retention began and the system
chemically ‘closed’ (Figure 2.1-2 A).

However, the temperature at which a system ‘closes’ depends on many factors.


Most important are cooling rate, cooling history (for samples which did not initially
cool from sufficiently high temperatures) and diffusional or annealing behaviour (and
thus mineral chemistry and grain size as well). For example the closure temperatures
reported for the methodologies relevant in this study range from 55-100°C for (U-
Th)/He dating of apatite [Farley, 2000; Farley, 2002] and 160-200 °C for (U-Th)/He in
zircon [Reiners et al., 2000; Farley and Stockli, 2002; Reiners et al., 2004], while fission
track dating of apatite has a reported closure temperature range of 90-120 °C [Reiners
et al., 2005; Ketcham et al., 1999] (Figure 2.1-2 B).

A B

Figure 2.1-2: A: Illustration of the closure temperature concept from Dodson (1973). As the temperature in the
system decreases the daughter/parent ratio D/P increases until a steady rate of growth proportional to the
decay constant for the number of parent isotopes present is reached. Backward projection of this slope onto the
time axis gives the observed time tc of when the system apparently ‘chemically closed’ and the age was
recorded. The corresponding temperature experienced by the system at t c is defined as the closure temperature
Tc. B: Closure temperatures in relation to cooling rate for the three methods used in this study [Reiners and
Brandon, 2006].

These large ranges were obtained by numerically computing effective closure


for grains of various sizes, chemical compositions and cooling histories and using
diffusion/annealing measurements on the minerals in question. For the purpose of

49 | P a g e
Methods of Thermochronology Chapter 2

first order approximation the values generally used to describe ‘cooling below X °C
since Y Ma’ are based on the properties of common mineral standards (utilized to
assess the accuracy of our data and cooling rates). Unless there is evidence to suspect
different cooling rates or mineral chemistry, as a first order approximation we often
assume ‘standard’ AHe, AFT and ZHe closure temperatures of 70°C, 120°C and 180°C
respectively [Farley, 2000; Ketcham et al., 1999; Reiners, 2005]. These estimates are
based on a typical average grain size of 60 µm equivalent spherical radii, average
mineral compositions and a cooling rate of 10°C/M.y..

In addition to introducing the closure temperature concept, Dodson (1973) first


described the temperature range where annealing or diffusion gradually slow down as
a ’continuous transition from total loss to undisturbed accumulation’. This concept is
still valid today, and now defined as the Partial Retention Zone (PRZ) for helium dating
[Wolf et al., 1998] and the Partial Annealing Zone (PAZ) for fission track dating
[Gleadow and Duddy, 1981; Gleadow et al., 1983]. The upper (low T) and lower (high
T) limits of these zones are generally defined by 90% and 10% of retention of the
produced daughter products in samples which are held at those temperatures for a
considerable amount of time [Reiners and Brandon, 2006]. Figure 2.1-3 shows these
‘loss only’ (no new production of 4He or tracks considered) PRZs/PAZs for the methods
applied in this study as a function of holding time for grains of average size and
chemistry, as computed by Reiners & Brandon (2006).

According to Wolf et al. (1998) samples which reside in the PRZ/PAZ for a long
time (how long depends on the temperature) will eventually equilibrate and reach a
steady state at which the rate of production and the rate of daughter loss cancel each
other out and the sample yields a steady age (Figure 2.1-4). Thus the upper and lower
limits of the PRZ/PAZ can also be defined as the temperature range where the
resulting measured age will be between 10% and 90% of the original age. In this study
not only loss but also ongoing production of daughter products were considered and
the resulting PRZ/PAZ values were found to be slightly higher than in Figure 2.1-3, with
e.g. their AHe PRZ ranging from 40-85°C for a holding time of 10M.y. in contrast to 25-
50°C observed by Reiners and Brandon (2006) for the same holding time.

50 | P a g e
Methods of Thermochronology Chapter 2

Figure 2.1-3: Modelled loss-only (ignoring renewed production of He or Ft) partial retention/annealing zone
(PRZ/PAZ) temperatures for AHe, AFT and ZHe dating vs. holding time at these temperatures. Upper (higher
temperature) limits show 10% of retention of daughter products while lower limits (cooler temperatures) are
drawn at 90% retention of daughter products. Redrawn after [Reiners and Brandon, 2006].

Figure 2.1-4: Development of helium ages of samples which are held isothermally for a prolonged amount of time
[Wolf et al., 1998]. Depending on the isothermal temperature at which a sample is held a steady state age is
reached after a specific holding time. Regardless of initial age (top: 0 Ma, bottom: 100 Ma) the same equilibrium
age is reached for equivalent temperatures. The only difference is the holding time required to reach the steady
state.

51 | P a g e
Methods of Thermochronology Chapter 2

As these varying approaches show, reported values vary, mostly based on the
application of differing diffusion kinetics when computing these ranges. The cited
ranges which were computed for Durango apatite chemistry are given as 40-85°C
[Wolf et al., 1998] for the AHE PRZ, 70-110°C [Ketcham et al., 2000] for the AFT PAZ
and 130-190°C for the ZHe PRZ [Reiners, 2005; Reiners and Brandon, 2006]. When
interpreting thermochronological data using thermal modelling (to be discussed in
section 2.5) the preferred model for the anticipated geological setting, mineral and
mineral chemistry must be chosen.

A B

Figure 2.1-5 Example on how different cooling histories which all result in an average Helium age of 40Ma at the
surface (Figure A) produce distinctly different age vs. depth relationships in a theoretical modelled vertical
section (Figure B). In (A) dotted lines display the evolution of the average age over time while the solid line
depicts the temperature path taken. The relationships in (B) show what ages could be expected from samples at
depth for each of the Histories presented in (A) [Wolf et al., 1998]. The example shown was computed for the
AHe method, but similar results can be expected for all other thermochronometers. The depths were calculated
for a geothermal gradient of 20°C/km.

Clearly interpreting cooling ages is rarely as simple as assigning a temperature


to a specific point in time based on the measured age. The average age is always the
result of the evolution of the daughter/parent ratio over time which is influenced by
residence times at higher temperatures (PAZ or PRZ), cooling rates and mineral
chemistry (e.g. U and Th content). Unless rapid cooling through all relevant
temperatures occurred (as could be the case for extrusive volcanism for example), an
observed age rarely relates to a specific geological event in time. To illustrate this,
Figure 2.1-5 shows how a range of different cooling histories (A) results in the

52 | P a g e
Methods of Thermochronology Chapter 2

development of different average AHe age as the sample is being exhumed (B) and yet
all 5 histories yield an average age of 40 Ma (Figure from Wolf et al 1998).

2.1.2 Combining Methods and Sampling Vertical Transects

One of the most useful approaches for determining the thermal history of rock
units is the combination of multiple mineral thermochronometers. Figure 2.1-1 shows
the wide range of temperature ranges that can be covered by the available methods.
When combined with an appropriate geothermal gradient, the thermal history can be
converted into a quantifiable burial and/or exhumation history.

Furthermore, applying multiple thermochronometers to multiple samples from a


(near) vertical transect in mountain ranges or from boreholes adds valuable
constraints to the thermal history. At its simplest, cooling ages can be plotted against
elevation so the gradient corresponds to exhumation rate, assuming a constant
geothermal gradient. Samples from mountain sides typically have a significant
horizontal distance component. Thus interpreting data, in particular AHe ages, in
purely exhumation terms requires knowledge of topographic development.
Furthermore, where exhumation is rapid (e.g. > 1.0 mm/yr), the assumption that low
temperature isotherms in the crust remain horizontal may not be appropriate.
Numerical models have shown that rapid changes in burial depth (due to e.g. incision
or structural exhumation/fault movements) distort the thermal field and decrease the
distance between isotherms and/or bring them closer to the surface [Ehlers, 2005].
Steep topography will also influence the subsurface horizontality of near surface
isotherms.

Despite these additional considerations, using multiple thermochronometers


and sampling from vertical profiles to determine age-elevation relationships remains
one of the main advantages of the methodologies. For this project, both approaches
have been utilized and will be discussed in the relevant chapters. For most samples
both AHe and AFT data was acquired. In a smaller number of samples ZHe ages were

53 | P a g e
Methods of Thermochronology Chapter 2

also determined. Several near-vertical profiles of up to 2.5 km relief were sampled.


The complications mentioned above, such as potential distortion of the thermal field
and local thermal or structural variations will be discussed in greater detail with the
interpretation of the results in chapter 5.

2.2 General Sample preparation

The minerals used in this study are apatite (Ca5(PO4)3(F,OH,Cl)) and zircon
(ZrSiO4). Both are accessory phases present in low abundances in most igneous and
metamorphic rocks, and both can arrive in sediment as a detrital mineral phase,
although zircon is much more common. Both minerals are commonly rich in uranium
and thorium, making them ideal for radiometric dating techniques.

Most samples were collected from intrusive rocks or gneisses of broadly


granitoid composition (granite to tonalite and some more mafic phases), and
metasedimentary rocks such as schists and quartzites. Sample locations were chosen
based on accessibility and spacing along sampling profiles. Care was taken to avoid
sampling rocks that might have been reheated to temperatures above their closure
ranges by hot springs or recent volcanic activity. At each location approximately 1.5 kg
was collected and care was taken to avoid weathered surfaces. A full list of samples
with more detailed rock descriptions can be found in Appendix 1.

Mineral separations were conducted using standard heavy-liquid separation


techniques using the mineral separation laboratories of NIGL Keyworth and at
Birkbeck College, London. A total of 58 samples were crushed using a jaw crusher and
subsequently milled using a disk mill. After sieving, material smaller than 350 µm was
separated further using a Rogers shaking table which concentrates minerals of high
specific gravity. The dense fraction was passed through a Frantz magnetic separator at
multiple field strengths up to 1.7 A. The non-magnetic fraction passed through heavy
liquid separation using Bromoform (2.84 g/cm3) and Diiodomethane (3.31 g/cm3) to
remove residual quartz or mica and to obtain pure apatite (density of 3.2 g/cm3) and

54 | P a g e
Methods of Thermochronology Chapter 2

zircon (density of 4.7 g/ cm3) separates. Further hand picking of grains under a
microscope was done to produce a 100% pure mineral separate, and to choose the
grains for analysis. This sample set was complemented by samples, mostly separates,
from existing sources. These are indicated in the sample list (Appendix 1) and where
appropriate in the following chapters. These mineral separates were obtained using
the same standard techniques.

2.3 (U-Th)/Helium Dating

2.3.1 Age Calculation and Temperature Sensitivity

238 232 235


(U-Th)/He dating utilizes the radioactive decay of U, Th and U which
produces 8, 7 and 6 α-particles (4He) respectively, before the stable isotopes 206
Pb,
208
Pb and 207Pb are produced. A further He contribution is from the decay of 147Sm to
143
Nd with the production of 1 α-particle. Because of the much smaller decay constant
(only 6.54x10-12y-1 compared to 1.55x10-10y-1 for 238U) the contribution is generally less
than 1% of the measured helium and thus mostly negligible. At high temperatures the
thermal energy in the system allows for He diffusion which can escape the crystal
lattice of the host mineral. The energy in the system causes vibrations within the

238 235 232 147


Figure 2.3-1: α-decay of U, U, Th and Sm produces Helium atoms which accumulate within the crystal
lattice of the host mineral.

55 | P a g e
Methods of Thermochronology Chapter 2

atomic bonds, helping the helium atoms to overcome the barrier energy holding them
in place at a mineral specific ‘site’ (a position where the atom can be stored under
surface temperature conditions) and jumping towards the next one [Gautheron and
Tassan-Got, 2010]. In this manner helium can migrate through the crystal and
eventually be lost if it reaches the edge, escaping into a fluid phase and leaving the
rock volume.

With decreasing temperature and thus decreasing energy in the system this
diffusion becomes more and more sluggish, until with sufficiently low temperature He
is quantitatively retained (Figure 2.3-1). Consequently, by measuring the produced 4He
concentration and the concentration of parent isotopes 238U, 232Th and 235U the time
at which the system dropped below Tc can be calculated using the following
relationship:

𝟒
Equation1: 𝐇𝐞 = 𝟖 𝟐𝟑𝟖𝐔(𝐞𝛌𝟐𝟑𝟖 𝐭 − 𝟏) + 𝟕 𝟐𝟑𝟖𝐔 / 𝟏𝟑𝟕. 𝟖𝟖(𝐞𝛌𝟐𝟑𝟓𝐭 − 𝟏) + 𝟔 𝟐𝟑𝟐𝐓𝐡 (𝐞𝛌𝟐𝟑𝟐𝐭 − 𝟏)

Where λ238, λ235 and λ232 are the specific decay constants of 238U, 235U and 232Th
respectively, t is the time since cooling below Tc and the amount of 235U is calculated
using the ratio of 238U/235U = 137.88 of naturally occurring uranium.

Tc for (U-Th)/He dating is generally cited as 70°C (PRZ 45-80°C) for apatite and
180 °C (PRZ 130-190°C) for zircon. However, detailed methodological stepwise heating
experiments have shown that closure temperatures range from 55-100 °C for apatite
[Farley, 2002] and 160-200 °C for zircon [Reiners et al., 2000; Farley and Stockli, 2002].
They were found to depend on various factors including cooling rate, grain size and
mineral chemistry (e.g. chloro-, hydroxy- or fluoro-apatite). The latter is difficult to
determine in each grain which is to be analysed for helium without destroying it.
Therefore uncertainty in mineral chemistry is a caveat which has to be considered
when interpreting cooling ages. If the average chemical composition of a sample can
be determined independently (e.g. by measuring D-par (track-surface-interception
aperture) on fission track mounts produced for the same sample or analysing the same

56 | P a g e
Methods of Thermochronology Chapter 2

mounts by microprobe analysis, see section 2.4.6 for more detail), and the results
indicate ‘normal’ fluor-apatite chemistry, standard closure temperatures can be
assumed as a first order approximation. When modelling the thermal history of
samples using standard thermal history programs such as HeFTy (section 2.5), a choice
of diffusion models is offered and together with the grain measurement and U and Th
content calculate all possible cooling paths for the diffusion parameter implied –
rather than simply assuming one Tc and correlating this to the age. This is especially
valuable in combination with fission track length data, and thus will be discussed in
subsequent sections.

2.3.2 Alpha Recoil Correction

A major component of the (U-Th)/He method is correcting for the potential


loss of radiogenic 4He (alpha particles which are produced by decay) near the edge of
the crystals hosting the parent isotope. The energy released during α-decay causes
recoil of the 4He away from the parent nucleus. The distance the 4He atom recoils
varies with each decay series due to the differing energies released. In apatite
radiogenic 4He from 238U-series decay has an average stopping distance of 19.7µm,
while α-particles created by the 235U- and 232Th-decay series display mean stopping
distances of 22.8 µm and 22.4 µm respectively. The denser crystal structure of zircon
means recoil distances are slightly shorter with averages of 16.6 µm, 19.6 µm and 19.3
µm for 238U, 235U and 232Th respectively [Farley et al., 1996a]. Therefore upon
production α-particles will recoil and come to rest anywhere within a theoretical
sphere, with the centre being the parent nucleus and the radius equalling the mean
stopping distance (Figure 2.3-2). This is termed the recoil sphere.

Thus, if the parent nucleus was located exactly at the edge of the host crystal,
the 4He atom would have a 50 % chance of being ejected out of the crystal upon
production. The probability of this occurrence reduces if the parent nucleus is located
further from the edge, as the recoil sphere moves further inside the grain. It reaches 0

57 | P a g e
Methods of Thermochronology Chapter 2

% and thus complete retention of 4He if the distance of the parent nucleus to the grain
edge exceeds the stopping distance. This gradient has to be taken into account when
comparing measured parent isotope concentrations to measured helium, because the
parent concentration includes parent isotopes whose decay products might have been
ejected out of the grain. Consequently if this phenomenon were not considered, it
would result in a significant underestimation of the calculated age.

Figure 2.3-2: Schematic figure of α-ejection, retention and possible implantation [Farley, 2002] based on the
positioning of the recoil sphere.

Techniques developed in the early stages of (U-Th)/He dating attempted to


remove this bias by mechanically abrading or chemically etching the outer ‘shell’ of
potential Helium loss away. Modern studies however apply quantitative modelling to
correct for this He loss based on grain dimensions and stopping distances. To obtain
the recoil correction factor Ft, the dimensions of each grain have to be measured
carefully in order to calculate the surface area and volume of each grain and thus
determine the volume of the ‘shell’ of potential helium loss. Thus an obvious caveat of
this method is the assumption that the parent nuclei are uniformly distributed
throughout the host grain. Assuming this to be true, in a first approximation for a

58 | P a g e
Methods of Thermochronology Chapter 2

spherical host grain the ejection can be calculated using equation 2 [Farley et al.,
1996a]:

Equation 2: FT = 1 – 0.75 (s/r) + 0.0625 (s/r)3

Where s is the mean stopping distance and r is the radius of a theoretically spherical
grain. Figure 2.3-3 depicts how this relationship evolves with changing r and shows
that the difference between the individual decay series is minimal.

Naturally apatite grains have a more complex geometry in nature, and thus Ft
has to be estimated using the surface to volume ratio β:

Equation 3: Ft ≈ 1 – β (s/4)

Farley et al. (1996) have shown that this relationship estimates Ft well for β >
0.07, but underestimates it for β < 0.07. Thus it is important to choose grains of
sufficient size and euhedral approximately cylindrical shape to guarantee that the F t
generated is representative. For apatites this implies choosing grains with a width of
minimum 70 µm (90 µm is preferred where possible, but depending on grain quality
this could not always be guaranteed) and avoiding oblate grains at all costs. Zircon
grains can be smaller (minimum width of 40 µm) since the stopping distances are
shorter and oblate grains are also much less common.

Figure 2.3-3: Numerical Ft factor model results for theoretical spherical grains with varying radii. This illustrates
that retention is similar for all decay chains and highly dependent on the size (radius) of the grain [Farley et al.,
1996b].

59 | P a g e
Methods of Thermochronology Chapter 2

The estimated Ft factor is then applied to the uncorrected age (as given by
Equation 1) using the following relationship:

𝑀𝑒𝑎𝑠𝑢𝑟𝑒𝑑 𝐴𝑔𝑒
Equation 4: 𝐶𝑜𝑟𝑟𝑒𝑐𝑡𝑒𝑑 𝐴𝑔𝑒 = 𝐹𝑡

It has to be noted that the α-ejection process can also be reversed, i.e. 4He could be
implanted into the grain (as depicted in Figure 2.3) by a parent nucleus located in a
directly adjacent grain within the original rock. When dealing with mineral separates
which have been released from their original crystalline rock assemblage, this is
impossible to determine and thus has to be considered as a caveat in the method.

2.3.3 Standards

In order to ascertain the accuracy of the helium data, in-house mineral


standards of known age were measured and dated along with each batch of samples.
For apatites the commonly used standard is the Durango apatite with an accepted age
of 31.6 Ma [Jonckheere et al., 1993]. Because Durango age standard grains are
fragments of one larger crystal and thus exclusively from the inside, so no Ft-correction
has to be applied. For zircon the standards used were Fish Canyon Tuff zircons with an
accepted age of 27.3 Ma [Reiners et al., 2002]. FCT zircons are individual crystals and
thus require Ft-correction.

2.3.4 Apatite Grain Selection

In preparation for (U-Th)/He analysis euhedral apatites with a grain size smaller
than 300 μm were picked from mineral separates using a 225x cross-polarized Nikon
SMZ1500 binocular microscope. Of the 58 mineral separates available only 32 proved
to contain apatites of sufficient quality and size for (U-Th)/He dating. For each of these

60 | P a g e
Methods of Thermochronology Chapter 2

samples three to five crystals were chosen and dated by single grain analysis to
quantify the reproducibility of bedrock samples.

Great care was taken to choose inclusion-free grains to avoid effects of He-
implantation from inclusions of minerals with higher U-Th concentrations. Similar to
the potential implantation of helium from crystals formerly in contact with the now
separated mineral grain (as previously discussed in section 2.3.2), common U- or Th-
rich inclusions within apatites such as zircon or monazite needles as well as fluid
inclusions of unknown composition could potentially add helium. This implanted
contribution of α-particles would also be considered ‘parentless’ because such
inclusions of more resistant minerals will not dissolve in standard apatite dissolution
procedures required in preparation for chemical analysis using the ICP-MS. Therefore
they would not contribute U and Th to the parent isotope measurement, resulting in
an overestimation of the age. Until recently, a lack of inclusions was believed to be
one of the most important criteria for grain selection. However, although numerical
models have shown that inclusions are not trivial as a source for excess or ‘parentless’
helium, they have been proven to contribute generally less than 5 % (unless the
inclusions take up more than 10 % of the host grain) and so are negligible for most
naturally occurring apatites [Vermeesch et al., 2007]. Nevertheless inclusion-free
grains were preferably chosen where available to avoid any possible contamination
altogether.

Furthermore, where grains with two intact terminations were selected as


broken ends expose internal surfaces which complicate the Ft-correction as it cannot
be determined when the grain broke (i.e. did ejection occur through the broken end or
did it break during the mineral separation process and we are overcorrecting for
ejection as a result?). Examples of apatites chosen/not chosen for analysis and
two/one termination grains are shown in Figure 2.3-4.

Most grains had a minimum diameter of 60-90 μm and lengths of 90-230 µm.
The x- and y- grain dimensions of each were measured on digital photographs using
the GXCapture software. The third dimension was not measured directly; however

61 | P a g e
Methods of Thermochronology Chapter 2

care was taken to only pick grains which have similar depth and width. To simplify the
grain volume calculation it was then assumed that depth equals width. Oblate grains
were avoided. The photographs and dimensions of all grains picked can be found in
Appendix 2 (accompanying CD).

2.3.5 Apatite Analysis

Following careful selection and measurement, the grains were individually


packed in Pt-tubes. The arithmetic mean for all aliquots from each sample was
determined to find the mean sample age and the standard deviation of the single grain
ages was used to provide the uncertainty and thus assess the precision of the mean
age. Analysis took place at the (U-Th)/He laboratory at SUERC, East Kilbride using a in-
house designed gas extraction line (Figure 2.3-5). Pt tubes containing the sample
aliquots were heated with a 808 nm iodide laser beam to approximately 900-1000 °C,
degassing the crystals for 60 seconds [Foeken et al., 2006]. The active gas was
removed by exposure to active charcoal cooled by liquid nitrogen. After 5 minutes the
residual gas was admitted into the Hiden HAL3F quadrupole mass spectrometer for
analysis. The measured voltage was converted to an absolute value in volume of He by
peak comparison with a 4He standard calibration volume. The measured values were
blank-corrected and blank levels were generally less than 4% of the sample signal. To
ensure the crystals have degassed fully a re-heat of each aliquot was performed and
accepted if the measured 4He was less than 5% of the initial measurement. Otherwise
further reheating was performed until this could be confirmed and all reheat
measurements were added up to account for the total helium released.

Following the helium measurements the samples were dissolved by


submerging the Pt tubes in 10% HNO3 overnight. The solutions were spiked with 30 µg
of HNO3 containing known concentrations of 235U and 230Th to enable the analysis of
238
U and 232Th content using the Inductively Coupled Plasma Mass Spectrometer (ICP-

62 | P a g e
Methods of Thermochronology Chapter 2

1 100µm 1 100µm

A B

100µm
100 µm 100µm
100 µm

C D
Figure 2.3-4: Apatite grain examples, photographed at different magnifications using the Nikon SMZ1500
binocular microscope at Leicester. A: Array of good quality euhedral apatite grains, some containing inclusions
(red arrows pointing towards examples). B: Example of a poor quality sample where apatites are not euhedral
but rather broken shards. Samples like these were not analysed. C: Example of a grain chosen for analysis. The
grain displays a sub-euhedral shape (slightly rounded edges, typical for plutonic rocks) and two intact
terminations. D: Example of a one-termination grain chosen for analysis.

MS) at SUERC. With this technique the sample solutions were converted into an
aerosol in a stream of argon, ionised using a plasma torch and subsequently
accelerated into a quadrupole mass spectrometer where the intensities of desired
masses 235, 238, 232 and 230 were measured in counts per second (CPS). By
comparing the peak intensities the 235/238 and 230/232 isotope ratios can be
determined and because the concentrations of masses 235 and 230 are known the

63 | P a g e
Methods of Thermochronology Chapter 2

desired quantities of 238U and 232Th can be calculated. In some cases the spiking
procedure was repeated with a solution of known 149Sm concentration and the 147Sm
sample content determined in the same way.

Figure 2.3-5: Schematic of the (U-Th)/He line at SUERC, East Kilbride. Samples are stored in sample chamber
(green) under high vacuum maintained by the turbo pump and ion pump 1 (purple) and heated with the ion
beam laser to degas. The gas will equilibrate in the chamber and manifold sections (blue, valves V5, V9, V10 and
V1 closed). After 5 min the mass spectrometer section is opened (V1), allowing the gas to equilibrate with the
line. V1 is then closed and the gas composition in the pink section measured using the Quadrupole Mass
Spectrometer. To empty this section after measurement it is pumped empty to high vacuum again using the
second ion pump (opening V3 or V7). Equally after heating a sample the manifold and chamber sections are
pumped by opening V5. At the beginning and end of each session and between every 2 samples a calibration is
4
performed by measuring a known quantity (pipette volume) of He from the calibration tank (yellow section)
using the same procedure minus the laser heating. This not only allows for evaluation of consistency of
measurements over longer periods but is also used to determine the absolute value of the returned
measurements by peak comparison with the standard.

2.3.6 Zircon Grain Selection

Between 7 and 10 zircons from 8 samples were selected for each sample as
described in section 2.2.4. All grains had a minimum width of 40 µm and the longest

64 | P a g e
Methods of Thermochronology Chapter 2

grain measured ~400 µm in length. Photographs and measurements of all analysed


grains can be found in Appendix 2 (accompanying CD). The chosen crystals were
predominantly colourless with both terminations preserved. Because of the typically
high U and Th content, a complete lack of inclusions is not a major criterion for zircons,
as inclusions are typically small and thus deemed to have little influence on the
resulting age. Larger inclusions were avoided, yet many selected grains inevitably
contained 1-15 inclusions of sub micrometre scale.

As with apatite, the potential ejection of 4He along the edges of the zircon
crystals has to be considered when calculating the helium age based on the
daughter/parent ratio. For the calculation of the Zr Ft factor the alpha-Ft-ejection
factor program provided by Cecile Gautheron (Université Paris Sud, program can be
downloaded at http://hebergement.u-psud.fr/flojt/ [Gautheron and Tassan-Got,
2010; Ketcham et al., 2011a]) was used instead of the manual calculation described in
section 2.3.2. Although this program uses the same principle calculation, it allows the
user to specify the exact the grain shape (habit) including types and presence of
terminations of each grain before calculating Ft. The resulting volume and surface
areas calculated are more exact than the generalized grain shape formula previously
discussed, which is particularly important for zircon since the grains tend to be smaller
and thus a larger error in Ft could potentially increase the uncertainty in the age by up
to 30% (as simple calculation comparisons have shown). To produce the most accurate
ages possible the dimensions and shape of each grain were entered to calculate the Ft
factor for each individual grain.

2.3.7 Zircon Analysis

Zircons were individually packed into Pt-tubes and analysed at SUERC. The
procedure is only slightly different to that required for apatite analysis described in
section 2.3.5. Zircons were heated to approximately 1200 °C for 5 minutes and the gas
purification process took a further 10 minutes. Afterwards the 4He concentration was

65 | P a g e
Methods of Thermochronology Chapter 2

determined using the quadrupole mass spectrometer as previously described for


apatite analysis.

Zircons were first unpacked under a binocular microscope and transferred to


individual Teflon beakers. After spiking with known quantities of 230Th and 235U, the
samples were dissolved in a 50% solution of HF and HNO3 in a Teflon-lined steel bomb,
heated to 225°C overnight. The resulting solution was then analysed for U and Th and
Sm content using the ICP-MS at SUERC following the same procedure as previously
described for apatite in section 2.3.5.

2.3.8 Caveats

Many potential caveats, such as He-loss through ejection, He-gain through


implantation or high U-Th inclusions have already been discussed in the sections
above. While we can correct for the He-loss through alpha ejection using the Ft
correction factor, no such correction can be applied for He-implantation. However
studies have shown that this is a rare effect which would require direct contact with a
high U-Th mineral over a large surface within the original rock to impact the
measurement significantly [Brown et al., 2013]. Also by analysing multiple individual
grains for each sample, potential outliers or ‘rogue grains’ which might have been
influenced by processes like implantation or large inclusions can be identified and
excluded from mean aliquot ages. However, several caveats remain that require
consideration before interpreting (U-Th)/He ages.

One of the biggest topics of current research in the theoretical


thermochronology community is the role of diffusion kinetics on ages. The physical
process of helium migration through the crystal lattice is barely understood and thus
estimating necessary diffusion coefficients such as the required activation energy Ea,
the relevant diffusion domain a and diffusivity at infinite temperature D0 are very
challenging. However, because all estimates of closure temperatures involve
assumptions about diffusivity, it is important to understand the underlying processes

66 | P a g e
Methods of Thermochronology Chapter 2

to improve our closure temperature estimates and thus our interpretation of


thermochronology data. Many recent studies have attempted to quantify diffusive
behaviour of helium in crystals by performing laboratory-timescale stepwise heating
and degassing experiments as well as numerical simulations to extrapolate their
results to geologically relevant timescales. Assuming that these extrapolations are
accurate, results indicate that kinetic coefficients vary depending on grain size
(diffusion domain a was found to equal the grain), grain shape (due to slight
dependency of diffusion on crystallographic orientation, more relevant for zircon than
apatite), cooling rate, presence of zonation in crystals as well as presence and amount
of radiation damage (discussed in more detail below)[Wolf et al., 1996; Farley, 2000;
Reiners et al., 2002; Guenthner et al., 2013]. However, most of these experiments
were conducted using standards or carefully selected samples and have to be treated
carefully when dealing with natural and often less chemically pure or
crystallographically perfect samples. For example: the activation energy Ea was
estimated to be around 33kcal/mol for Durango apatite [Farley, 2000], however values
of natural samples found by Wolf et al. (1996b) vary from 30-39 kcal/mol. Farley
(2000) also found that the deviation in diffusivity for apatite from a simple linear
Arrhenius relationship at low temperatures to lower diffusion rates above 265°C is
irreversible, suggesting that the currently used values might be incorrect for samples
that have experienced cooling, reheating to temperatures higher than that and
subsequent secondary cooling in their thermal history.

As hinted at above the presence of internal U-Th concentration variations, i.e.


zonation, can induce uncertainties in He age. Backscattered electron images of
particularly zircons but also many apatites commonly show at least some degree of
zonation caused by changes in chemistry during the initial growth of these minerals.
Consequently the distribution of helium produced in strongly zoned grains will be
inhomogeneous as well. This could imply zones of varying radiation damage and thus
varying diffusion characteristics as well as leading to over-or under-correction of
Helium loss by the Ft factor if the inner core and outer shell of the grain have
significantly different parent isotope concentrations. However, because of our current

67 | P a g e
Methods of Thermochronology Chapter 2

inability to mount and analyse chosen grains for zonation before attempting (U-Th)/He
dating we have no choice but to make a crude assumption of uniform distribution of
parent isotopes across the crystal.

2.3.8.1 Age Dispersion and Effective Uranium

While the above discussed issues such as implantation, inclusions and zonation
can explain individual outliers in datasets, they cannot fully explain the degree of age
dispersion which is sometimes observed in (U-Th)/He datasets. According to the
underlying theory of thermochronology, crystals originating from the same rock
sample should yield the same age as the aliquots have experienced the same cooling
history. In practice however, considerable age dispersion in aliquots originating from
the same sample is sometimes observed. In recent years studies of over-dispersed
datasets have revealed that one of the major controls on the variation of aliquot ages
is the total content of radiation damage from uranium and thorium [Flowers et al.,
2009; Flowers et al., 2007], which is defined as eU= 238U(ppm)+ 0.235 232Th(ppm). For
old samples or samples with extremely high U and Th concentrations, ongoing
radioactive decay including actinide series decay (α-, β- and γ-decay and resulting
cascading atom displacement) and spontaneous fission, can cause damage of the
crystal lattice [Weber et al., 1994; Ewing et al., 1995; Weber et al., 1997]. This can
reach an extent at which radiation damaged areas create vacancies and even
interconnect within the crystals. This process results in the formation of traps that
collect and retain Helium which enters them and thus potentially produce a falsely old
age in comparison to undamaged crystals [Shuster and Farley, 2009; Flowers et al.,
2009]. Furthermore it has been theorized (although not yet fully quantified) that with
extreme damage this effect could be reversed and diffusivity increased by potentially
forming interconnecting ‘tunnels’ through which the gas can escape [Farley, 2000;
Nasdala et al., 2004].

68 | P a g e
Methods of Thermochronology Chapter 2

The result of this relationship is a dependence of the closure temperature and


thus the observed (U-Th)/He age of each individual crystal on the uranium and
thorium content (eU) of that crystal. Figure 2.3-6 (A) shows the resulting PRZ ranges
for multiple realistic eU scenarios, ranging from 4-150 ppm, as calculated by Flowers et
al. (2009) using their ‘Radiation Damage And Annealing Model’ (RDAAM). If, for
example, a sample has been held at 70°C for 75 Ma (holding time of the model results
in this plot) and subsequently instantly cooled to surface temperatures, 3 apatite
grains analysed and found to have different eU concentrations of 28, 60 and 100 ppm
would yield (U-Th)/He ages of approximately 3 Ma, 15 Ma and 34 Ma respectively. The
resulting arithmetic mean age would be 17.3 Ma with a large 2σ error of ± 12.8 Ma.

A B

Figure 2.3-6: Dependence of closure temperature and (U-Th)/He age on eU concentration. (A): Temperature
ranges of (U-Th)/He PRZ for several apatite eU concentrations (4-150ppm), including the Durango standard
(dashed line) and the average apatite chemistry of 28ppm (thick black line) which is traditionally assumed to
apply to observed ages. For comparison the AFT PAZ is plotted as well (thick grey line). (B): The difference in
closure temperature (and thus PRZ range) decreases with increasing cooling rate, indicating that eU is a more
important control on slowly cooled samples while the dispersion caused in rapidly cooled rocks (>10°C/M.y.) is
minimal. The eU examples in plot B are 4, 10, 28, 60, 100 and 150ppm. Both figures redrawn after [Flowers et al.,
2009] where the curves were calculated using their RDAAM model.

This effect could explain many apparently over-dispersed aliquot assemblages,


particularly for slowly cooled samples. As Figure 2.3-6 (B) shows the difference in Tc
between samples of varying eU decreases with increasing cooling rate, indicating that
eU is a more important control on slowly cooled samples while the dispersion caused
in rapidly cooled rocks (>10°C/M.y.) is minimal. Over-dispersion in samples caused by
eU can be identified by plotting observed age vs. eU concentration and where a

69 | P a g e
Methods of Thermochronology Chapter 2

positive correlation can be established, looking at the individual grain ages could yield
more information about the samples’ cooling history than simply reporting the mean
sample age. Where for example a thermal history model is used to calculate the likely
range of cooling histories, modelling 3 different crystals with varying eU and thus
varying Tc individually can help narrow potential cooling paths and improve model
results (details of these models to be discussed in section 2.5).

Furthermore, the resulting mean age for samples with considerable eU


variation is almost always older than the ‘true’ age an ‘average’ apatite (28 ppm eU)
would display. Thus it could be the cause of inverted age relationships, both in
individual samples (where the AFT age for example is considerably younger than the
AHe age of the same sample) and age-elevation relationships of multiple samples
(where e.g. samples of higher elevation display older ages than lower elevation
samples).

The eU of samples are reported with the data in in chapters 3 and 4. For all
samples eU vs. age plots were created and where positive correlations could be
established a minimum of two individual grains were modelled during thermal
modelling (where applicable, using the built-in RDAAM annealing model by Flowers et
al. 2009) rather than using the mean age and chemistry. Where inverted age-
relationships were observed eU was analysed and reported if found to be a potential
cause for the observed disparity.

Lastly, we know from annealing experiments how the chlorine and fluorine
content of apatite can affect both etching and annealing behaviour due to the
difference in ion radii between the two atoms [Barbarand et al., 2003]. It seems
reasonable to suggest that variations in apatite chemistry beyond differences in eU,
could be causing a spread in ages as the larger ion radius of chlorine could influence
the diffusional behaviour of Helium within the crystal. Because most diffusion
experiments were conducted on Durango fluor-apatite [Farley, 2000], closure
temperatures for chlorine rich apatites might prove to be underestimated (as they
would require more energy for equal diffusion). This has not been studied to a great

70 | P a g e
Methods of Thermochronology Chapter 2

effect and needs further investigation within the thermochronology community. As we


are currently still unable to measure Cl/F content in each grain analysed for (U-Th)/He
by the standard laser and dissolution technique, we cannot report these values for the
crystals analysed for this study. However, recent developments in the technique such
as in-situ laser ablation (U-Th)/He dating [Vermeesch et al., 2012] might enable these
measurements in the near future.

Despite these caveats described above, (U-Th)/He thermochronology on


apatite and zircon remains a very useful method to address surface-near questions
related to exhumation and erosion of rock surfaces. The uniquely low temperature
sensitivity of these two methods allow us to identify small scale variations is cooling
history horizontally and vertically across the uppermost crust which allows for
conclusions regarding recent exhumation and erosional processes to be drawn. No
other thermochronological method is capable of resolving thermal patterns of the
recent geological past for the upper 3 km of the crust.

2.4 Fission Track Dating

2.4.1 Fission Track Formation

Highly energetic spontaneous fission of radioactive 238U atoms at 180° into


isotope fragments with masses between 140-150 and 87-93 produces small zones of
damage by rapid ejection of fission fragments, such as e.g. 56Ba. This explosive process
causes the atoms along its path to be ionized and pushed apart by the resulting mutual
repulsion, forming a track of amorphous material [Fleischer et al., 1965; Naeser, 1967;
Naeser, 1976; Gallagher et al., 1998]. In apatite this ‘path’ is typically 15-20 µm long
and approximately 8 nm wide and can be made visible under microscopic conditions
by exposing and etching internal crystal surfaces (Figure 2.4-1). Thermal energy allows
those tracks to ‘heal’, as the misplaced atoms can jump back into a proper

71 | P a g e
Methods of Thermochronology Chapter 2

crystallographic position. This process slows down as the temperature is lowered


through the Partial Annealing Zone (PAZ) and becomes negligible well below the Tc.

238
Figure 2.4-1: Schematic diagram of how spontaneous fission at 180° of U causes the formation of a fission
track. These tracks can be made visible by mounting crystals in epoxy resin, exposing internal surfaces by
polishing and etching the resulting surface to widen the tracks. An example of a microscopic view of this is
depicted on the right.

This study uses fission track dating of apatite, which has a closure temperature
range of 90-120 °C [Reiners et al., 2005]. As for (U-Th)/He dating this is a range rather
than a defined temperature. The reasons for varying closure temperatures in fission
track dating include cooling rate and importantly mineral chemistry, which will be
discussed in more detail below (Sections 2.4.6 and 2.4.7).

2.4.2 Sample Mount Preparation

For this study 46 apatite sample separates were prepared for fission track
analysis. In order to expose internal surfaces of the sample grains and make tracks
visible the mineral separates were mounted on a glass slide using Araldite epoxy resin.
Once hardened, the surface of the resin was polished with carbide paper and alumina
to expose internal surfaces of the grains on the slide. Great care was taken not to
apply too much pressure or grains might fracture and also not to remove too much
material, otherwise the grains might fall out of the mount. Ideally polishing would
reveal a substantial cross section of the grain.

72 | P a g e
Methods of Thermochronology Chapter 2

Once exposed, the sample mounts were submerged in 5 molar HNO3 for 20
seconds at room temperature to etch and reveal the zone of crystal damage caused by
spontaneous fission; in practice this visible etched zone is a few microns wide and up
to 18µm in length.

2.4.3 External Detector Method

Like in the previously described (U-Th)/He Method the age calculation of


Fission Track Dating is based on the ratio of parent isotope to decay product. While
the latter, i.e. Fission Tracks, can be simply counted under a binocular microscope to
determine the density of tracks per cm2, obtaining the 238U concentration relevant to
the section exposed is more difficult. While current research in Fission Track
Thermochronology is exploring the use of in situ Laser Ablation ICP-MS measurements
for Uranium analysis [Hasebe et al., 2004], the traditional external detector method
remains the most widely used technique and was thus also utilized in this study (Figure
2-8).

By this method, the polished and etched grain mounts were cut to
approximately 1 cm x 1 cm and a sheet of uranium free muscovite mica with a
thickness of 50-100 µm, acting as a detector, was placed tightly on top. Several of
these sample mounts and detector ‘sandwiches’ were then stacked together with
standard uranium (natural ratio) glass dosimeters (CN5 , uranium content 12ppm) at
either end of the sample stack and irradiated with thermal neutrons (to avoid
inducing fission in 238U and 232Th) using the FRM11 reactor at Garching, University of
Munich. This process induces fission in the 238U isotopes within the apatite grains. If
the mica detector was placed correctly on top with no gap left in between, half of the
tracks that would have been produced within an unpolished volume of apatite would,
as a result of polishing, travel into the mica detector causing tracks to be present
within the detector. Subsequently by etching the mica detectors in 48 % hydrofluoric
acid for 25 min at 20 °C, these can be made visible and thus make it possible to

73 | P a g e
Methods of Thermochronology Chapter 2

determine the track density within the mica detector. This is proportional to the 238U
concentration in the exposed surface of the apatite grain which induced them.

Figure 2.4-2: Work flow illustration of the external detector method used in this study. [Hurford and Carter,
1991]

The etching effectively reveals a mirror image on the mica of the track
distribution within the initial apatite grain, and the shape of the grain will be mirrored
by the induced tracks in the detector. The number of induced tracks (Ni) over a
distinct area (A) of this ‘print’ is defines as the induced track density ρi. Likewise for the
spontaneous tracks on the sample grain (Ns) over the same distinct area (A) the
spontaneous track density ρs can be calculated. Identical mirror image areas are
counted for their track density, so that any variations in U concentration can be taken
into account.

74 | P a g e
Methods of Thermochronology Chapter 2

2.4.4 Age Calculation

Comparing spontaneous and induced track densities enables the calculation of


the age of the sample since cooling below Tc:

1 𝜆𝑑 𝜑𝜎𝐼ρ𝑠 g
Equation 5: 𝑡= 𝑙𝑛 [1 + ]
𝜆𝑑 𝜆𝑓 𝜌 𝑖

Where

λd = total decay constant for all uranium isotopes [(1.55125±1.25)*10-10y-1]

φ = neutron flux [n/cm2]

σ = thermal neutron capture cross section of 235U [580.2*10-24cm2]

I = Natural isotope ratio 235U/238U [7.2527*10-3]

ρs = spontaneous track density of sample

g = geometry correction factor [0.5 for the external detector method]

λf = decay constant for spontaneous fission of 238U [(8.51 ± 0.18)×10−17 yr−1]

ρi = induced track density on external mica detector surface.

However, in practice equation 5 presents two problems which make a simple


mathematical solution difficult. The accurate evaluation of the neutron flux even
within a well-thermalized reactor is very difficult [Hurford and Green, 1983] to
determine, partly because of variations in neutron flux spatially, both radially and
along the path of neutrons that constantly undergo absorption. In order to address
this problem a piece of dosimeter glass of known homogeneous uranium content
together with another mica detector placed against it is fixed on top and at the
bottom of the sample mount – mica detector stack. By counting the average track
density on the dosimeter glass detector the average variation in induced track density,
i.e. the neutron flux ρd can be determined along the path of neutrons.

75 | P a g e
Methods of Thermochronology Chapter 2

Furthermore, at the time when the method was developed reported values for
238
λf, the decay constant for spontaneous fission of U, varied by as much as 25%
[Bigazzi, 1981]. Reported values ranged from 7*10-17y-1 to 8.5 *10-17y-1, thus increasing
the uncertainty and comparability in measured ages between different labs and
analysts, depending on the constant used. In order to circumvent this variable entirely,
the zeta calibration method was developed, which is an empirical method of flux
assessment using mineral standards of known fission track age. Although recent
technological advances have enabled more accurate measurements and have
confirmed that λf of 238U is (8.51 ± 0.18)×10−17 yr−1 [Yoshioka et al., 2005], the zeta
calibration method is still the standard method used by fission track
thermochronologists because it also includes a personal factor related to counting
efficiency.

2.4.4.1 Zeta Calibration

Not unlike in the previously described (U-Th)/He method the analytical


problems can be addressed satisfactorily by the introduction of mineral standards of
known age and of the same type as the unknowns to be dated. In the case of Fission
Track Dating each analyst has to calibrate their counting technique against a large
number of standard samples of known age to acquire their individual empirical ζ-factor
[Fleischer and Hart, 1972; Hurford and Green, 1983]. This factor is influenced not only
by mineral type (separate ζ-factors have to be acquired for each new mineral used),
but also by the microscope used, individual eye-sight and judgement as to whether a
linear etched feature in a grain is classified as a track or for example an etched
crystallographic defect, as well as the etchant used and the age range of the samples
available – to name but a few. For a standard of known age ζ can be calculated using
the following equation:

[𝑒 𝜆𝑑 𝑡 𝑠𝑡𝑑 −1]
Equation 6: 𝜁= 𝜌
𝜆𝑑 [ 𝜌𝑠 𝑔] 𝜌𝑑
𝑖 𝑠𝑡𝑑

76 | P a g e
Methods of Thermochronology Chapter 2

where tstd is the known age of the standard sample. A total of 15 standards where
counted by the analyst (Annika Szameitat), including Durango apatite (age: 31.6 ± 0.5
Ma [Jonckheere et al., 1993]), Fish Canyon Tuff apatites (age: 27.3 ± 0.7 Ma [Reiners et
al., 2002]) and the previously unmentioned Mt Dromedary Laccolith (age: 98.7 ± 0.6
Ma [Boesen and Joplin, 1972]). All standard zeta values are plotted and once a
relatively stable zeta level is reached an analyst-specific weighted mean ζ can be
calculated (Figure 2.4-3).

MEAN ZETA
500

400

300
ZETA

200

100
0 5 10 15 20

STANDARDS

Figure 2.4-3: Zeta calibration of the analyst (Annika Szameitat) of this study, resulting in a mean weighed ζ-factor
of 310.5 ± 10.83. Several standards, including the still existing outliers (e.g. no. 6 and 10) were re-counted
multiple times to ensure consistency and accuracy of the high values.

The ζ-factor of the analyst derived for this study is 310.5 ± 10.8 and was
determined based on the protocols and facilities of the Birkbeck-UCL London
Geochronology Centre . In conjunction with the dosimeter track density ρd, ζ thus
replaces the unknown λf, σ and the neutron flux φ in the age equation, which now
presents as:

1 ρ𝑠
Equation 7: 𝑡= 𝑙𝑛 [1 + 𝜆𝑑 𝜁 𝑔𝜌𝑑 ]
𝜆𝑑 𝜌𝑖

77 | P a g e
Methods of Thermochronology Chapter 2

2.4.4.2 Counting Procedure

The acquisition of ρs and ρi of a sample requires the use of an automated stage


system to precisely pinpoint the counting area A on the grain as well as on the
mirrored ‘print’. Both grain mount and mica detector were mounted on a glass slide
and the stage was calibrated using features which are distinct enough to be identified
in both mount and print. This enables the stage to automatically lead the ocular to the
print of a chosen grain and back to the mount when commanded, making it possible to
count both track densities over exactly the same area.

Grains selected for counting had to be cut in a c-axis parallel fashion to avoid
mixing grains with different etching velocities related to crystallographic orientation as
this would lead to the inclusion of grains in the dataset with different and unknown
geometry factors in relation to the mica detector. Grain oriented parallel to the c-axis
have the lowest bulk etch rate (surface dissolution) and is comparable to the mica
detector ensuing a geometry factor of 0.5. Furthermore all grains and prints had to
present a homogeneous track distribution. Inhomogeneous distributions can be
caused by zonation, inclusions or fractures in the grain, which cause bias and
uncertainty when choosing an area for counting. Such areas were avoided if found.

For each sample 20-25 apatites and their corresponding prints were counted
and a central age for each sample was calculated [Galbraith and Laslett, 1993]. This
number will increase if the sample records provenance with multiple age modes or if a
sample is very young. In the latter case some grains will have a zero track count and
therefore to capture the natural Poisson distribution it becomes necessary to count
many more grains, typically 30-60. A statistical χ2 test [Galbraith, 1981] was
performed to assess the dispersion of ages within the sample.

78 | P a g e
Methods of Thermochronology Chapter 2

2.4.4.3 Age Dispersion and Central Age

According to the initial theory of thermochronology all crystals separated from


a single rock should yield the exact same cooling age because they all - in theory - have
experienced the same physical cooling history. However, as this chapter has already
illustrated in previous sections, factors such as grain size and mineral chemistry can
lead to differences in diffusional behaviour between crystals and thus cause
considerable age variations between aliquots. In the case of Helium dating the issue is
addressed by dating multiple grains for each sample and calculating the arithmetic
mean for a best estimate of the overall cooling age (section 2.3). From the early
development of the fission track method a similar approach has been applied by
counting tracks on a minimum of 20 crystals and averaging the resulting age. Initially
this average was calculated by combining all spontaneous tracks N s of all crystals and
comparing them to the sum of all induced tracks Ni found on the mica detectors to
calculate one overall ‘pooled age’. However, when the age dispersion deviates from a
normal Poisson distribution this approach is no longer valid. To assess for over-
dispersion the chi squared test is traditionally used, however this only indicates when
there is over-dispersion, it does not say how much. Also this test does not work if track
counts are low. To overcome this [Galbraith and Laslett, 1993] constructed a random
effects model that gives a central age (grain population geometric mean age) estimate
and a relative standard deviation of the population of ages (also known as the age
dispersion and normally expressed as a percentage variation). Typically a sample with
20% dispersion is considered to be of mixed age. Over-dispersed samples can be found
in meta-sedimentary samples or modern detrital samples, where mineral populations
of different origin and thus chemical composition might be present. Because the
dispersion often contains various different non-poissonian populations, it inherently
can also contain valuable information about different age populations

To visualize the dispersion of ages within a sample the acquired data can be
plotted on a radial plot (2-variable scatter plot), which is shown in Figure 2.4-4.

79 | P a g e
Methods of Thermochronology Chapter 2

Figure 2.4-4: Principle of a radial plot to visualize fission track age distributions, redrawn after [Galbraith, 1990]
based on the visual interface of the RadialPlotter program by [Vermeesch, 2009] which was used to create radial
plots for this study. Individual ages (a and b) are plotted in the age (radial y scale) vs. total track count (x-scale)
with ± 2σ of the standard deviation projected onto the age scale. Increasing number of total tracks increases the
precision and so the resulting ± 2σ error of age b on the age scale is slightly smaller. As an additional feature the
RadialPlotter program offers colour coding of D-par values to identify potential chemical populations within the
data.

Galbraith (1988) designed radial plots for visualising the distribution of populations of
grain ages. Single grain ages (z) with standard error σ are plotted (point x,y) according
to x (precision)=1/σ and y (standard estimate)= (z-z0)/σ, where z0= the central age.
The error attached to each point is standardised on the y scale. The value of the age (z)
and the 2σ uncertainty can be read off the z scale by extrapolating lines from point 0,0
through the plotted age (point x,y).

Because of the option to distinguish populations within 2 variables, such plots


can help analyse the dispersion and enable the user to spot age populations or
chemical populations much better than a simple 1-variable histogram could. Examples
on how this can be applied for real samples are shown in Figure 2.4-5.

80 | P a g e
Methods of Thermochronology Chapter 2

Figure 2.4-5: Radial plot of real samples 12-AS-14 (left) and 12-RP-AS-04 (right). At first glance the data of sample
12-AS-14 appears to show a much larger spread, however the moderate precision (here shown as the σ/t (t=age)
2
ratio rather than total no. of counts) and the resulting large 2σ error causes the sample to pass the χ test and be
0% dispersed. Statistical peak analysis confirms that under these circumstances 100% of the data result in a
central age of 26.1 ± 1.6 Ma. Although sample 12-RP-AS-04 appears to show a narrower cluster of data, it fails
2
the χ test with 57% dispersion due to a small 2σ error range and lower precision. Peak analysis for this sample
shows that the data, which combined leads to a central age of 33.8 ± 4 Ma, can be statistically divided into two
peaks: 86.8% of the data centre around and age of 27.2 ± 2.2 Ma while 13.2% of the data result in a smaller yet
older peak at 94 ± 13 Ma. As both samples have low D-par values indicating simple F-apatite chemistry, chemical
variations are unlikely to be the cause for this division. The main peak caused by over 80% of the crystals
indicates, however, that the true age of the sample might be slightly younger than the central age.

2.4.5 Track Length Measurements

As described in section 2.3.1, when subjected to high temperatures fission


tracks anneal due to thermal energy present in the system. This process starts at the
tips of a track, moving towards the middle over time and thus shortening it. The length
of a fission track is primarily a function of the maximum temperature to which it has
been exposed (the duration of heating has a secondary influence), and because tracks
are forming continuously, individual tracks will experience and therefore relate to
different portions of a sample's thermal history. The track length distribution of a
sample is therefore the result of a sample’s thermal history.

81 | P a g e
Methods of Thermochronology Chapter 2

Since a decrease in fission track length causes a reduction in the probability of


a track intersecting a mineral surface, a short mean track length will cause a reduction
in track density, resulting in a reduced or apparent age that has little direct geological
meaning. Consequently, to interpret fission track data properly it is essential to know
if the measured age reflects a true normal full-length distribution or, is an apparent
age as a result of track shortening due to exposure to elevated temperatures.

Throughout the 1980s a succession of papers were published that used


confined track length data to monitor fission-track annealing in laboratory
experiments [Laslett et al., 1987; Duddy et al., 1988; Green et al., 1989] and in the
natural geological environment [Gleadow and Duddy, 1981]. These resulted in a
quantitative predictive model of fission-track annealing, based on the Durango apatite
(composition Cl/F ratio of ~ 0.1), which can be used to predict partial annealing
temperatures (~60-110°C) to ~±10°C for time-scales between 106-108 years. More
recent work has produced annealing models that take into account apatite
composition [Ketcham et al., 2007a] as it has been shown that the actual
temperatures that define the partial annealing relate to apatite chemistry. Partial
annealing refers to temperatures where there is track length reduction but not loss.

Track lengths are measured only on complete horizontal confined tracks,


avoiding the need for correcting angle of dip. The tracks are confined because the
etchant has found its way into the crystal via a cleavage crack to reveal a Track IN
CLEavage (‘tincle’), or a Track IN Track (‘tint’). They are easily identified as both ends of
a track will be in focus at the same time. Track lengths are measured in crystals of the
same orientation as for counting, i.e. parallel to the 'c' crystallographic axis. This again
avoids etching anisotropy effects. Since cleavage develops in a preferred orientation it
follows that ‘tincles’ will also have an angular bias as most confined track tend to be
clearly visible at angles near to perpendicular to the cleavage. Therefore, only tints are
measured to avoid angular bias

Where a sufficient number of these ‘tints’ was present, between 50 and 100
track lengths were measured for each sample. The angle of all measured lengths with

82 | P a g e
Methods of Thermochronology Chapter 2

the c-axis was also recorded, because observation of track lengths is biased with
crystallographic orientation and tracks oriented perpendicular to the c-axis are wider
and more visible compared to long thin tracks parallel to the c-axis. Annealing is also
anisotropic such that tracks perpendicular to the c-axis shorten faster. This angular
information is used to make an orientation correction known as the c-axis projection
method, which is incorporated in the track annealing model of Ketcham et al. (2007)
(see section 2.5).

2.4.6 Mineral Chemistry

The chemical composition of naturally occurring apatite (Ca5(PO4)3(F,OH,Cl))


can vary significantly between fluoro-, chloro- and hydroxyl-apatite endmembers. As
mentioned above, the ratio of chlorine to fluorine ions within the crystal plays a major
role in influencing the temperature at which the system closes and tracks are
preserved. The larger ion radius of chlorine slows down annealing and thus Tc for
chlorine-rich apatites is higher than for fluorine-rich endmembers. The larger ion
radius also significantly increases the bulk etch rate of the apatite crystal. While the
abundance of OH-groups is not known to significantly influence the closure
Temperature, it may also increase the bulk etch rate in a similar fashion to chlorine.

These empirical etching characteristics can be used to estimate the chemical


composition of a sample and identify variation within the individual grains of the
sample by obtaining the average D-par value for each sample. D-par, defined as the c-
axis parallel length of an etch pits (the opening where tracks intersect with the
exposed surface of the grain), for Fluor-apatite typically ranges from 1-2µm, while
both Chlorine- and Hydroxyl-apatites show etch pits of up to 5µm length due to higher
etching velocities. For each analysed sample D-par values were obtained.

Although D-par provides a good approximation to grain composition/annealing


sensitivity, 6 samples representative of the range of D-pars encountered during
analysis were also analysed for chlorine and fluorine content using an electron analysis

83 | P a g e
Methods of Thermochronology Chapter 2

at the Laboratory of the Birkbeck College, University of London, to verify the D-par
measurements. Durango apatite was used as an internal standard for these
measurements.

2.4.7 Caveats

There are several factors that can complicate fission track analysis or even
make the acquisition of a meaningful age impossible. Zonation is the most common
problem, since most apatites grown in a natural setting will have experienced some
degree of chemical variation during growth and thus present with varying track
densities across a crystal.

Furthermore in very old samples or samples with high U-Th content it is


possible that the number of tracks – spontaneous for old samples, both spontaneous
and induced for high U-Th samples – increases beyond a number which is still
countable. If too many tracks are present they become very hard to distinguish and
the remaining crystal structure around them can even be etched away when
submerged in acid. In this case a fission track mount will become virtually impossible
to count, as all tracks interconnect and the uncertainty introduced by attempting to
count them would likely be large.

In contrast, young samples or samples extremely low in U and Th can contain


grains which show no tracks at all. While zero counts still have to be ‘counted’ towards
the central age to obtain a statistically valid Poisson distribution, such samples render
track length measurements impossible as there are simply not enough tracks present
to allow for them to intersect and highlight surface parallel tracks at depth. In addition
uncertainties of the counting statistics on samples with few total counts can approach
100%. Hence although very young samples with sufficient U and Th, low track densities
yield robust maximum ages, their uncertainties can be large.

84 | P a g e
Methods of Thermochronology Chapter 2

2.5 Thermal modelling of Thermochronology Data

Interpreting thermochronological datasets obtained by the methods described


above is a complex issue. While using multiple chronometers to maximize the
information gained from a given sample, quantifying the range of thermal histories
using the data record is a challenge. In order to identify thermal histories, which are
consistent with the obtained datasets (including any geological constraints) without
bias, computational tools such as statistical tests and numerical models are required.
Currently two of the most commonly used programs to interpret cooling histories are
HeFTy [Ketcham, 2005] and QTQt [Gallagher, 2012]. For the purpose of this study
HeFTy was used to analyse all thermochronological data obtained as well as to forward
model the thermal histories of the vertical profiles sampled. Furthermore QTQt was
used to model 3 vertical profiles to compare the results to the ones given by HeFTy.
Both programs will thus be described in more detail in this section.

2.5.1 HeFty

Hefty can perform both forward and inverse modelling of thermochronological


datasets. It has become a well-established tool for the thermochronology community
over the last decade and has been used in many studies to interpret
thermochronological datasets. A detailed description of all options provided by this
program can be found in Ketcham (2005). This section will mainly focus on the
methods and tools used in this particular study.

2.5.1.1 Forward modelling

Forward modelling is not only a good educational tool which helps to visualize
the principles of thermochronological dating, but it also assists with first order

85 | P a g e
Methods of Thermochronology Chapter 2

exploration of varying thermal histories and data patterns that could be expected from
them. A theory can be very quickly tested by simply drawing a time-temperature path
in the ‘time-temperature-history’ field and the model will predict a fission track mean
age, the age of the oldest track and the overall track length distribution as well as a (U-
Th)/He mean age and diffusional profile which could be expected from a real sample
for this particular history.

In order to be comparable to a real dataset these ‘predictions’ or ‘forward


models’ are corrected for various biases introduced by the technique as outlined
previously. As described in section 2.3.7, the most important factor in fission track
dating is the strong anisotropy in annealing and etching behaviour depending on a
track’s orientation in relation to the crystallographic c-axis [Green and Durrani, 1977;
Donelick, 1991]. Tracks perpendicular to the c-axis anneal faster than those parallel to
the c-axis. This implies that the likelihood of intercepting a track at a high angle to c
will decrease over time, skewing the observed track length distribution towards longer
tracks which present at orientations closer to the c-axis. Etching velocities on the other
hand increase as the angle to the c-axis increases, meaning that high angle tracks tend
to be wider and thus easier to spot. It is therefore important that the analyst remains
unbiased and consistent when measuring track lengths so to not introduce further
bias. For both apatite and zircon the program offers several annealing models which
can be chosen via a drop down menu and will then be used to calculate expected track
length distributions. In order to achieve a realistic length distribution all lengths
measured in real samples are corrected for their orientation using the correction
method introduced by [Ketcham et al., 2007a] and presented as c-axis projected
lengths. A forward model produced by HeFTy will also take this bias into account and
produce a distribution of c-axis projected lengths which could be expected.
Additionally, chemical indicators such as Chlorine content or D-par measurements can
be varied to explore the effects on the resulting track length distribution and fission
track ages.

Forward modelling of the (U-Th)/He data is done as a finite difference solution


for an idealized spherical grain. This simplification is necessary to keep computing time

86 | P a g e
Methods of Thermochronology Chapter 2

at a minimum and thus enable the user to observe changes swiftly as the T-t path is
changed. For this system the model predicts a mean helium age based on the basic
age equation (equation 1, section 2.2.1) as well as a (grain size normalised) diffusional
profile for helium. As discussed in section 2.2 estimates of diffusion kinetics vary and
HeFty thus enables the user to choose the mineral used and their preferred diffusion
model via a drop down menu as well as offering the possibility to input their own
values. Furthermore concentric zonation can be included by manipulating zonation
profiles for both U and Th. An example of the HeFty user interface and forward
modelling results is shown in Figure 2.5-1.

Figure 2.5-1: HeFTy user interface for forward modelling. 1: Time-Temperature history window, where a T-t path
can be chosen and altered by dragging the nodes. 2: Thermochronometer tabs where preferred
annealing/diffusion models can be chosen. 3: Predicted results for selected models.

2.5.1.2 Inverse modelling

Inverse modelling using HeFty allows the user to simultaneously input actual
data results of up to seven thermochronometers for each sample and will provide a
graphic output of all possible solutions to time-temperature paths which are

87 | P a g e
Methods of Thermochronology Chapter 2

consistent with the data at hand. Data can be input as apatite or zircon fission track
data (input: Ns, Ni, zeta, rho d and rho s), track lengths (input: lengths measured and
angle to c-axis if available), (U-Th)/He age data (for apatite, zircon or titanite, input:
grain size (equivalent spherical radius), measured age and U/Th/Sm ppm), He-diffusion
profiles (degassing data), zonation profiles (e.g. SEM measurements) and %Ro vitrinite
reflectance data by adding tabs to the user interface provided (Figure 2.5-1).

In addition to allowing for data input, the tabs also offer various drop-down
menus where preferred diffusion- or annealing-kinetics models can be chosen as well
as c-axis projection parameters for FT data and e.g. various stopping distances for
Helium. Over the past two decades the rapid increase in low temperature
thermochronology data and the increasing controversies over interpretation of said
data have emphasized our present lack of understanding of the physical principles of
diffusion and annealing behaviour in crystals. This in turn led to an increase in research
into influencing factors on annealing and diffusion which has seen the development of
several diffusion models which are now widely used and showcased in Table 2.5-1 for
comparison. While early annealing models used kinetics based on step-heating
experiments of chemically untypical yet widely used mineral standards to predict how
an age would change with varying temperature history [Laslett, et al., 1987; Ketcham,
1999], the most recently developed models are calibrated against a wide variety of
chemical compositions and thus more true to natural samples of both apatite
[Ketcham et al., 2007] and zircon [Yamada et al., 2007].

This understanding of physical principles that influence a predicted age


becomes even more crucial in the case of diffusion models (U-Th)/He dating and in
particular their interaction with radiation damage caused by fission track formation (as
described in section 2.3.8.1). Recent studies have shown that differences in effective
uranium (eU) content can cause large dispersion within samples (as much as 200%)
based on the interaction of increased helium retention through damage to the crystal
lattice (‘traps’) introduced by fission tracks [Gautheron et al., 2009; Flowers et al.
2009]. In the case of zircon it has even been suggested that, due to the generally
higher eU content, this increased retention might reverse if the damage becomes too

88 | P a g e
Methods of Thermochronology Chapter 2

great and creates ‘tunnels’ for Helium to escape the crystal [Guenthner et al., 2013].
While these principles are still heavily researched, three diffusion models which take
these effects into account as far as the current state of knowledge permits are
generally used: For apatite there are two RDAAM (Radiation damage and annealing
model) versions available, namely the models after Gautheron et al. (2009) and
Flowers et al. (2009). While the first model uses a more physically rigorous approach
with heavy emphasis on grain properties it generally produces less good fits to natural
data. In comparison the model after Flowers et al. (2009) emphasizes the importance
of eU and seems to produce better fits with natural data and dispersion, however it is
not yet fully understood by the thermochronology community why that is and thus
warrants further investigation. For zircon (U-Th)/He data the latest model by
Guenthner et al. (2013) combines and RDAAM model with the reversal of retention
after damage becomes too great, and is thus the most advanced option yet.

For the purpose of the models created in this study, the AFT data was modelled
in HeFty using the latest annealing model by Ketcham et al. (2007) and 5.0 molar
etching acid parameters from the same study. The goodness of fit was assessed using
the Kuiper’s Statistic option with a 95% confidence level. The AHe models were
calibrated using diffusion kinetics after Farley et al (2000) while for ZHe the
parameters of Guenthner et al (2013) were used. Both helium methods used stopping
distances and alpha ejection correction after Ketcham et al. (2011) and the corrected
age was reported as the result [Ketcham et al., 2007; Farley, 2000; Guenthner et al.,
2013; Ketcham et al., 2011].

In order to run an inverse model an initial temperature (starting point) and a


current surface position and temperature have to be marked on the time-
temperature-history field. A further option provided is the possibility to draw time
temperature boxes through which the thermal history has to pass (Figure 2.5-1). Based
on the published biotite Ar-Ar data (as discussed in chapter 1) which indicate that the
Columbia Mountains experienced peak metamorphism during the upper cretaceous,
all models created for this study were produced with an initial condition ‘box’ at 300-

89 | P a g e
Methods of Thermochronology Chapter 2

Table 2.5-1: Overview and comparison of currently relevant annealing/diffusion


models available in HeFTy and QTqt

Annealing/Diffusion
Features Caveats and Advantages
Model
Improved and currently the most advanced
Multicompositional annealing
apatite annealing model, incroporates
Ketcham et al., algorithm which uses input grain
varying chemistry and was calibrated on
2007 chemistry and dimensions to calculate
numerous natural samples from published
annealing for apatite fission track ages
data and annealing experiments
Currently the most advanced zircon
FT Annealing Models

Yamada et al.,
Annealing experiment based annealing annealing model, incroporates varying
2007
algorithm for zircon fission track data chemistry and was calibrated on numerous
(only QTqt)
natural samples from published data
Multicompositional annealing
algorithm (earlier version) which uses
Ketcham et al., Older version of Ketcham, 2007 (above).
input grain chemistry and dimensions
1999 Contains less data for calibration.
to calculate annealing for apatite
fission frack ages
Initial algorithm for Durango chemistry,
Durango standard annealing however although widely used as age
Laslett et al.,
experiment based annealing algorithm standard, most sample chemistry differs
1987
for apatite fission track data greatly from Durango apatite and thus this
model might not be applicable to most data.
Compositional diffusion model for
zircon, includes 'trapping' of helium in
caveats created by fission tracks AND
Guenthner et Currently the most advanced model for
eventual speeding up of diffusion once
al., 2013 zircon, the only RDAAM for this mineral.
radiation damage threshold is crossed
(needs FT data as proxy to recalibrate
diffusion)
Only correlates diffusion to FT data, no
explicit use of eU input. However the
Gautheron approach (when compared to
Compositional diffusion RDAAM model Flowers) suggests more emphasis on grain
(U-Th)/He Diffusion Models

for apatite, includes 'trapping' of geomoetry. No reversal of retention with


Gautheron et helium in caveats created by fission increasing radiation damage (as for zircon in
al., 2009 tracks (needs FT data as proxy to Guenthner model). Although based on
(only QTqt) recalibrate diffusion), extremely rigorous principles this model often
rigourous linear model following produces bad fits with natural data when
known physical constraints compared to the Flowers approach (below).
Both this and the Flowers model are
currently widely applied to apatite data, the
choice comes down to author preference.
Correlates diffusion to FT data and strongly
The original RDAAM (Radiation to eU. No reversal of retention with
damage and annealing model), non- increasing radiation damage (as for zircon in
linear compositional diffusion model Guenthner model). For high U/Th
for apatite, includes 'trapping' of uncertainties this could be problematic,
Flowers et al., helium in caveats created by fission however both this and the Gautheron
2009 tracks (needs eU and FT data as proxy model are currently widely applied to
(only QTqt) to recalibrate diffusion) and the effect apatite data, the choice comes down to
of FT annealing in interaction with author preference. This model tends to fit
radiation damage. Calibrated on natural data better, however the reasons
natural apatites of varying composition why are not completely understood and
and thus widely applicable. require further research before this can be
confidently chosen as the preferred model.

90 | P a g e
Methods of Thermochronology Chapter 2

310°C between 45 Ma and 80 Ma. The current surface temperature was set to 10 ± 10
°C. No other constraints were forced on the model to allow the full range of possible
and geologically valid histories to be explored.

Using these start and finish conditions, the HeFTy inverse model run will then
randomly choose a user-defined number of paths between start and finish and
through every drawn box, each time evaluating the statistical fit to the data. For
Fission Track length distributions a Kuiper’s Statistical test is performed which assesses
the statistical chance of a resulting thermal history being more than 5% different from
the measured data [Press et al., 1990; Willett, 1997]. Thus a T-t path that passes the
Kuiper test at the 95% confidence level is permitted. The goodness of fit (GOF) of age
data is tested according to [Ketcham et al., 2000] and also deemed a good fit when
less than 5% of a randomly chosen subset of ages differs from the predicted age. Both
tests are performed by the program in a combined merit function and the results are
displayed on the final time temperature plot. Where both statistical tests were passed
at the 95% confidence level a time-temperature-path is a ‘good fit’ to the data and
plotted in magenta colour. Where the statistical precision limit of 50% confidence level
is met, paths are deemed an ‘acceptable fit’ and plotted in green (Figure 2.5-2).

2.5.1.3 Caveats

The most evident caveat of any model has to be that any given result can only
be as good as the data used to produce it. Time constraints, sample quality, availability
or financial limits generally cause datasets to be limited and thus only a sample of the
actual information contained within the rock are available to be used. Thus a model
result will only be able to show a range of possible solutions for the data provided
assuming a limited data set accurately represents the sample.

Furthermore a dataset produced in one laboratory may differ slightly from


another one produced somewhere else, based on analyst, microscope, etching

91 | P a g e
Methods of Thermochronology Chapter 2

procedure etc. as discussed in previous sections. These uncertainties are therefore


included in the results any thermal model will present.

Figure 2.5-2: Example of an inverse modelling run using AFT and ZHe data. A: Boxes in the time-temperature-field
determine constraints for the modelled paths while the measured data is displayed on the right. B: Result of a
inverse model of 100.000 randomly chosen paths which were tested against the data input. Where the randomly
chosen paths produced permissible thermal histories they are displayed as lines, whith statistically ‘acceptable’
(50% of ages agree with this history) paths displayed in green and ‘good’ paths (95% areement) displayed in
magenta. The wide spread at ages older than 55Ma is caused by a lack of data, while the temperature range
where sufficient data is present the history is better constrained.

Where the statistical analysis is concerned, it has been suggested that the
‘Frequentist’ statistical tests used by HeFTy during forward modelling (such as the K-S-

92 | P a g e
Methods of Thermochronology Chapter 2

test and the GOF test) and their assigned 95% confidence limit are not the most
rigorous solutions. For very large datasets other approaches, such as the Bayesian
Multi Domain Monte Carlo algorithms used in Kerry Gallagher’s program QTQt should
be utilized to avoid over-interpretation of results [Vermeesch and Tian, 2014].
However, for smaller studies such as this one, the mathematical solutions and inbuilt
quality-control of data in HeFTy (e.g. not permitting physically impossible datasets to
be modelled) are the better option if the overall ‘garbage in – garbage out’ principle is
considered during the interpretation of results. As mentioned before, the results of a
model are only as good as the data used to produce them.

However, a key advantage of this program, when compared to QTQt, is it’s


user-friendliness for beginners and also the ability of a user to quickly test a presumed
thermal history in the forward modelling mode. It is the best teaching and exploration
tool, as it allows real time manipulation of all input parameters and gives instant
thermal history results. This uniquely enables the user to see how a dataset would
react to the changes in parameters and which settings might be the most suitable for
the data at hand. For these reasons the main thermal modelling of this work was done
using HeFTy.

2.5.2 QTQt

QTQt is another thermal history modelling program created by Kerry Gallagher


[Gallagher, 2012] which is based on the same physical principles as its rival HeFTy. In
many ways it offers the same options as HeFTy for modelling fission track data and
potential additional information such as helium data, vitrinite reflectance or helium
diffusion profiles. The principal distinction between the two programs is the approach
to modelling and subsequent statistical analysis of the data. While HeFTy uses a
frequentist 1 dimesional finite difference model to calculate cooling ages [Ketcham,
2005], QTQt contains a Bayesian algorithm which follows a Markov Chain Monte Carlo
(MCMC) approach (‘random walk’ model rather than ‘finite difference’) [Gallagher,

93 | P a g e
Methods of Thermochronology Chapter 2

2012]. While in HeFTy all T-t solutions which cannot be rejected within the given
confidence intervals (95 and 50%) are presented as a solution, QTQt instead explores
and evaluates the model space, ranking the found solutions by likelihood and plotting
the most likely ones. As an extensive review and comparison between both programs
done by Vermeesch et al. (2014) showed, this simple distinction can have severe
consequences for data interpretation, and thus care has to be taken by the user when
interpreting data from either program. While HeFTy’s approach may be more easily
understood and the interface easier to use, the randomly assigned confidence
intervals and less rigorous statistics can lead to unreasonably few or no solutions for a
model which is based on either a very small or, even more problematic, a very large
high precision dataset [Vermeesch et al., 2014]. In such cases QTQt’s approach will still
provide the same objective assessment of the data and provide solutions that are
consistent with the data which was input. For either model, regardless of sample size,
the ‘garbage in, garbage out’ principle applies.

In the past, a more rudimentary user interface and, a complicated system of


data input where the user had to build their own files following only sparse
instructions, meant that QTQt was often neglected by early career researchers who
predominantly chose HeFTy for its beginner’s usability. However as of version 5.4.6
(January 2016, and thus in time for the final submission of this thesis and therefore
described here) it also offers a window style user-interface where the user can choose
and manipulate parameters at will, as well as a newly added feature of forward-
modelling (much in the same style as described for HeFTy above), which was
previously not possible. In the latest QTQt version which was used for this study a data
file can be built using a data input window much like in HeFTy , starting with fission
track inputs and in the following windows options to add further thermal history
information. An inverse data run is then conducted using the MCMC run function after
defining thermal parameters and run parameters. It is also, unlike in HeFTy, possible to
load numerous samples at once, e.g. for a vertical profile, and run them together.

In addition to the same plotting options as HeFTy, QTQt offers some additional
options to help with data analysis, such as letting the user chose between different

94 | P a g e
Methods of Thermochronology Chapter 2

statistical tests, prediction models (different statistical bases will display different
ranges of possible solutions) as well as numerous plot options to make data analysis
more convenient. Upon completion the quality of the forward model run can be
immediately assessed by examining the available likelihood chain graph, which shows
how well the MCMC chain performed in response to the data. If this is acceptable
(steady, no trends visible) the user can then choose to plot whole profile or individual
sample T-t histories, summary predicted vs. observed age/MTL plots as well as track
length histograms and statistical fitting plots for several different models. These
include the maximum likelihood model (best fit to the actual data), maximum
posterior model (max. likelihood multiplied by the prior probability distribution of the
Bayesian approach), maximum mode modelled (combines the probability distribution
of all models sampled) and the expected model (weighed mean model, weighed for
the probability obtained by the posterior model). (For further description of all QTQt
options a comprehensive userguide is available for download alongside the program
after contacting K.Gallagher (kerry.gallagher@univ-rennes1.fr)).

For the purpose of this study temperature history plots were created for 3
vertical profiles in order to compare them to the HeFTy results. These are shown along
with predicted versus observed data plots in Chapter 5 (section 5.3.1) and compared
to the results produced by HeFTy for the same localities to examine the difference in
results between the two programs. The inverse modelling needed to create these
plots was done using the same Ketcham (2007) annealing model for apatite fission
track as used in HeFTy, as well as the Flowers (2009) and Guenthner (2013) diffusion
models for apatite and zircon helium data respectively (see section 2.5.1.2).

2.6 Chapter Summary and Conclusions

This chapter has illustrated how low-temperature thermochronology,


particularly with the combination of several methods and minerals, is uniquely capable

95 | P a g e
Methods of Thermochronology Chapter 2

of shedding light on the recent cooling and thus exhumation history of the uppermost
parts of the crust. These methods enable us to investigate spatial and temporal
patterns of cooling across mountain ranges both in horizontal and vertical direction,
provided great care is taken in sample selection and data interpretation as described
in this chapter.

For the purpose of this study the above described methodologies apatite (U-
Th)/He dating, apatite fission track dating and zircon (U-Th)/He dating with closure
temperatures of ~70°C, 120°C and 180°C respectively were applied to investigate the
cooling history of the South-Eastern Canadian Cordillera. AHe and AFT and selectively
ZHe cooling ages of 72 rock samples spanning a vertical elevation range of almost
2500m were determined. Where possible fission track length distributions were
acquired as described here and the most likely thermal histories of these samples
were modelled using HeFTy (and for 3 vertical profiles also QTQt), taking into account
the newest diffusion kinetics and various factor such as grain size, habit and chemistry.
A total of 8 vertical profiles were collected to assess the age-elevation relationship at
these locations and also to investigate how this relationship changes across the range.
All thermochronology result produced will be presented in the following two chapters.

96 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

Chapter 3:
Thermochronology of the Columbia
Mountains of the Southern Canadian
Cordillera

3.1 Introduction

As described in detail in chapter 1, the south eastern Canadian Cordillera


includes the Rocky Mountains east of and the Columbia Mountains west of the Rocky
Mountain Trench. The Columbia Mountains in a wider sense comprise of the Cariboo
Mountains, the Quesnel Highland, the Monashee Mountains, the Shuswap Highland,
as well as the Selkirk and Purcell Mountains (Figure 3.1-1 and sections 1.2.1 and 1.2.2
in chapter 1 for more detail).

Briefly summarized they display a varied geology of Proterozoic to Eocene


igneous, metamorphic, and (meta-) sedimentary rocks of both the North American
craton and its continental margin, and arc and oceanic basin terranes of Palaeozoic
and Mesozoic age, the latter accreted mainly in Jurassic time. Orogenic events mainly
of Mesozoic and early Cenozoic age caused widespread deformation and
metamorphism of this composite assemblage of rocks. Predominant lithologies of the
area are clastic and volcanic assemblages of the Proterozoic Kootenay Terrane and the
mainly clastic continental margin sediments of the Windermere Formation (chapter 1,
Fig 1.2.2). Furthermore, several extension-related core complexes, metamorphosed
during Late Cretaceous – Eocene time, were exhumed in the Eocene during a shift
from E-W contraction to NW-SE extension. Examples are the Valhalla Complex
(paragneisses of the southern Selkirk Mountains), the Monashee Complex (Monashee
Mountains), and the smaller Malton Gneiss west of the northern end of Kinbasket
Lake. Both of the latter expose Proterozoic crystalline rocks of the craton beneath the

97 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

Neoproterozoic Windermere Supergroup. A more comprehensive and detailed


description of the geology and geological history is provided in chapter 1.

Figure 3.1-1 Physiographic map of the south eastern Canadian Cordillera (after (Mathews 1986), showing the
main ranges. The study area of this chapter includes all ranges west of the Rocky Mountain Trench (yellow
dashed line). (Background image: Google Earth, 2015)

Existing high temperature thermochronology data established cooling below


700°C of the deeper levels of the Monashee and Valhalla Complexes between ~70 and
55 Ma while higher level tectonic units such as the Selkirk Allochthon and the Cariboo
Mountains experienced peak metamorphic events between ~170-125 Ma respectively
(Parrish 1995). Ar-Ar ages utilizing hornblende and mica within the Monashee and
Valhalla Complexes record cooling from 500 to 300°C during the Eocene, with most
ages between 50 and 60 Ma. Once again some higher units, such as the Cariboo and

98 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

Selkirk Mountains recorded older, mainly Cretaceous and for the Selkirk Allochthon
Jurassic ages (see comprehensive review of existing data in chapter 1 (Table 1.3.1) for
references and additional detail).

Few low temperature thermochronology studies are available for this region.
Zircon and apatite fission track data from the Shuswap complex indicates cooling from
200 to below 120°C during the Eocene and early Oligocene, with ages ranging from 54-
38 Ma (ZFT) and 49-28 Ma (AFT) (Lorencak et al. 2001). A similarly localized study
analysing the exhumation along the Lewis thrust near the American border found
apatite fission track ages of 45-85 Ma (Osadetz et al. 2004).

This chapter presents a new comprehensive and widely spread low


temperature thermochronology dataset of Apatite Fission Track- (AFT), Apatite (U/Th)-
He – (AHe) and Zircon (U/Th)-He (ZHe) ages from the western part of the Columbia
Mountains in order to improve our understanding of their Paleogene to Neogene
cooling and exhumation history.

3.2 Results

In order to explore the overall thermal history of the south-western Canadian


Cordillera four widely spaced transects in addition to seven vertical profiles A-G were
sampled (Figure 3.2-1). The resulting AFT, AHe and ZHe ages are described in detail,
transect by transect from north to south in the sections below.

3.2.1 Trench Transect

The ‘Trench’ transect follows the Rocky Mountain Trench along the Fraser River
from Prince George in the north to Valemount at the northern shore of Kinsbasket

99 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

Lake in the south. Here the profile bends slightly southward, not following the RMT
but a course parallel to it until Mt Albreda, the location of vertical profile E, is reached.
Striking roughly NW – SE and thus being oriented along the axis of the orogen this
transect offers insight into the along strike cooling history at low elevations from the
flat landscape of the interior plateau in the north, to the valley bottoms below some of
the highest peaks (>3000m) in the core of the range on either side of the profile line.

Figure 3.2-1: Overview map of samples collected west of the Rocky Mountain Trench. Horizontal transects to be
discussed in this chapter are indicated by grey boxes, vertical profile locations A-G are indicated in pink. Blue
markers indicate bedrock samples, green markers indicate modern sediment samples. Individual samples of
vertical profiles are not displayed, see sections below for detailed locations. (Background image: Google Earth,
2015).
The five sampling locations along this transect range in elevation from 564 m to
875 m a.s.l. and cover a distance of approximately 340 km. Unlike the other, more
southern transects most samples of the trench transect are modern detrital river
sediments. They were sampled along river banks of modern rivers to gain insight into

100 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

the AFT age range present in their catchment area, i.e. the area where the sediment
could have originated. The spatial extent of these catchments is shown in Figure 3.2-2.

Figure 3.2-2: Trench transect sample location map. The thick white line indicates the path of the elevation
profile. White squares indicate sampling locations of modern detrital river sediments (samples GF31-39) while
black squares show the locations of bedrock samples. For samples GF35-39 the yellow shaded areas show the
extent of corresponding catchment areas where sediment could have originated. For sample GF31 the potential
catchment area is indicated by the thick yellow line. The location of vertical profile E is shown in pink.
(Background image: Google Earth, 2015)

The three detrital samples in the middle of the transect, GF 35, GF 37 and GF
39 all presented Oligocene AFT central ages, with 30.1 ± 4.3 Ma, 28 ± 2.4 Ma and 33.7
± 1.3 Ma respectively (Figure 3.2-3). However these central ages contain multiple age
populations within their age distribution (for radial plots of all modern detrital samples
presented here see Appendix 3.2).

The catchment of sample GF 35 covers an area of 530km2 and reaches a


maximum elevation of 2300m, draining the north-eastern part of the Cariboo
Mountains. While two thirds of the crystals analysed for this sample yielded Oligocene
ages and formed a population centred on 28.6 ± 3.2 Ma, five crystals presented with a
much younger Miocene age, centred on 16.5 ± 3.5 Ma. Similarly, draining the south-

101 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

eastern Cariboo Mountains, sample GF 39 with a catchment area of approximately


150km2 and a maximum elevation of 3420m displays Oligocene (35 ± 1 Ma) and
Miocene (17.4 ± 3.1 Ma) subsets. In addition to those and most likely due to a higher
maximum elevation, it also contains and older population with 46.5 ± 4 Ma, and thus
early Eocene age. Despite a central Oligocene age, sample GF 37 (catchment area
690km2 and maximum elevation 2880m, draining mainly the Eastern Rocky
Mountains) only seems to contain Miocene (22.3 ± 1.1 Ma) and Eocene (44.9 ± 4.8 Ma)
populations, emphasizing the importance of analysing age populations when dealing
with detrital samples.

Figure 3.2-3: Elevation profile of trench transect. For the modern detrital samples (white markers) GF31-39 the
central AFT age and populations (1)-(3) are shown, with [number] indicating the number of crystals which
contributed to each population. For the bedrock sample 12-AS-22 (black marker) results of apatite (U-Th)/He
analysis are presented in blue. Errors given are 1σ for AFT and 2σ for AHe ages. Vertical profile E connects to this
transect in the south-east.

vSample GF 31, being located at the edge of the range and sampled from a
river which is created by the confluence of most streams draining the northern part of
these mountains, has the largest catchment area of at least 40.000km2 reaching a
maximum elevation of 3420m and containing all smaller catchments previously
mentioned. It is thus not surprising that just like the smaller catchments included in
the sampling area, the largest age population found in this sample yielded an
Oligocene age of 28.9 ± 1.7 Ma. However in addition to that this sample also
presented two so far unidentified age groups: upper Cretaceous (71.7 ± 11.9 Ma) and

102 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

even earliest Jurassic (201.4 ± 28.9 Ma). The Miocene age appears to have been
statistically unrepresented given the relatively small number of crystals dated.

The only bedrock sample along this transect, 11-AS-22, also yields an Oligocene
age, although unfortunately its uncertainty is very large. This age, however, was
obtained by Apatite (U-Th)/He dating which has a much lower closure temperature.
This could indicate that at least the southern part of this transect experienced cooling
through both the AFT and AHE closure isotherm (180° and 70°C) during the Oligocene.

3.2.1.1 Vertical Profile E

At the south eastern end of the trench transect vertical profile E was sampled
along the northern flank of Mt Albreda over a horizontal distance of approximately
18km and covering an elevation range of more than 1400m (Figure 3.2-4). Sample
PCA-200-83 was later added from archived samples, which is why this profile is so
widely spaced, because of its high elevation and the thus added information it was
nevertheless included. This profile is located on the smallest basement complex, the
early Proterozoic Malton Gneiss. Like the southernmost bedrock sample of the trench
transect (11-AS-22) which was sampled only 7km to the north-west of this profile, the
apatite (U-Th)/He ages obtained here are Oligocene in age (within error), ranging from
20.3 ± 8.8 Ma to 37.4 ± 0.1 Ma. Both zircon (U-Th)/He ages obtained for the lowest
and mid-elevation samples also record this cooling phase with ages of 27.9 ± 4.5 Ma
and 34.2 ± 3.2 Ma.

Surprisingly, the apatite fission track ages, which could only be obtained for the
highest and lowest elevation samples, consistently show ages approximately 10 m.y.
younger than the corresponding apatite helium age (Figure 3.2-4, top right). At first
glance and observed on their own, the AFT data indicate cooling through the AFT
isotherm during the upper Miocene. However, with the AFT closure isotherm (~120°C)
being located in between the AHe (~70°C) and ZHe (~180°C) isotherms, this cooling
history appears to be physically impossible without having been recorded by the AHe

103 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

ages as well. Possible reasons for this discrepancy include a potential short term
reheating event which was unable to affect the AHe ages (as suggested by T. Ehlers
and D. Whipp during personal conversation) or excessively old apatite and zircon (U-
Th)/He ages caused by radiation damage due to high effective uranium content.
However, there is no evidence for the first hypothesis nor were strong positive eU
correlations established for the He-data. This will be discussed in greater detail in
section 3.2.5. The age elevation relationship obtained from AFT data suggests a
cooling of 255m of rock through the relevant closure isotherm per one million years, a
rate which seems to be similar to what the AHe and ZHe data suggest, albeit shifted
towards younger ages.

y = 255.46x - 1929.6
2500 R² = 1

2000
Elevation (m)

1500

1000 AHe
AFt
500
ZHe

0
0 10 20 30 40 50
Age (Ma)

Figure 3.2-4: (Top left) Sample locations and elevation profile line of vertical profile E on the northern flank of Mt
Albreda. Samples 12-17 tp 12-19 were sampled along a logging road whilst sample PCA-200-83 was collected by
helicopter from a nearby peak. The location of PCA-200-83 was approximated by elevation on the elevation
profile; (bottom) Elevation profile with apatite (U-Th)/He, apatite fission track and zircon (U-Th)/He results; (top
right) Age elevation plot resulting from this data. Errors given are 1σ for AFT and 2σ for AHe ages. (Background
image: Google Earth, 2015)

104 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

3.2.2 Mica Creek Transect

Figure 3.2-5: Mica creek transect sample location map. The thick white line indicates the path of the elevation
profile. Black markers show the locations of bedrock samples. The location of vertical profiles A,B,D and F are
shown in pink. RMT: Rocky Mountain Trench. The location of the Sellkirk-Shuswap transect, which intersects this
transect north of Revelstoke, is marked by the white dashed line. (Background image: Google Earth, 2015)

Unlike the previously described Trench Transect which follows the strike of the
central axis of the range, all remaining transects are transverse to the RMT. The Mica
Creek Transect extends from Mica Dam, which is located near the centre of Kinsbasket
Lake and thus at the edge of the RMT, towards Revelstoke in the south, following the
course of the Columbia River. All samples were from the eastern flank of the valley
and are located in the hanging wall of the Columbia River normal Fault (and thus part
of the Selkirk Allochthon), which defines this valley. The strike of this transect is N-S
and all samples are bedrock samples from roadside outcrops at elevations between
454 m and 780 m a.s.l..

Both the apatite (U-Th)/He and fission track ages obtained along this transect
show Miocene ages, ranging from 10.1 ± 5.1 to 19.4 ± 12.9 Ma for AHe and a single

105 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

AFT age of 18.3 ± 1.1 Ma (Figure 3.2-6). Furthermore it appears that the AHe ages are
older towards the south. However the errors are also larger which is why this
observation is tentative. The zircon (U-Th)/He ages observed are Eocene, 40.8 ± 4.6
Ma in the north nearer the RMT and 44.9 ± 4.6 Ma observed in the south. There might
be a suggestion of younging to the north, but the zircon (U-Th)/He ages are the same
within uncertainty. The most striking feature of these dates is the young AHe ages at
low elevations of ~10-20 Ma.

Figure 3.2-6: Mica creek transect elevation profile with apatite (U-Th)/He (blue), apatite fission track (pink) and
zircon (U-Th)/He (green) ages at sample locations. Errors given are 1σ for AFT and 2σ for AHe ages. Pink bars A-G
indicate where the vertical profiles either directly connect to or can be projected perpendicular onto this
transect.

3.2.2.1 Vertical Profile F

Vertical profile F is located at the northern end of the Mica Creek transect near
the Mica Creek dam and consists of 3 samples: two roadside outcrop samples at
relatively low elevations (610m and 780m) and a sample obtained from the peak of Mt
Trident (Figure 3.2-7) at 3100m a.s.l., the highest elevation sample of the sample suite.
The profile covers an elevation range of 2500m over a vertical distance of
approximately 35km; sample 11-AS-12 however was not sampled along the direct
profile line and its location was projected onto it perpendicular to the transect.

106 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

3500 y = 67.36x - 452.69


R² = 1
3000
2500

Elevation (m)
2000
1500 AHe

1000 AFt

500 ZHe

0
0 10 20 30 40 50 60 70
Age (Ma)

Figure 3.2-7: (Top left) Sample locations and elevation profile line of vertical profile F south of Mica creek, from
the level of Lake Revelstoke to the top of Trident Mountain; (bottom) Elevation profile with apatite (U-Th)/He,
apatite fission track and zircon (U-Th)/He results. The position of sample 11-AS-12 was projected perpendicular
to the profile line onto the elevation profile; (top right) Age elevation plot resulting from this data. Errors given
are 1σ for AFT and 2σ for AHe ages. (Background image: Google Earth, 2015)

The wide spacing of this ‘profile’ is not ideal, as a typical ‘vertical profile’ should be
sampled over a small distance and as vertical as possible. It was nevertheless included
because it is the only location where samples with considerable elevation difference
were available in the most central region this close to the RMT.

The two low elevation samples were also part of the Mica Creek transect and
thus previously shown to have Miocene AHe and AFT and Eocene ZHe ages. The high
elevation sample at 3100m has an Eocene AFT age of 52.7 ± 7.6 Ma, indicating that the
higher part of this region cooled below ~120°C during this time (Figure 3.2-7). The ZHe
age at significantly lower elevations indicate that lower elevation rocks cooled below
~180°C during the Eocene and subsequently crossed the AFT PAZ at approximately
~120°C in the Oligocene/lower Miocene, followed by a cooling below ~70°C during the
upper Miocene, as indicated by the low elevation AHe age. These later phases of low

107 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

temperature cooling were not recorded by the high elevation AFT age, indicating that
the higher elevations had already cooled below ~120°C during the earliest Eocene
cooling phase. Unfortunately there is no AHe age for higher elevations available to
consolidate this further and shed light on the lowest temperature history. Based on a
lack of sufficiently dense data along this profile it is also not possible to determine the
rate of cooling through the closure isotherms or whether this occurred as one slow or
multiple stages of cooling. The best guess age elevation relationship based on two
available AFT samples equals a rate of 10 Ma increase in age with each 220 m increase
in elevation.

Figure 3.2-8: HeFTy thermal history inverse model result and apatite fission track length distribution for sample
11-AS-12. The only constraints imposed on the inverse models are the known mica Ar-Ar cooling ages ranging
from 45-80 Ma at 300°C (box constraint used during modelling, not depicted on scale) and the present day
average surface temperature of 10 ± 10°C. Green envelope marks the range of acceptable T-t paths found (50%
confidence level, so 50% of observed data (single grain ages, not displayed) could have been produced by these
paths) and magenta the range of good paths (95% confidence level). The model was run until 7 good and 106
acceptable paths were found. The thick black line indicates the best fit cooling path for the dataset.

The thermal history model for sample 11-AS-12 produced by inverse modelling
the acquired data using Hefty is shown in Figure 3.2-8. For all of the HeFTy models
presented in this study the boundary conditions were set based on an average annual
surface temperature range of 10 ± 10°C (for all elevations) and previously published
Ar-Ar (biotite) data form the Columbia Mountains which indicates cooling below 300°C
between 45-80 Ma (see chapter 1 for detailed summary and references). The inverse
modelling result indicates that the sample was initially exhumed through the ZHe PRZ

108 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

during the upper cretaceous, followed relatively slow cooling at a rate of 4°C/m.y.
during the Eocene and Oligocene. This slow cooling through the AFT Partial annealing
zone (120-90°C) concurs with a wide track length distribution and a relatively short
mean track length of 12.5µm. In the upper Miocene a sharp increase in cooling rate to
9°C/m.y., eventually bringing the sample up to surface temperatures. Although no AHe
age was available for this sample, the nearby sample 11-AS-13 at a similar elevation
has an AHe age of 10.1 ± 5.1 Ma and thus agrees with model predicting rapid cooling
below ~70°C at this time.

3.2.2.2 Vertical Profile G

Vertical profile G is located approximately 50 km west of the mica creek


transect line at the edge of the Monashee Mountains and was collected by A.
Halverson and B. Guest (University of Calgary). The profile extends for approximately
8km from the Yellowhead Highway to the top of Ptarmigan Mountain, covering an
elevation range of just over 1600m (Figure 3.2-9). Both AHe and ZHe data obtained for
this profile show Eocene ages throughout, ranging from 31.8 ± 4.6 Ma to 51.1 ± 32.6
Ma for apatite and from 39.3 ± 13.8 Ma to 50.6 ± 22.6 Ma for zircon. Although the 2σ
errors for both AHe and ZHe data are particularly large for these samples, the overall
age-elevation relationship still shows a positive correlation with the age increasing by
9 Ma with every 1000m climbed.

3.2.2.3 Vertical Profile B

Vertical profile B includes eight bedrock samples, covering an elevation range


of 1500m and sampled over a distance (from lowest to highest elevation sample) of
12.5km. It is located in the Monashee Mountains on the logging road leading up to
Pettipiece Pass, where the profile section continues south-eastward towards Feline
Peak, where the highest sample is located.

109 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

2500
AHe
2000
ZHe

Elevation (m)
1500

1000 y = 109.15x
- 3342.5
R² = 0.5103
500
0 20 40 60 80
Age (Ma)

Figure 3.2-9: (Top left) Sample locations and elevation profile line of vertical profile G, on the western flank of
Ptarmigan Mountain; (bottom) Elevation profile with apatite (U-Th)/He and zircon (U-Th)/He results; (top right)
Age elevation plot resulting from this data. The trendline shown was calculated for ZHe data. Errors given are 1σ
for AFT and 2σ for AHe ages. Data for this profile was obtained and kindly provided by my co-authors
A.Halverson and B.Guest. (Background image: Google Earth, 2015)

With the exception of the highest elevation sample, AHe ages of this profile are
predominantly Oligocene in age, ranging from 19.3 ± 2.6 Ma to 36.3 ± 2.3 Ma and
increasing with elevation (Figure 3.2-10). Interestingly, the top of Feline Peak at 2320
m (sample JC-311-94) gives a middle Eocene age of 45.3 ± 5 Ma causing the AHe age-
elevation relationship to ‘bend’ to a shallower gradient at the top, indicating a possible
‘exhumed PRZ’. In such a case samples from a narrow elevation range which
experienced considerable time in the PRZ temperature window would result in a sharp
increase in age with little elevation climbed which would manifest in a near horizontal

110 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

age-elevation relationship. Unfortunately, no higher elevation samples could be


obtained to confirm this. The Oligocene cooling recorded by the AHe data is in
agreement with the AFT data, which also increases with elevation (within error) from
21.2 ± 1.2 Ma to 33.3 ± 2.1 Ma. The average ZHe ages are both lower Eocene (52.7 ± 3
Ma and 51.3 ± 1.7 Ma). The age elevation relationship for AFT data suggests a steep
increase in age of 10 Ma/ 950 m of elevation.

3000 y = 79.032x - 885.65


R² = 0.4989
2500

Elevation (m) 2000

1500
AHe
1000
AFt
500
ZHe

0
0 10 20 30 40 50 60 70 80
Age (Ma)

Figure 3.2-10: (Top left) Sample locations and elevation profile line of vertical profile B, following the logging
road eastward to Pettipiece Pass and then towards the SE to the top of Feline Peak; (bottom) Elevation profile
with apatite (U-Th)/He, apatite fission track and zircon (U-Th)/He results; (top right) Age elevation plot resulting
from this data. The trendline shown was calculated for AFT data. Errors given are 1σ for AFT and 2σ for AHe ages.
(Background image: Google Earth, 2015)

The thermal history model for sample JC-60C-94 produced by inverse


modelling the data using Hefty is shown in Figure 3.2-11. The best fit result (black line)
indicates that the sample experienced a short period of fast cooling of >10°C/m.y.
through the apatite fission track partial annealing zone (PAZ) around 35 Ma ago. This

111 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

in confirmed by a relatively narrow track length distribution. Average cooling since 30


Ma is approximately 3°C/m.y., although the wide range of suitable solutions (as
indicated by the pink envelope) might mask modest changes in rate over this period of
time.

Figure 3.2-11: HeFTy thermal history inverse model result and apatite fission track length distribution for sample
JC-60C-94. The only constraints imposed on the inverse models are the known mica Ar-Ar cooling ages ranging
from 45-80 Ma at 300°C (box constraint used during modelling, not depicted on scale) and the present day
average surface temperature of 10 ± 10°C. Green envelope marks the range of acceptable paths found (50%
confidence level, so 50% of observed data (single grain ages, not displayed) could have been produced by these
paths) and magenta the range of good paths (95% confidence level). The model was run until 500 good paths
were found. The thick black line indicates the best fit cooling path for the dataset.

3.2.2.4 Vertical Profile A

Vertical profile A is located approximately 3km east of Three Valley Lake and
covers an elevation range of 550m over a distance of 1.5 km. Stratigraphically its
location is interesting because it is located in the Monashee Mountains, in the hanging
wall of the Monashee decollement fault which separates the Proterozoic sediments of
the cover from the paragneissic basement complex. Even though of the four samples
collected along this profile only two could be dated, this profile was nevertheless
included because it returned an interesting difference in ages. The AFT results showed
the oldest measured AFT data of the study area at relatively moderate elevations:

112 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

Eocene and upper Palaeocene ages of 47.3 ± 2.4 Ma and 61.2 ± 6.5 Ma (Figure 3.2-12).
In contrast the one AHe age of the lower elevation sample is one of the youngest in
the entire suite: 10.1 ± 0.8 Ma, and thus late Miocene. Although the highest elevation
AFT result appears to be too old compared to data previously presented in this
chapter, the lower elevation AFT result agrees well with a previously published age at
almost the same locality (44.3 ± 4.2 Ma, (Lorencak et al. 2001)) and therefore confirms
the consistency of these ages. There was no AHe age reported in that study, which is
why no comparison can be made here, however AHe ages described above on the
nearby Mice Creek Transect (Figure 3.2-6) are similarly young.

2000

y = 38.058x
1500 - 858.12
Elevation (m)

1000
AHe
500
AFt

0
0 20 40 60 80
Age (Ma)

Figure 3.2-12: (Top left) Sample locations and elevation profile line of vertical profile A between Victor lake and
Three Valley Gap; (bottom) Elevation profile with apatite (U-Th)/He and apatite fission track results; (top right)
Age elevation plot resulting from this data. Errors given are 1σ for AFT and 2σ for AHe ages. (Background image:
Google Earth, 2015)

The thermal history models for both samples are shown in Figure 3.2-13. The
high elevation sample 12-AS-05 shows rapid >10°C/m.y. cooling through the apatite

113 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

fission track PAZ (120-90°C) occurred during the upper Cretaceous and this fast rate is
consistent with the narrow track length distribution observed. From the Palaeocene
through to the upper Miocene cooling was slow at <1°C/m.y. but increased slightly to
approximately 3°C/m.y. during the last 20 m.y.. For the lower elevation sample 12-AS-
06 the best fit result indicates that the sample experienced rapid cooling of

Figure 3.2-13: HeFTy thermal history inverse model result and apatite fission track length distribution for samples
12-AS-05 (top) and 12-AS-06 (bottom). The only constraints imposed on the inverse models are the known mica
Ar-Ar cooling ages ranging from 45-80 Ma at 300°C (box constraint used during modelling, not depicted on scale)
and the present day average surface temperature of 10 ± 10°C. Green envelope marks the range of acceptable
paths found (50% confidence level, so 50% of observed data (single grain ages, not displayed) could have been
produced by these paths) and magenta the range of good paths (95% confidence level). The model was run until
100 good paths were found. The thick black line indicates the best fit cooling path for the dataset.

approximately 10°C/m.y. during the upper Eocene and was subsequently held under
slow 1°C/m.y. cooling conditions at 100-80°C from Palaeocene to mid Miocene. This
cooling hiatus occurred at slightly higher temperatures than for the higher elevation

114 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

sample, causing a wider track length distribution and the shorter mean track length.
Cooling at lower elevations later increased to 7°C/m.y. over the past 12 m.y..

The elevation difference between these two samples is only 500 m and yet the
thermal histories illustrate several phases of exhumation. The Late Cretaceous-Early
Eocene rapid cooling observed for the high elevation sample occurred during the
known last phase of mountain building and the later lower Eocene extension which
lead to the exhumation of the core complexes and most certainly caused some rock
uplift and substantial erosion in this region. At this time the lower elevation sample
appears to have still been buried and remained at higher temperatures. The
subsequent slow cooling of this sample could be related to gradual erosion and
incision of the valley following the potential uplift. Both samples also display a later,
Miocene increase in cooling and thus likely increased erosion and incision of deeper
valleys within the area.

3.2.2.5 Vertical Profile D

Vertical profile D is located on the western flank of Mount Revelstoke. With the
six available bedrock samples covering an elevation range of almost 1000m over a
horizontal distance of only 6km this profile is the most narrowly sampled one and thus
the most ‘truly vertical’ profile example that could be obtained for this study (Figure
3.2-14). With the exception of the lowest bedrock sample (80-RP-R1b) all samples are
located on the hanging wall of the east-dipping extensional Columbia River Fault,
which strikes parallel to the river and defines the valley.

The apatite (U-Th)/He ages on this profile increase with elevation from early
Miocene (19.4 ± 2.9 Ma) to mid Eocene (45.2 ± 16.9 Ma) at the top, while the apatite
fission track ages are consistently Oligocene, ranging from 24.5 ± 2.5 Ma to 34.5 ± 4
Ma (Figure 3.2-14). The most reliable and well-constrained age-elevation relationship
is given by the AFT data during the Oligocene. In this case, the age increases by 19 Ma
for every kilometre increase in elevation. Similar to profile B the uppermost AHe age is

115 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

considerably older (45.2 ± 16.9 Ma) than the other available ages from lower
elevations, causing the age-elevation relationship to ‘kink’ and thus hinting at a
potential ‘exhumed PRZ’.

2500 y = 107.08x - 1894.2


R² = 0.5974
2000

1500

Elevation (m)
1000 AHe

AFt
500 ZHe

0
0 10 20 30 40 50 60 70
Age (Ma)

Figure 3.2-14: (Top left) Sample locations and elevation profile line of vertical profile D on the western flank of
Mt Revelstoke; (bottom) Elevation profile with apatite (U-Th)/He, apatite fission track and zircon (U-Th)/He
results. Black markers indicate bedrock samples, while the white marker for GF 66 indicates a modern detrital
river sample (AFT age is pooled age, single population) Samples that do not lie directly on the profile lines were
projected onto the elevation profile at the relevant elevation; (top right) Age elevation plot including AFT
trendline resulting from this data. Errors given are 1σ for AFT and 2σ for AHe ages. (Background image: Google
Earth, 2015)

The ZHe ages on this profile increase from mid Eocene (41.5 ± 5.2 Ma) to
lowest Paleocene (63.9 ± 27.9 Ma) with elevation. At first the curve of the ZHe age
elevation relationship appears to indicate a potentially exhumed partial annealing
zone. However, the large error of the highest sample (which is due to a positive eU
correlation, to be discussed in section 3.2.5) leaves open the possibility that the slope
of the ZHe age-elevation relationship is similar to the AFT result.

116 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

The thermal history model results for the mid elevation sample 12-AS-14 are
shown in Figure 3.2-15. Despite being located at a relatively low elevation, the best fit
cooling result shows gradual cooling at a rate of approximately 4°C/m.y. since the
upper Cretaceous until present day. Previously presented thermal histories of nearby
samples of similar elevation (e.g. sample 12-AS-05, vertical profile A) showed much
more distinct phases of cooling intermitted by phases where extremely low cooling
rates dominated. Furthermore the long mean track length of 13.3 µm and the
relatively sharp track length distribution appear to contradict the thermal model
result, as slow cooling should have enabled earlier tracks to heal over time, causing a
wider distribution and reducing the mean length.

Figure 3.2-15: HeFTy thermal history inverse model result and apatite fission track length distribution for sample
12-AS-14. The only constraints imposed on the inverse models are the known mica Ar-Ar cooling ages ranging
from 45-80 Ma at 300°C (box constraint used during modelling, not depicted on scale) and the present day
average surface temperature of 10 ± 10°C. Green envelope marks the range of acceptable paths found (50%
confidence level, so 50% of observed data (single grain ages, not displayed) could have been produced by these
paths) and magenta the range of good paths (95% confidence level). The model was run until 100 good paths
were found. The thick black line indicates the best fit cooling path for the dataset.

The modern detrital sample GF 66 taken from the shore of the Columbia River
in the town of Revelstoke presented with a single population of 30.0 ± 1.8 Ma. The
catchment area for this sample is not depicted, as the Columbia River flows south at
this point, having exited Kinbasket Lake at Mica Dam (near vertical profile F) and thus
effectively drains the entirety of the eastern Columbia and western Rocky Mountains
(i.e. the Rocky Mountain Trench). Thus a detrital sample presenting with a single

117 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

population (for radial plots of all detrital samples see Appendix 3.2) could be
surprising, however the gradient and thus flow rate is slow, potentially inhibiting long
distance sediment transport and preventing mixing of different age populations.

3.2.3 Selkirk-Shuswap Transect

Figure 3.2-16 Selkirk-Shuswap transect sample location map. The thick white line indicates the path of the
elevation profile. White squares indicate sampling locations of modern detrital river sediements (samples GF61,
63 and 67) while black squares show the locations of bedrock samples. For samples GF63 and GF67 the yellow
shaded area shows the extent of the corresponding catchment area where sediment could have originated. For
sample GF61 the potential catchment area is indicated by the thick yellow line. RMT: Rocky Mountain Trench.
The location of the Mica Creek transect, which intersects this transect north of Revelstoke, is marked by the
white dashed line. The locations of previously discussed vertical profiles A and D are marked by pink boxes.
(Background image: Google Earth, 2015)

This transect extends from the Rocky Mountain Trench, approximately 30 km


north of Golden, perpendicular to the RMT towards the west for over 175 km, crossing
the Selkirk Mountains and a narrow part of the Monashee Mountains, ending on the
Shuswap Highland near Vernon. Although the maximum elevations of these ranges

118 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

reach almost 3000m, all samples along this transect were collected at relatively low
elevations in valleys, ranging from 375 m to 839 m a.s.l. (Figure 3.2-16).

Figure 3.2-17: Selkirk-Shuswap transect elevation profile with apatite fission track (pink) ages at sample
locations. Black markers indicate bedrock samples and white markers show the location of modern detrital
samples (GF61, 63 and 67). Both modern sediments GF 61 and 63 presented a single population of AFT ages,
while GF 67 is comprised of 2 populations, indicated by (1) and (2). Errors given are 1σ for AFT ages. The locations
of previously discussed vertical profiles A and D are marked by pink boxes.

The first two samples in the east are modern detrital sediment samples. GF 61
was collected on the upper Columbia River near Donald, before the river enters
Kinbasket Lake. At this point its catchment area spreads over almost 11800km 2 and
covers elevations up to 3200m. The catchment area of GF 63 is much smaller, draining
the northernmost part of the Purcell Mountains and the eastern flank of the Selkirk
ranges, covering elevations up to 3377 m (Mt Dawson) over and area of 580km 2.
Despite their difference in catchment size and drainage area location both samples
yielded Oligocene AFT ages (30.5 ± 1.5 Ma and 29.7 ± 3.9 Ma) in single populations
(Figure 3.2-17). Modern detrital sample GF 67 is located near the western end of the
transect. The sampling location is the Perry River below the Trans-Canada Highway
near the town of Malakwa. It drains approximately 740 km2 of the western Monashee
Mountains with topography including Bourne Glacier, Jordan Range and Cat Peak that
reach a maximum elevation of 2740 m a.s.l. It mainly includes the western side of the
crest of the range. The central age for this sample was determined to be 37.7 ± 2.2
Ma, containing two major populations of 29.3 ± 1.5 Ma (21 crystals) and 49.2 ± 3.5 Ma
(10 crystals). The bedrock samples along this transect are lower in elevation in general,
and despite the elevation further decreasing towards the west the AFT ages increase

119 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

from lowest Miocene (20.1 ± 2.9 Ma) to upper Eocene (39.7 ± 5 Ma) in the same
direction.

3.2.4 Southern Transect

Figure 3.2-18: Southern transect sample location map. The thick white line indicates the path of the elevation
profile. Samples that do not lie directly on the line were projected perpendicularly onto the profile. Black squares
show the locations of bedrock samples. The location of vertical profile C is shown by the pink square.
(Background image: Google Earth, 2015)

The 10 samples collected in the south can be grouped into one NE-SW trending
transect (9 samples) and one additional bedrock sample at the southern end of
Kootenay Lake (Figure 3.2-18). The transect elevation profile extends westward from
the trench, perpendicular to the main axis i.e. the RMT, crossing the northern part of
the Purcell Mountains and continues on towards the southwest crossing the southern
part of the Selkirk Mountains (Valhalla Complex). The observed ages along this
transect are displayed in Figure 3.2-19.

120 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

Figure 3.2-19: Southern transects elevation profile with apatite (U-Th)/He (blue) and apatite fission track (pink)
ages at sample locations. Errors given are 1σ for AFT and 2σ for AHe ages. The location of vertical profile C is
indicated by the pink bar.

The samples along this transect range in elevation by 1860 m and thus the ages
are expectedly varied as well. Low elevation AFT ages range from 24.9 ± 3.5 Ma (in the
middle of the transect) to 34.0 ± 2.7 Ma at the south-western edge of the range
(Figure 3.2-19). With rising elevation the AFT ages increase up to 44.4 ± 2.4 Ma at 2300
m (sample PCA-121-84). The AHe ages are mostly from medium elevation samples and
range from 22.8 ± 3.3 Ma in the trench to 37.6 ± 1.7 Ma in the south west, increasing
in age with increasing distance to the RMT. The only exception is the apparently
inverted age elevation relationship of samples 12-AS-11 and 12-AS-12, with the lower
elevation sample showing an older AHe age (31.5 ± 10.3 Ma) being older that the
sample 500 m above (24.7 ± 6.4 Ma). However the 2 σ errors are large and overlap,
making the ages plausible. Similarly, sample 12-RP-AS-07 (AHe age 34.3 ± 3.4 Ma and
AFT age 28.5 ± 1.8 Ma) and the isolated sample 12-RP-AS-04 (AHe age of 48.0 ± 6.2 Ma
and AFT age of 33.8 ± 4.0 Ma, ages on Figure 3.2-18) which is not part of this transect
but rather located 75km to the east at the southern end of Kootenay Lake, show
inverted ages within the same sample. In these cases the AHe 2σ and AFT 1σ error,
despite being large, do not overlap. This phenomenon is technically impossible,
because the AHe closure temperature of 70°C is lower than the AFT Tc of 120°C, and
thus heating the rock to at least partially reset the AFT age would inevitable reset the
AHe age as well, yet is has been occasionally observed in thermochronological
datasets. Potential reasons for these observations, such as short term re-heating or

121 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

radiation damage, will be discussed in the discussion of the data below in section
3.2.5.

The thermal history model of sample PCA-58-84 (at the SW end of the transect)
indicates rapid cooling (9°C/m.y.) through the AFT PAZ during the upper Eocene. It
furthermore remained at slightly higher temperatures of approximately 60°C until it
was rapidly exhumed (>10°C/ m.y.) over the past 5 m.y..

Figure 3.2-20: HeFTy thermal history inverse model result and apatite fission track length distribution for sample
PCA-58-84. The only constraints imposed on the inverse models are the known mica Ar-Ar cooling ages ranging
from 45-80 Ma at 300°C (box constraint used during modelling, not depicted on scale) and the present day
average surface temperature of 10 ± 10°C. Green envelope marks the range of acceptable paths found (50%
confidence level, so 50% of observed data (single grain ages, not displayed) could have been produced by these
paths) and magenta the range of good paths (95% confidence level). The model was run until 500 good paths
were found. The thick black line indicates the best fit cooling path for the dataset.

3.2.4.1 Vertical Profile C

Vertical Profile C was sampled on the eastern edge of the Valhalla complex at
the southern end of Slocan Lake. The three samples along this profile cover and
elevation range of 1750 m over a horizontal distance of 15 km (Figure 3.2-21). The
individual ages have already been presented as part of the southern transect (above).
The age-elevation relationship from the obtained data, shown in Figure 3.2-21,
indicates an increase in age of 85 Ma with every kilometre elevation climbed.

122 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

2500.0
y = 84.997x - 1439.9
2000.0 R² = 0.9099

Elevation (m)
1500.0

1000.0
AHe
500.0
AFT

0.0
0 10 20 30 40 50
Age (Ma)

Figure 3.2-21: (Top left) Sample locations and elevation profile line of vertical profile C crossing the slocan lake
valley from east to west; (bottom) Elevation profile with apatite (U-Th)/He and apatite fission track results; (top
right) Age elevation plot for this data. Errors given are 1σ for AFT and 2σ for AHe ages. (Background image:
Google Earth, 2015)

The thermal history results for the two samples along this profile where AFT
track lengths could be measured and consequently a useful model result could be
determined are shown in Figure 3.2-22. The higher elevation sample PCA-121-84
begins with an lower Eocene rapid >10°C/ m.y. cooling event, after which the sample
gradually cooled through lower temperatures from 40°C to surface temperatures at a
rate of just below 1°C/ m.y.. The lower elevation sample PCA-529-83 indicates rapid
Oligocene cooling (11°C/ m.y. ) through the AFT PAZ followed by a hiatus (1°C/ m.y.)
during the 80-60°C cooling interval and final rapid cooling (>10°C/ m.y.) over the last 5
m.y.. Similar to vertical profile A (Figure 3.2-13), the later exhumation of the lower
elevation sample and the longer residence at higher temperatures has caused a
slightly wider track length distribution and an overall shorter mean track length.

123 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

Figure 3.2-22: HeFTy thermal history inverse model result and apatite fission track length distribution for samples
PCA-529-83 and PCA-121-84. The only constraints imposed on the inverse models are the known mica Ar-Ar
cooling ages ranging from 45-80 Ma at 300°C (box constraint used during modelling, not depicted on scale) and
the present day average surface temperature of 10 ± 10°C. Green envelope marks the range of acceptable paths
found (50% confidence level, so 50% of observed data (single grain ages, not displayed) could have been
produced by these paths) and magenta the range of good paths (95% confidence level). The model was run until
500 good paths were found. The thick black line indicates the best fit cooling path for the dataset.

3.2.5 Data Discussion

3.2.5.1 (U-Th)/He Data

Traditionally, (U-Th)/He ages presented in publications are aliquot averages of


3-5 single grain analyses and the errors presented are that given by the resulting
standard deviation. This ‘average’ is applied because in Helium dating the analytical
errors are typically very small (<3%) compared to the error introduced by the

124 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

reproducibility of individual grain ages. Often a significant spread (multiple 10s%) of


ages amongst individual aliquots can be observed.

All of the sample aliquot ages presented in this study were obtained by
measuring 2-5 crystals and most the individual ages within samples agree fairly well,
however a closer look at the individual grain data (Table 3.3-2 for AHe and Table 3.3-3
for ZHe) reveals that there are a few samples where ages vary by up to 30Ma for
apatite and 70Ma for zircon. As described in chapter 2 the carefully conducted picking
procedure should eliminate complicating factors such as U/Th rich inclusions
(crystalline or fluid) which could cause changes in observed ages. Furthermore Helium
implantation was deemed to be unlikely to have a significant influence on variations in
data because it was found to be a rare effect (Brown et al. 2013). Because all AHe and
ZHe data was acquired from bedrock samples, all grains from the same sample should
yield the same age. Even metasedimentary samples which might contain minerals
from varying initial rock types should yield similar ages for these low T
thermochronology methods, because these are sensitive to temperatures well below
low grade metamorphism which they experienced before being exhumed.
Furthermore metasedimentary samples containing apatites and zircons from different
initial sources often contain crystals with varying U/Th content. These, too, should
yield the same ages since the (U-Th)/He method uses parent-daughter isotope ratios
rather than absolute values to determine the age.

However, as previously described (see chapter 2), variability of U/Th content,


more specifically radiation damage effective uranium (eU= 238U(ppm)+ 0.235
232
Th(ppm)) can cause considerable variation of closure temperatures between
individual aliquots and thus, especially for slowly cooled samples, induce a wide
spread of ages within the same sample. This is due to an increase in radiation damage
with increasing eU, causing Helium traps to form in the crystals and blocking 4He from
diffusing out of the crystal, making the age older compared to undamaged samples
(Flowers et al. 2009). Trapped Helium also requires a higher energy to re-enter the
crystal structure and finally diffuse out of it, which means that with increasing

125 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

radiation damage the closure temperature of the crystal increases as well, recording
an older age.

In order to evaluate to what extent this effect might be responsible for the
spread in aliquot ages presented in Table 3.2-2 and Table 3.3-3 eU versus age plots
were created for all samples and their correlations analysed by plotting trend lines for
each sample. All plots which resulted in positive correlations are presented in Figure
3.2-23 (eU vs age plots for all samples can be found in Appendix 3.1). Examples for
strongly positive correlations within the AHe data are 11-AS-13 and JC-321-94, where
aliquot ages of these samples differed from the averaged mean age by up to 30%. The
eU positive correlations indicate that radiation damage is most likely the reason for
the observed spread in ages, as higher eU samples have higher closure temperatures.
Likewise for the ZHe dataset (Table 3.3-3) samples 11-AS-11 and 12-AS-16 show large
variation between aliquot ages and produce a strongly positive eU correlations,
indicating that they have experienced some degree of radiation damage.

It has to be noted that none of the individual ages can be considered ‘wrong’
and thus excluded from the mean age, they are simply a reflection of how the
chemistry affects closure temperatures. Aside from calculating an individual closure
temperature for each crystal (which – although time consuming - can be done using
thermal modelling software, as described in chapter 2), currently the only way to
ensure a completely comparable dataset would be to date numerous (20+) crystals for
each sample and chose 3-5 aliquots per sample which show a similar chemical
composition (eU) and thus all would have the same closure temperature. In practice,
because of financial and temporal constraints, this is not feasible for a normal project
and thus we are currently left with the option to date multiple crystals and averaging
the age, assuming that the resulting (U-Th)/He dates will be comparable to an
‘average’ apatite chemistry. Furthermore, even though current research is being
conducted on the topic, we have not yet been able to find a satisfactory explanation
for occasionally observed strongly negative correlations which are sometimes

126 | P a g e
AHe positive eU correlations ZHe positive eU correlations
100 1000
11-AS-11
12-AS-16
12-AS-17
zAH49-1

Age (Ma)
Age (Ma)

11-AS-1
10 11-AS-13 100
12-AS-08
12-AS-11
12-AS-12
12-AS-13
JC-311-94
JC-321-94
AH48-2
AH48-1
1 10
1 10 100 10 100 1000 10000
eU (ppm) eU (ppm)
Figure 3.2-23: Positive correlations of effective Uranium (eU) vs. age for AHe (left) and ZHe (right) data. Positive correlations indicate that the spread in single ages is most likely attributed
to radiation damage and thus He-retention. All other eU plots for samples showing no or negative correlations can be found in the Appendix.

127 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

observed (an example from this study is sample 11-AS-22, see Table 3.3-3 and eU plot
in Appendix x). Although it has been suggested that extreme damage could reverse the
eU effect by effectively destroying the crystal lattice and letting 4He escape much
faster (Farley 2000, Nasdala et al. 2004), it is not yet known when this relationship
‘turns’. Additionally experimental research has shown that not only the total eU
content but also the distribution of eU, i.e. zonation of U/Th and the resulting
radiation damage may play a part in increasing diffusion in some areas within the
crystal and inhibiting it in others (Farley 2000). This has been found to be responsible
for various modest age dispersions found in AHe datasets (Ault and Flowers 2012),
and thus could also explain some of the variability in the ages observed in this study.

A lot of the recent research in the field of thermochronology has investigated


these effects and we clearly still need to gain a better understanding in order to
improve our interpretations and eventually even use the observed effects to our
advantage. For now we can only report eU and remind ourselves that mean (U-Th)/He
ages with large 2σ errors are not ‘bad’, but that they can be the result of combining
aliquots with varying closure temperatures.

3.2.5.2 Fission Track Data

The apatite fission track data for samples collected by this study are presented
in Table 3.2-4, while Table 3.2-5 (additional bedrock samples) and Table 3.2-6 (modern
detrital river sediments samples) contain data obtained from R. Parrish, with the latter
data being measured in ~1982 using the population method without track length
measurement. It has to be noted that the average AFT ages given in Table 3.2-5 (1982
data) are pooled ages, which were calculated by ‘pooling’ all induced and spontaneous
tracks of individually measured grains and calculating the resulting ‘averaged age’.
Because this method does not contain any form of statistical weighting of individual
data and introduced bias when the age distribution within a sample deviates from a
normal poisson distribution, it was gradually replaced by the ‘Central age’ during the

128 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

early 1990s. This method presents a weighted mean value of the logarithmic normal
distribution of individual grain ages and is thus more robust in regards to outliers and
non-poissonian age distributions (Galbraith and Laslett 1993, Gallagher, Brown, and
Johnson 1998). Data presented in Table 3.2-4 and Table 3.2-6 (new data) are
therefore Central ages (see chapter 2 for more detail on method and on how these
ages are produced).

As described in chapter 2, the chemical composition of apatites can have


significant influence on the temperature sensitivity of the fission track system.
Depending on the dominant halogen endmember total annealing temperatures can
vary up to ±20°C, with higher temperatures being related to a higher proportion of
Chlorine end-members (Burtner, Nigrini, and Donelick 1994, Carlson, Donelick, and
Ketcham 1999). In order to ensure the quality of the observed ages the chemical
composition of all samples was estimated by measuring D-par, the average dimensions
of the track etch pit. Due to the different etching velocities of fluoro- chloro- and
hydroxyl- apatite this ‘D-par’ value is a first order estimate of the overall grain
chemistry. The observed D-par in Table 3.2-4 shows that most samples fell between
Dpars of 1.44µm and 2.08 µm and are thus typically Fluoro-apatites (Donelick 1993),
with the exception of sample 12-AS-08, which had a Dpar of 7.33µm. This extremely
high value could indicate OH-groups as the dominant halogen endmember, because
values for Chloro-apatites typically range from 4-5µm.

In order to confirm the observation of predominantly fluorine dominated


apatites and also to investigate the abnormally high Dpar of sample 12-AS-08, six
selected samples (Dpar > 2 or relative error > 10%) were analysed for their chlorine
and fluorine content (as a percentage of the composition of total material analysed)
using microprobe analysis. The results in Table 3.2-1 illustrate that all samples were
indeed fluorine dominated. Only sample 12-AS-05 showed a slightly elevated chlorine
content of 0.12 wt%, however this is not enough to significantly influence the AFT
annealing behaviour, since chlorine only becomes a dominant control on crystal-
structure above 0.35 wt% (Barbarand et al. 2003). Even sample 12-AS-08 which gave a
Dpar of 7.33 shows no significant Chlorine content and thus the large Dpar is most

129 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

likely attributed to the presence of Hydoxyl-endmembers. Because the presence of


OH-endmembers in apatites does not significantly influence the closure temperature
or annealing behaviour this does not pose a problem for the interpretation of the
presented results. We can therefore assume that all AFT data presented in this chapter
adheres to a standard closure temperature of ~120°C with the PAZ ranging from 90° to
120°C.

Table 3.2-1 Chemical analysis results for selected Apatite samples

AFT age 1σ error rel. error Dpar


Sample N F wt% Cl wt%
[Ma] [±Ma] [%] [µm]
12-AS-08 32.4 1.6 5 20 7.33 3.07 0.01
12-RP-AS04 33.8 4 11.8 20 1.81 2.32 0.01
12-AS-05 61.2 6.5 10.6 20 1.66 2.66 0.12
12-AS-15 34.5 4 11.7 20 1.82 3.10 0.03
PCA-307-83 52.7 7.6 14.5 21 2.03 2.15 0.02
PCA-200-83 16.5 1.9 11.7 20 1.48 2.76 0.01

Averaged microprobe chemistry results for 8 selected samples, chosen for high Dpar relative error
>10%. N= number of crystals analysed. All apatites are Fluorine dominated and only 12-AS-05
shows significant Chlorine content, however not enough to exert a dominant control on annealing
behaviour (<0.35wt%).

Lastly, two locations in the presented dataset showed AHe ages significantly
older than the corresponding AFT age of the same sample. Similar to the age
dispersion observed in the AHe data, the difference could also be attributed to the
effect of radiation damage on the AHe age. However, at vertical profile E where the
age difference of 10 Ma was consistent over the elevation range sampled, neither of
the samples 12-AS-19 and PCA-200-83 displayed positive correlations. Much rather the
observed age-eU relationship resulted in strongly negative correlations, indicating that
falsely old AHe ages due to radiation damage are likely not the reason for the
disparity. Similarly, two samples of the southern transects, 12-RP-AS-07 and 12-RP-AS-
08, which also presented AHe ages significantly older than AFT ages display neither
strongly positive nor negative eU correlations and thus it is unlikely that these results
are due to excessive radiation damage. An alternative explanation in these cases could

130 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

be a late and short term reheating event which has not yet significantly affected
diffusional Helium loss (based on discussions with colleagues D.Whipp and T.Ehlers at
AGU 2015). Because annealing of fission tracks happens at a much faster rate, short
bursts of increased temperatures could notably reduce the overall track density.
Especially for low U/Th crystals which display low track densities by nature, this could
significantly decrease the number of tracks observed on a randomly cut internal crystal
surface and thus decrease the AFT age calculated. In published literature, events such
as intrusion of thin and quickly cooled veins, the vicinity of a fast cooling and thin
basalt cover or even wildfires have been suggested as potential causes for such
curiously inverted age relationships (Stockli, Farley, and Dumitru 2000, Reiners et al.
2007). Although great care was taken not to sample near proven sources of heat, such
as hot springs or recent intrusive/extrusive rocks, we cannot entirely exclude these
causes in light of the observed data. Also, although no recent signs of fires were
present at the sampling locations and erosion in the humid and mountainous climate
of British Columbia is most likely to have removed rock surfaces that were exposed to
fires in the distant past, we cannot dismiss this possibility altogether either.

131 | P a g e
Table 3.2-2: Apatite (U-Th)/Helium replicate analyses and mean ages

Corr. Analytical Averaged Standard


width length Th Raw Age
Sample 4He (mol) U (ppm) Th/U eU+ Ft Age* Error** sample Deviation
(µm) (µm) (ppm) (Ma)
(Ma) (±2σ) Age (Ma) (Ma)

11-AS-1-1 120.0 210.2 4.07E-13 30.82 68.52 2.2 46.9 24.4 0.79 26.1 1.6 22.8 3.3
11-AS-1-2 86.0 113.4 2.41E-14 19.03 23.85 1.3 24.6 17.1 0.68 19.5 1.2
11-AS-1-3 90.7 137.9 4.37E-14 13.02 8.74 0.7 15.1 20.7 0.70 22.7 1.4

11-AS-9-1 79.9 103.4 2.68E-15 5.17 13.76 2.7 8.4 13.1 0.65 15.8 0.9 19.1 6.3
11-AS-9-2 89.9 79.3 8.12E-15 3.45 8.80 2.5 5.5 23.4 0.65 26.4 1.6
11-AS-9-3 82.0 111.0 3.25E-15 6.05 8.11 1.3 8.0 12.8 0.66 15.1 0.9

11-AS-11-1 76.6 113.2 4.56E-15 16.69 5.61 0.3 18.0 10.1 0.64 12.3 0.7 19.4 12.9
11-AS-11-2 83.6 106.2 1.47E-14 1.77 9.69 5.5 4.1 31.2 0.70 34.3 2.1
11-AS-11-3 98.0 134.7 3.73E-15 6.96 5.54 0.8 8.3 10.0 0.72 11.7 0.7

11-AS-13-1 79.8 109.7 1.56E-16 2.67 7.04 2.6 4.3 3.1 0.66 4.3 0.3 10.1 5.1
11-AS-13-2 63.1 78.2 6.78E-16 4.27 16.79 3.9 8.2 10.0 0.56 13.7 0.8
11-AS-13-3 70.8 78.7 7.23E-16 3.93 18.88 4.8 8.4 9.1 0.59 12.4 0.7

11-AS-15-1 74.9 103.2 1.84E-15 8.64 60.74 7.0 22.9 7.0 0.64 9.5 0.6 10.5 1.4
11-AS-15-2 67.6 146.8 2.33E-15 9.87 7.83 0.8 11.7 9.2 0.63 11.5 0.7

11-AS-22-2 114.3 119.4 5.65E-15 1.03 4.54 4.4 2.1 37.4 0.74 50.5 1.1 33.6 16.9
11-AS-22-3 129.2 187.9 1.21E-14 2.26 2.46 1.1 2.8 13.2 0.79 16.6 1.6

12-AS-06-1 96.0 250.8 1.91E-14 7.66 5.26 0.7 8.9 7.4 0.75 9.8 0.2 10.1 0.6
12-AS-06-2 98.0 218.3 1.91E-14 7.99 3.89 0.5 8.9 8.2 0.75 11.0 0.6
12-AS-06-3 127.1 169.5 6.09E-14 18.97 14.92 0.8 22.5 7.4 0.78 9.5 0.1

132 | P a g e
Corr. Analytical Averaged Standard
width length Th Raw Age
Sample 4He (mol) U (ppm) Th/U eU+ Ft Age* Error** sample Deviation
(µm) (µm) (ppm) (Ma)
(Ma) (±2σ) Age (Ma) (Ma)

12-AS-08-1 141.0 275.2 8.22E-14 6.23 5.40 0.9 7.5 15.5 0.82 18.9 0.2 19.3 2.6
12-AS-08-2 103.9 195.2 1.96E-14 5.45 4.91 0.9 6.6 12.2 0.76 16.2 0.7
12-AS-08-3 94.5 158.9 1.78E-14 5.73 5.16 0.9 6.9 17.0 0.73 23.4 1.6
12-AS-08-4 94.5 146.8 1.38E-14 5.72 6.91 1.2 7.3 13.5 0.72 18.8 0.7

12-AS-09-3 94.8 149.8 2.35E-14 7.02 5.22 0.7 8.2 19.3 0.72 26.7 3.2 30.2 3.5
12-AS-09-4 83.8 161.5 1.53E-14 4.76 5.77 1.2 6.1 23.5 0.70 33.7 2.1

12-AS-10-1 84.7 129.7 2.78E-14 7.45 21.15 2.8 12.4 20.5 0.69 29.8 0.3 26.2 2.6
12-AS-10-2 91.2 137.7 4.31E-14 10.45 28.18 2.7 17.1 17.6 0.71 24.9 0.2
12-AS-10-3 100.8 158.7 8.18E-14 14.24 32.89 2.3 22.0 17.7 0.74 24.0 0.2

12-AS-11-1 88.8 166.0 1.23E-13 25.98 90.06 3.5 47.1 15.1 0.71 21.2 0.2 24.7 6.4
12-AS-11-2 129.8 205.8 3.85E-13 27.40 93.08 3.4 49.3 16.7 0.80 20.9 0.2
12-AS-11-3 140.2 367.7 6.12E-13 19.88 69.08 3.5 36.1 17.3 0.83 20.8 0.2
12-AS-11-4 100.6 287.3 5.84E-13 27.39 114.38 4.2 54.3 27.4 0.77 35.8 0.3

12-AS-12-1 93.0 175.8 6.76E-14 8.45 14.40 1.7 11.8 31.2 0.73 42.9 0.7 31.5 10.3
12-AS-12-2 116.9 175.5 4.04E-13 36.65 50.01 1.4 48.4 25.9 0.77 33.6 0.4
12-AS-12-3 98.8 149.6 1.33E-14 2.48 7.37 3.0 4.2 13.7 0.77 17.9 0.3

12-AS-13-1 145.2 169.3 1.43E-13 11.68 5.92 0.5 13.1 23.5 0.80 29.3 0.6 25.6 8.8
12-AS-13-2 116.0 124.9 2.29E-14 9.46 7.58 0.8 11.2 10.0 0.75 13.4 0.5
12-AS-13-3 107.8 154.5 6.03E-14 9.10 7.32 0.8 10.8 25.6 0.75 34.1 1.2

12-AS-14-1 185.2 301.2 3.13E-13 7.35 3.88 0.5 8.3 27.5 0.86 32.1 0.4 30.7 2.5
12-AS-14-2 132.9 200.6 6.22E-14 5.76 3.55 0.6 6.6 21.5 0.80 26.9 1.1
12-AS-14-3 174.5 192.4 1.10E-13 4.85 1.92 0.4 5.3 27.9 0.83 33.5 1.1
12-AS-14-4 104.3 188.9 6.47E-14 8.54 10.86 1.3 11.1 23.0 0.76 30.4 0.3

133 | P a g e
Corr. Analytical Averaged Standard
width length Th Raw Age
Sample 4He (mol) U (ppm) Th/U eU+ Ft Age* Error** sample Deviation
(µm) (µm) (ppm) (Ma)
(Ma) (±2σ) Age (Ma) (Ma)

12-AS-15-1 89.8 143.4 3.01E-14 8.74 5.67 0.6 10.1 23.3 0.71 32.9 2.7 32.9 2.7

12-AS-16-1 78.5 198.5 1.55E-13 34.86 8.62 0.2 36.9 26.4 0.69 38.1 2.7 45.2 16.9
12-AS-16-2 115.1 163.8 1.67E-13 11.04 3.02 0.3 11.7 52.4 0.77 68.5 12.8
12-AS-16-3 97.0 135.2 3.86E-14 10.80 6.70 0.6 12.4 20.8 0.72 29.0 2.6

12-AS-17-1 152.1 187.7 3.23E-14 1.64 2.21 1.3 2.2 30.6 0.82 37.5 0.8 37.4 0.1
12-AS-17-2 152.2 288.7 1.55E-13 5.30 1.58 0.3 5.7 31.1 0.83 37.3 1.2

12-AS-18-1 102.8 110.4 5.42E-15 2.00 9.17 4.6 4.2 14.9 0.71 20.8 0.5 32.8 12.4
12-AS-18-2 137.8 150.2 2.49E-14 1.33 4.51 3.4 2.4 39.3 0.79 49.9 1.0
12-AS-18-3 116.5 123.6 4.23E-14 4.63 23.58 5.1 10.2 20.8 0.75 27.8 0.4

12-AS-19-1 142.7 243.0 1.43E-13 7.32 4.49 0.6 8.4 26.7 0.82 32.6 0.3 20.3 8.8
12-AS-19-2 135.5 287.7 7.84E-14 6.97 7.46 1.1 8.7 13.1 0.82 16.0 0.1
12-AS-19-3 92.2 181.4 1.80E-14 9.48 6.20 0.7 10.9 9.0 0.73 12.3 0.4

12-RP-AS-04-2 106.2 217.0 1.89E-13 14.03 18.27 1.3 18.3 31.9 0.77 41.7 0.3 48.0 6.2
12-RP-AS-04-3 108.7 181.4 1.30E-13 6.85 19.87 2.9 11.5 41.3 0.76 54.2 0.5

12-RP-AS-07-1 82.2 98.4 4.18E-14 18.88 29.48 1.6 25.8 20.4 0.65 31.2 0.4 34.3 3.4
12-RP-AS-07-2 85.0 187.1 1.59E-13 26.19 25.51 1.0 32.2 28.2 0.71 39.6 0.4
12-RP-AS-07-3 99.2 167.3 1.18E-13 17.13 19.25 1.1 21.7 25.7 0.74 34.9 0.4
12-RP-AS-07-4 95.3 149.6 2.13E-13 31.88 85.30 2.7 51.9 22.8 0.72 31.6 0.3

12-RP-AS-08-1 109.8 138.9 4.42E-14 5.94 8.82 1.5 8.0 27.5 0.75 36.9 0.5 37.6 1.7
12-RP-AS-08-2 131.9 156.9 6.35E-14 4.73 5.50 1.2 6.0 31.4 0.78 40.0 0.6
12-RP-AS-08-3 81.1 185.3 3.65E-14 7.09 12.75 1.8 10.1 25.1 0.70 35.9 0.5

134 | P a g e
Corr. Analytical Averaged Standard
width length Th Raw Age
Sample 4He (mol) U (ppm) Th/U eU+ Ft Age* Error** sample Deviation
(µm) (µm) (ppm) (Ma)
(Ma) (±2σ) Age (Ma) (Ma)

JC-16-03-1 92.0 144.9 8.33E-15 1.88 6.46 3.4 3.4 24.1 0.71 33.8 0.7 27.6 6.4
JC-16-03-2 86.7 142.8 3.71E-15 1.85 6.80 3.7 3.5 13.1 0.70 18.7 0.4
JC-16-03-3 94.1 104.8 1.55E-15 1.36 4.21 3.1 2.3 21.0 0.69 30.4 2.4

JC-311-94-1 126.4 270.9 5.74E-13 22.26 16.53 0.7 26.1 37.6 0.80 46.7 0.4 45.3 5.0
JC-311-94-2 82.9 171.1 6.29E-14 14.26 6.76 0.5 15.8 27.1 0.70 38.7 2.3
JC-311-94-3 115.7 185.7 3.60E-13 25.68 8.55 0.3 27.7 39.1 0.77 50.6 0.7

JC-321-94-1 86.5 123.9 2.59E-14 8.58 9.77 1.1 10.9 24.0 0.69 34.8 1.4 29.8 5.0
JC-321-94-3 70.1 146.4 7.85E-15 6.68 5.65 0.8 8.0 16.1 0.65 24.9 5.0

JC-60C-94-1 128.8 134.9 5.92E-14 6.05 10.60 1.8 8.5 25.5 0.77 33.1 0.5 36.3 2.3
JC-60C-94-2 137.8 183.4 9.87E-14 4.37 12.59 2.9 7.3 30.9 0.80 38.6 0.6
JC-60C-94-3 109.7 189.8 1.80E-13 14.40 28.88 2.0 21.2 28.5 0.77 37.2 0.5

PCA-200-83-1 102.4 164.6 1.01E-13 6.40 39.38 6.2 15.7 29.6 0.74 39.8 0.4 26.8 8.9
PCA-200-83-2 85.7 137.1 3.50E-14 7.61 30.30 4.0 14.7 20.3 0.69 29.3 0.3
PCA-200-83-3 70.8 153.5 9.88E-14 27.37 167.05 6.1 66.6 14.7 0.65 22.6 0.2
PCA-200-83-4 91.5 169.8 4.52E-14 10.55 48.76 4.6 22.0 11.4 0.72 15.7 0.2

PCA-304-83-1 83.6 135.1 7.99E-14 22.19 25.74 1.2 28.2 23.9 0.69 34.8 0.3 30.1 4.7
PCA-304-83-2 79.9 121.6 5.08E-14 20.28 46.10 2.3 31.1 16.9 0.67 25.4 0.2

AH48-2-1 128.0 168.0 5.36E-08 17.43 1.39 0.1 18.1 49.3 0.77 64.3 3.9 49.8 14.6
AH48-2-2 128.0 176.0 9.02E-09 12.89 1.40 0.1 13.5 27.0 0.77 35.2 2.1

AH49-3-2 96.0 140.0 7.07E-08 40.67 2.58 0.1 42.0 24.5 0.70 34.8 2.1 47.1 12.3
AH49-3-4 109.0 196.0 1.11E-07 27.56 2.53 0.1 28.8 44.4 0.75 59.4 3.6

135 | P a g e
Corr. Analytical Averaged Standard
width length Th Raw Age
Sample 4He (mol) U (ppm) Th/U eU+ Ft Age* Error** sample Deviation
(µm) (µm) (ppm) (Ma)
(Ma) (±2σ) Age (Ma) (Ma)

AH48-1-1 141.0 148.0 3.32E-07 29.02 2.34 0.1 29.7 75.4 0.78 97.3 5.8 51.1 32.6
AH48-1-2 118.0 234.0 2.67E-09 9.04 0.73 0.1 9.4 21.3 0.77 27.7 1.7
AH48-1-3 119.0 216.0 1.09E-08 18.06 1.22 0.1 18.6 21.8 0.77 28.4 1.7

AH47-1-1 136.0 143.0 9.46E-08 37.85 19.05 0.5 42.4 28.2 0.76 37.2 2.2 31.8 4.6
AH47-1-2 120.0 171.0 6.01E-08 42.76 26.19 0.6 49.0 19.5 0.75 26.0 1.6
AH47-1-3 127.0 160.0 1.18E-08 16.02 4.34 0.3 17.2 24.5 0.76 32.3 1.9

AH49-1-1 125.0 187.0 2.22E-08 12.87 1.30 0.1 13.3 43.5 0.77 56.6 3.4 41.6 9.0
AH49-1-2 128.0 139.0 3.38E-09 8.47 0.46 0.1 8.7 26.1 0.76 34.5 2.1
AH49-1-3 126.0 164.0 1.57E-08 17.14 6.15 0.4 18.7 26.1 0.76 34.5 2.1
AH49-1-4 121.0 128.0 5.93E-09 9.61 0.66 0.1 9.9 30.3 0.74 40.9 2.5

+ 238 232
eU= U(ppm)+ 0.235 Th(ppm)
* Corrected for alpha ejection after Farley et al (1996)
** Analytical error for single grain analyses are 6% (2σ) and representative of the reproducibility of laboratory standards.
Samples beginning AH… were provided by Calgary MSc student A. Halverson, supervised by B. Guest

136 | P a g e
Table 3.2-3: Zircon (U-Th)/Helium replicate analyses and mean ages

Corr. Analytical Average


width length Raw Age StDev
Sample 4He (mol) U (ppm) Th (ppm) Th/U eU+ Ft Age* Error** sample
(µm) (µm) (Ma) (Ma)
(Ma) (±2σ) Age (Ma)

11-AS-11-2 123.3 315.7 8.23E-13 25.3 12.6 0.50 28.3 30.6 0.841 36.4 0.3 44.9 4.6
11-AS-11-3 133.0 238.8 2.79E-11 860.2 49.9 0.06 871.9 37.7 0.836 45.1 0.5
11-AS-11-4 95.1 322.0 2.52E-11 1154.3 79.7 0.07 1173.0 36.8 0.806 45.7 0.5
11-AS-11-5 88.2 390.1 2.57E-11 1107.7 64.3 0.06 1122.8 37.7 0.798 47.2 0.5
11-AS-11-6 101.0 279.5 3.11E-11 1321.6 77.1 0.06 1339.8 40.5 0.810 50.0 0.5

11-AS-12-3 120.0 183.5 2.00E-13 11.4 4.9 0.43 12.6 31.3 0.805 38.9 0.5 40.8 4.6
11-AS-12-4 106.3 302.0 2.50E-12 84.8 42.3 0.50 94.7 38.7 0.820 47.2 0.4
11-AS-12-5 81.8 254.1 1.99E-12 179.0 122.7 0.69 207.8 28.1 0.773 36.4 0.3

12-AS-09-1 78.2 173.1 2.85E-12 284.7 161.1 0.57 322.5 41.7 0.746 55.8 0.5 52.7 3.0
12-AS-09-2 75.2 226.5 7.03E-12 620.5 215.8 0.35 671.2 40.7 0.753 54.1 0.5
12-AS-09-3 88.8 407.0 1.20E-11 428.8 256.0 0.60 489.0 38.2 0.799 47.8 0.4
12-AS-09-4 63.1 190.6 2.43E-12 361.9 265.6 0.73 424.4 37.7 0.708 53.2 0.5

12-AS-14-1 55.2 181.8 1.61E-12 576.4 319.0 0.55 651.4 22.3 0.675 33.0 0.3 41.5 5.2
12-AS-14-2 74.5 153.1 3.76E-12 616.4 121.2 0.20 644.8 34.2 0.728 47.0 0.4
12-AS-14-3 62.4 162.4 6.50E-12 1688.1 125.3 0.07 1717.6 29.9 0.699 42.7 0.4
12-AS-14-4 43.1 194.4 1.21E-12 593.8 197.9 0.33 640.3 26.3 0.610 43.2 0.4

12-AS-16-1 84.2 369.6 1.74E-11 373.1 76.4 0.20 391.1 84.1 0.788 106.8 1 63.9 27.9
12-AS-16-2 89.6 347.2 3.64E-12 175.0 107.5 0.61 200.3 32.6 0.799 40.7 0.3
12-AS-16-4 67.7 219.9 1.29E-12 210.5 100.0 0.48 234.0 27.5 0.731 37.6 0.3
12-AS-16-5 61.1 399.9 9.98E-12 573.8 302.6 0.53 645.0 51.6 0.732 70.5 0.6

137 | P a g e
Corr. Analytical Average
width length Raw Age StDev
Sample 4He (mol) U (ppm) Th (ppm) Th/U eU+ Ft Age* Error** sample
(µm) (µm) (Ma) (Ma)
(Ma) (±2σ) Age (Ma)
12-AS-17-1 70.1 301.9 7.74E-12 795.2 547.6 0.69 923.8 28.1 0.760 37.0 0.3 34.2 4.2
12-AS-17-2 96.5 242.3 3.60E-12 302.9 79.2 0.26 321.6 24.8 0.798 31.0 0.3
12-AS-17-3 113.8 302.4 1.79E-11 651.5 187.6 0.29 695.6 32.8 0.830 39.5 0.3
12-AS-17-4 84.9 228.4 2.51E-12 315.8 87.0 0.28 336.3 22.7 0.775 29.2 0.3

12-AS-19-1 113.8 320.0 7.61E-12 354.5 115.9 0.33 381.7 24.0 0.831 28.9 0.3 27.9 4.5
12-AS-19-2 132.9 375.9 1.49E-11 428.3 97.0 0.23 451.1 24.9 0.854 29.1 0.3
12-AS-19-3 127.4 348.8 1.53E-11 725.4 171.1 0.24 765.7 17.6 0.848 20.7 0.2
12-AS-19-4 89.0 237.7 9.93E-12 972.0 200.0 0.21 1019.0 25.8 0.785 32.9 0.3

PCA-304-83-3 47.5 160.3 1.23E-11 4997.4 311.8 0.06 5070.7 33.3 0.629 53.0 0.5 51.3 1.7
PCA-304-83-5 91.3 267.3 3.69E-11 2016.2 329.5 0.16 2093.6 39.3 0.793 49.6 0.5

zAH47-1-1 60.0 198.0 2.79E-09 4018.0 39.9 0.01 4027.21 38.12 0.718 53.1 4.2 39.3 13.9
zAH47-1-2 163.0 243.0 1.50E-08 966.5 13623.8 14.10 4110.73 21.82 0.859 25.4 2.0

zAH48-1-1 90.0 250.0 3.92E-09 1926.4 41.6 0.02 1935.96 39.08 0.801 48.8 3.9 47.2 1.6
zAH48-1-2 167.0 256.0 5.41E-09 743.5 12.8 0.02 746.41 39.96 0.876 45.6 3.7

zAH49-1-1 131.0 188.0 2.33E-10 100.2 81.4 0.81 119.00 23.91 0.833 28.7 2.3 50.6 22.6
zAH49-1-2 113.0 259.0 5.33E-10 150.3 144.4 0.96 183.50 34.11 0.826 41.3 3.3
zAH49-1-3 135.0 291.0 1.67E-09 162.2 62.0 0.38 176.52 69.86 0.854 81.8 6.5

zAH49-2-1 144.0 201.0 2.07E-10 65.0 26.9 0.41 71.17 27.53 0.849 32.4 2.6 43.5 10.5
zAH49-2-2 115.0 178.0 2.56E-10 85.1 20.3 0.24 89.76 47.25 0.820 57.6 4.6
zAH49-2-3 98.0 199.0 2.37E-10 141.6 39.4 0.28 150.68 32.45 0.802 40.5 3.2

+ 238 232
eU= U(ppm)+ 0.235 Th(ppm) ** Analytical error for single grain analyses are 6% (2σ), representative of reproducibility of laboratory standards.
* Corrected for alpha ejection after Farley et al (1996) Average Fish Canyon Tuff standard age for the analyses was 30.9 ± 2.5Ma
Samples beginning zAH… were provided by Calgary MSc student A. Halverson, supervised by B. Guest

138 | P a g e
Table 3.2-4: Bedrock apatite FT results

No. of Dpar Mean Length Std Dev (1σ, Central Age Error (1σ,
Sample ρs* Ns ρi* Ni ρd* Nd Ρ(χ2), %
Grains (µm) (µm)** µm) (Ma) Ma)

11-AS-12 0.416 292 5.794 4012 1.628 9025 21 90.3 1.57 12.5 (100) 0.2 18.3 1.1
12-AS-05 2.799 1182 11.860 4837 1.564 8673 9 0.0 1.66 12.6 (113) 0.1 61.2 6.5
12-AS-06 0.205 253 1.774 2254 1.564 8673 19 1.9 2.05 11.9 (101) 0.3 47.3 2.4
12-AS-08 0.265 484 1.923 3608 1.564 8673 21 47.3 7.33 - - 32.4 1.6
12-AS-09 0.452 394 5.122 4489 1.564 8673 22 15.4 1.55 - - 21.2 1.2
12-AS-10 0.384 367 3.503 3075 1.564 8673 20 2.8 1.73 - - 28.9 2.2
12-AS-11 0.414 569 3.384 4734 1.564 8673 20 18.2 1.73 - - 29.3 1.5
12-AS-13 0.366 283 3.048 2433 1.564 8673 20 23.6 1.44 - - 28.1 1.8
12-AS-14 0.228 312 2.087 2897 1.564 8673 20 97.1 1.73 13.3 (45) 0.3 26.1 1.6
12-AS-15 0.360 223 2.657 1870 1.564 8673 21 0.0 1.82 - - 34.5 4.0
12-AS-16 0.435 404 3.516 3153 1.564 8673 23 0.7 1.53 - - 30.9 2.3
12-AS-19 0.203 276 4.495 6089 1.564 8673 20 21.8 1.74 - - 11.0 0.7
12-RP-AS-04 0.181 358 1.318 2589 1.564 8673 35 0.0 1.81 - - 33.8 4.0
12-RP-AS-07 0.492 389 4.032 3337 1.564 8673 20 18.5 1.76 - - 28.5 1.8
12-RP-AS-08 0.218 405 1.549 2843 1.564 8673 25 0.2 1.8 - - 34.0 2.7
JC-60C-94 1.680 543 1.680 4374 1.680 9316 20 1.4 2.08 13.1 (51) 0.2 33.3 2.1
PCA-200-83 0.547 131 3.265 2044 1.680 9316 20 3.9 1.48 - - 16.5 1.9
PCA-307-83 0.063 110 0.301 550 1.680 9316 34 0.1 2.03 - - 52.7 7.6

*Track Densities are 106 tracks cm-2 ** Number of lengths measured in brackets (#)
Abbreviations used are: ρs: spontaneous track density; ρi: induced track density on mica detector; ρd: dosimeter standard glass track density; Ns: number of spontaneous
tracks counted; Ni: number of induced tracks counted; Nd: number of tracks counted on dosimeter standard glass
P(χ2) is the probability for obtaining χ2 value for n degrees of freedom, where n = (no. of grains - 1)
The Central Age is an averaged age, weighted for the precision of individual crystals (after Galbraith 1992)
Analyst: Annika Szameitat, ζ=310±13

139 | P a g e
Table 3.2-5: Bedrock apatite FT results of unpublished data, provided by R. Parrish

Mean
No. of Std Dev Pooled Age Error
Sample ρs* Ns ρi* Ni Nd Ρ(χ2), % r(si), % Length
Grains (1σ, µm) (Ma) (1σ, Ma)
(µm)**

80-RP-R1b 0.139 294 0.278 587 2691 50 - 3.2 - - 24.5 2.5


80-RP-R2 0.187 237 0.460 175 2691 30 - 11.4 - - 20.1 2.9
80-RP-R3 0.058 120 0.108 228 2691 49 - 7.2 - - 26.4 3.4
80-RP-R6 0.329 417 0.424 537 2691 30 - 9.0 - - 39.7 5.0
80-RP-R9 0.068 143 0.164 346 2691 50 - 7.1 - - 21.0 2.5
80-RP-K3 0.041 86 0.080 168 2691 50 - 9.8 - - 24.9 3.5
PCA-529-83 0.633 480 4.400 3332 8100 20 15.2 - 12.73 (106) 1.6 26.3 1.6
PCA-58-84 0.481 432 2.900 2607 8100 25 23.9 - 13.34 (66) 1.59 30.3 1.9
PCA-121-84 9.560 717 3.930 2948 8100 20 20.1 - 13.89 (133) 1.74 44.4 2.4

*Track Densities are 106 tracks cm-2


** Number of lengths measured in brackets (no)
Abbreviations used are: ρs: spontaneous track density; ρi: induced track density on mica detector; ρd: dosimeter standard glass track density; Ns: number of spontaneous
tracks counted; Ni: number of induced tracks counted; Nd: number of tracks counted on dosimeter standard glass
P(χ2) is the probability for obtaining χ2 value for n degrees of freedom, where n = (no. of grains - 1)
r(si) is the correlation coefficient to the relative standard error of induced tracks.
Pooled Age is and average age, calculated based on the ratio of spontaneous and induced tracks, use of this was discontinued in the 1980s when the Central Age was
introduced
Analyst samples RP-80-…: R. Parrish, zeta: 106±3.3

140 | P a g e
Table 3.2-6 : Modern river sediment apatite FT results of unpublished data

No. of Ρ(χ2), Central Age ±


Sample ρs* Ns ρi* Ni ρd* Nd Age populations ± 1σ error (Ma) and (no. of grains)
Grains % 1σ error (Ma)

GF 31 0.481 665 2.068 3263 1.291 5369 62 0 49.0 ± 5.6 28.9±1.7(32) 71.3±11.9(5) 201.4±28.9(2)
GF 35 0.411 210 3.298 1585 1.291 5369 15 0 30.1 ± 4.3 16.5±3.5(5) 28.6±3.2 (10)
GF 37 0.501 425 3.902 3153 1.291 5369 23 0 28.0 ± 2.4 22.3±1.1(19) 44.9±4.8(4)
GF 39 0.425 1754 2.695 11201 1.291 5369 62 0.0 33.7 ± 1.3 17.4±3.1(7) 35.0±1.0(53) 46.5±4.0(2)
GF 61 0.387 563 2.357 5369 1.291 5369 45 1.5 35.0 ± 1.8 Single population
GF 63 0.254 90 1.826 673 1.291 5369 5 18.8 29.7 ± 3.9 Single population
GF 66 0.339 585 2.469 4311 1.291 5369 33 1.6 30.0 ± 1.8 Single population
GF 67 0.620 898 3.278 5265 1.291 5369 31 0.0 37.7 ± 2.2 29.3±1.5(21) 49.2±3.5 (10)

*Track Densities are 106 tracks cm-2


Abbreviations used are: ρs: spontaneous track density; ρi: induced track density on mica detector; ρd: dosimeter standard glass track density; Ns: number of
spontaneous tracks counted; Ni: number of induced tracks counted; Nd: number of tracks counted on dosimeter standard glass
P(χ2) is the probability for obtaining χ2 value for n degrees of freedom, where n = (no. of grains - 1)
The Central Age is an averaged age, weighted for the precision of individual crystals (after Galbraith 1992)
Analyst: A.Carter, samples collected by G. Foster and R. Parrish

141 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

3.3 Discussion and Conclusions

The new dataset presented in this chapter has added valuable information to
the recent low-T thermal history of the Columbia Mountains. As described in chapter
1, little low temperature thermochronology data exists to date in published literature.
AFT studies previously conducted focussed on smaller study areas and smaller scale
problems. In addition to expanding the existing AFT data by interconnecting these
isolated study areas, this chapter contains a comprehensive and larger area covering
AHe and ZHe dataset, adding valuable information regarding the recent and lowest
temperature cooling history of these mountains.

The data presented clearly show the overall presence of a mainly Oligocene to
upper Eocene cooling phase at medium to low elevations. It can be observed at a first
instance on the data summary map in Figure 3.3-1. The bedrock apatite fission track
ages shown range from 11 Ma to 61 Ma with most samples falling between 25-35 Ma,
while Apatite (U-Th)/Helium ages range from 10.1 Ma to 48 Ma with most samples
falling between 20-37 Ma. For the higher temperature thermochronometer zircon (U-
Th)/Helium ages are mostly Eocene, ranging from 27-74 Ma. When compared to
previously published cooling ages, which were described in detail in chapter 1, it is
evident that the ages presented here are younger overall. Most previous data was
acquired using higher temperature methods such as Ar-Ar dating, and thus correctly
showed older, mostly early Eocene ages. However, the previously available AFT data
also showed predominantly Eocene ages older than 35 Ma. This can be attributed to
these studies focussing on the frontal thrusts in the eastern Rocky Mountains and thus
on areas with greater distance to the trench (Osadetz et al. 2004) or the higher
elevations of the Eocene Shuswap complex (Lorencak et al. 2001). The latter study also
presented some lower elevation data in the vicinity of Revelstoke (near vertical profile
D and the Selkirk-Shuswap transect presented in this chapter)

142 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

Figure 3.3-1: Summary map of all AHe, AFT and ZHe ages presented in this chapter. Due to the large scale of the
map vertical profiles B, D and G are shown in inset boxes. All data are presented regardless of sample elevation,
black markers indicate bedrock sample location while white markers show modern detrital sampling location.
(Background image: Google Earth, 2015)

with AFT ages ranging from 27 to 39 Ma, which is in agreement with the data found in
this study. Curiously, their study also found slightly older AFT ages (40-44 Ma) a few
kilometres to the west in the vicinity of Victor Lake, where our vertical profile A
showed early Eocene to Palaeocene ages as well. Lorencak et al. attributed this
variation of ages at similar elevations to potential small scale movement along existing

143 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

faults, causing blocks of slightly different ages to be placed next to each other.
However, in our opinion it is more likely that small scale variations in thermal patterns
caused by paleo-topography and the processes of erosion which led to the
exhumation of the present topography are responsible for the observed variations,
particularly since there is no evidence for recent fault activity. A more detailed
examination of this topic will be subject of chapter 5 and thus won’t be discussed any
further at this point. For now, we would like to focus on the overall cooling history that
could be identified.

In order to examine the presented dataset for patterns and trends, several
graphics were created which will be discussed in the remaining part of this chapter. In
general, as can be expected ages increase with increasing elevation for all three
methods used, which can be demonstrated by plotting age vs. elevation for all data
obtained (Figure 3.3-2 A, top row). It also appears that ages increase with increasing
distance to the Rocky Mountain Trench. In order to confirm this observation, all data
were plotted against the perpendicular distance (in degrees longitude) of the sampling
location to the Rocky Mountain Trench (Figure 3.3-2 B, middle row). The correlations
in the AFT and ZHe are clear, however the AHe dataset contains too much noise for a
clear observation. Because of the strong control elevation exerts on the observed
data, all ages were consequently normalized to an elevation of 700 m a.s.l. based on
the observed age-elevation relationship (Figure 3.3-2 B, bottom row). While the
correlation of the AHe dataset is now clearer (most ages plot in a cluster caused by a
similar distance to the RMT), many AFT ages now fall into the negative age field, which
is of course, impossible. The most likely cause for this is overcorrection of samples
based on an underestimated and generalized age-elevation-relationship. Nevertheless
the overall younging trend is still observed.

For the purpose of further investigating patterns of exhumation and resolving


the observed west to east younging trend spatially from north to south as well, maps
of various ‘elevation-slices’ were created. Each map (Figure 3.3-3 to Figure 3.3-5)
depicts the ages of samples with elevations within the slices and our interpretation of
the resulting age contours.

144 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

AHe AFT ZHe

2.5 4 2.5
A: Age vs. Elevation

y = 0.0351x + 0.0081
2.0 3 2
Elevation (km)

1.5 1.5
2
1.0 1
y = 0.0436x 1
0.5 0.5 y = 0.0276x
+ 0.1083
- 0.0686
0.0 0 0
0 20 40 60 0 20 40 60 80 20 40 60 80
B: Age vs. Distance to

Distance to RMT (degrees)

4 4 1.5
y = 0.0085x + 0.6993 y = 0.0757x - 0.6385
3 3
1
2 2
RMT

1 1 y = 0.0202x - 0.0222
0.5
0 0
-1 -1 0
0 20 40 60 0 20 40 60 0 40 80 120
C: Age normalised to 700m

Distance to RMT (degrees)

4 1.5
vs. Distance to RMT

y = 0.0114x 3
3 + 0.8877
2 1
2
1 1
0.5 y = 0.002x +
0 0 y = -0.0273x + 0.8761
2.2918
-1 -1 0
0 20 40 60 -60 -20 20 60 100 -40 0 40 80 120

Age (Ma) Age (Ma) Age (Ma)

Figure 3.3-2: (top row) Age elevation relationships for AHe, AFT and ZHe data of all samples west of the RMT
(excluding anomalous vertical profile E for ZHe and sample PCA-307-83 for AFT). (middle row) All ages plotted
against the distance in longitude perpendicular to the RMT in degrees. (bottom row) All ages normalised to an
elevation of 700m a.s.l. based on the age elevation relationship obtained in top row, plotted against the
perpendicular distance of sample location to the RMT.

First, all ages of samples with elevations below 800 m a.s.l. are shown in Figure
3.3-3. Because most of the data obtained for this study comes from low to medium
elevations, the coverage of this map is particularly good and the AHe and AFT age
contours could be drawn with confidence. Both thermochronometers (with closure
temperatures of ~70°C and ~120°C respectively) show a similar pattern: In addition to

145 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

increasing in age with increasing distance to the RMT there appears to be a centre of
youngest ages where the RMT intersects the 52° latitude, the northern end of the
Selkirk Mountains. This area near the centre of Kinbasket Lake, which extends for 200
km inside the RMT, has been mentioned previously as not only the locality of vertical
profile F but also the area where the youngest ages were found at the northern end of
the Mica creek transect. The age difference between the AHe and AFT data near this
contour focal point is approximately 5 Ma, with the youngest AHe ages younger than
15 Ma and the AFT ages below 20 Ma. At more distal locations to the south (e.g.
Valhalla complex, southern Selkirk Mountains) the difference between the AHe and
AFT contours appears to increase to approximately 10 Ma. The ZHe contours (closure
temperature of 180°C) in this figure appear to show an increase in age towards the
southeast and thus in the opposite direction to the AHe and AFT contours, but they
are based on only 3 ages and therefore less well constrained.

Similar patterns to the one that was just described can be observed in the next
two elevation slices (Figure 3.3-4). For samples ranging in elevation from 800-1200 m
a.sl. (depicted on the left) the AHe and AFT ages increase with increasing distance to
the RMT from < 20 Ma to > 40 Ma. While the AHe data show a similar concentric
pattern centred around 51-52° N, the pattern is less clear in the AFT data due to a lack
of data coverage. For samples ranging from 1200-1800 m a.s.l. (shown on the right)
good coverage of the AHe data leads to the same previously made observation of age
contours increasing away from the northern Selkirk Mountains (30-50 Ma). The little
ZHe data only allowed for one tentative contour to be drawn (40 Ma parallel to the
RMT with older ages in the west) and the AFT ages were too varied to make any
reasonable assumptions on contours.

146 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

Figure 3.3-3: AHe, AFT and ZHe ages of samples with elevations below 800 m a.s.l. and resulting age contours.
Both the AHe (blue) and AFT (pink) data show a similar pattern, overall increasing in age from northeast to
southwest with a slightly concentric pattern around the northern Selkirk Mountains (51-52°N, west of the RMT).
The ZHe data (green) appear to show a slight trend of increasing age from west to east, however this is based on
only 3 samples and thus not well constrained. (Background image: Google Earth, 2015)

147 | P a g e
Figure 3.3-4: (left) AHe, AFT and ZHe ages of samples with elevations ranging from 800-1200 m a.s.l. and resulting interpreted age contours. (right) AHe, AFT and ZHe ages of samples with
elevations ranging from 1200-1800 m a.s.l. and resulting age contours. For both plots ages increase from east to west, with a concentric pattern emerging in the AHe contours centered
around the northern Selkirk Mountains (51-52°N). (Background image: Google Earth, 2015)

148 | P a g e
Figure 3.3-5: (left) AHe and AFT ages of samples with elevations ranging from 1800-2000 m a.s.l. and resulting age contours. For both thermochronometers ages increase from northeast
to southwest. (right) ZHe ages of samples with elevations higher than 1800 m a.s.l. and the resulting age contours. Ages also increase from northeast to southwest. AHe and AFT ages
higher than 2000 m plotted for comparison. (Background image: Google Earth, 2015)

149 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

For the high elevation samples (Figure 3.3-5) the AHe and AFT ages with
elevations ranging from 1800 – 2000 m (shown on left) result in only slightly curved,
almost trench-parallel age contours with ages increasing away from the trench. The
interpreted AHe contours range from 25-30 Ma and the AFT contours shows ages
increasing from < 20 Ma to > 40 Ma from east to west. Samples with elevations higher
than 2000 m were excluded, because as mentioned for vertical profiles B and D, above
2000m a.s.l. AHe ages increase strongly with little elevation (indicating the beginning
of a partially exhumed PRZ) which means that even samples in close vicinity can have
widely varied ages, making finding contours difficult. Furthermore the only AFT age
>2000 m was excluded because with an elevation of 3100 m a.s.l. and an age of 52.7
Ma it had a similar effect when compared to lower elevation samples, throwing off the
observed pattern. These ages are shown on the second plot (right side), which also
contains the ZHe data of all ages with elevations higher than 1800 m (total range of
ZHe sample elevations is 1843-2347 m a.s.l.). Similar to the lower elevation data the
ZHe results show contours increasing in age from 50-60 Ma with increasing distance to
the northern Selkirk Mountains, however the data density is low and thus these
interpretations are unfortunately not well constrained.

To summarize the most important observations made from these contour plots
we can say it appears that for lower elevations and lower T(°C) thermochronometers
contours show a distinct pattern of increasing age away from the northern Selkirk
Mountains and the RMT. This pattern weakens for higher elevation samples of low
T(°C) chronometers and is not present in the high T(°C) ZHe data, with all plots
showing near RMT parallel contours of ages increasing with distance to the RMT.

In addition to looking at the spatial distribution of ages, the thermal history


models produced by HeFTy provide insight into the temporal distribution of cooling.
The last summary plot, Figure 3.3-6, shows all thermal history model results presented
in this chapter, separated into two columns showing different cooling events. Samples
of generally lower elevations are shown on the left (with JC-60C-94 at 2220 m being
the exception) and samples of higher elevations (except for the lowest one, sample 11-

150 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

AS-12 at 780 m) on the right. Both columns are arranged by elevation, increasing from
bottom to top on each side.

Figure 3.3-6: Overview of thermal history models presented in this chapter. The left column shows that low to
medium elevation samples (440m to 942m) and one higher elevation sample (top left, 2220m) display a distinct
upper Eocene/Oligocene cooling event (indicated by grey shading), cooling through the AFT PAZ and with the
exception of 12-AS-06 also the AHe closure of 70°C during this period. Higher elevation samples in the right
column (780m to 2300m) show cooling through similar temperatures sooner, during the upper Cretaceous to
lower Eocene. For both columns samples at lower elevations also show a sharp increase in cooling over the last
10Ma while the higher elevation samples in each column display gradual cooling to surface temperatures.

151 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

The first trend in the medium to low elevations (left column) is a period of
rapid (>10°C/m.y.) cooling between 35-28 Ma (upper Eocene – mid Oligocene), which
is present in all four locations as indicated by the grey shaded overlay. The generally
higher elevation samples (right column) do not show this cooling phase but rather
indicate medium rate cooling (3-5°C/m.y.) for lower elevation samples and rapid
cooling (>10°C/Ma) for higher elevations during the Cretaceous and Palaeocene (75-45
Ma). For both columns the samples at lower elevations which were collected in
present day valleys show a further recent phase of extremely rapid cooling (>10°C/
m.y.) over the last 10 m.y. (also indicated by grey shading in Figure 3.3-6).

In order to take these observations a little further we can attempt some first
order approximations of erosion by estimating the local geotherm and thus calculating
burial depth at the time recorded by the presented ages. As described in chapter 1 the
Rocky Mountain Trench marks the transition of thick old cratonic continental crust in
the east to a much thinner, accreted crust in the west. Being located on this ‘mobile
belt’ (Hyndman, Currie, and Mazzotti 2005) surface heat flow west of the RMT has
been averaged at 75mW/m2 and thus a steady state crustal geotherm of
approximately 32°C increase in temperature per kilometre depth (for the upper few
km of the crust (Ehlers 2005)) can be calculated for the present time. Using this
relationship, we can generally conclude that those areas which now present AHe and
AFT ages from 35 - 28 Ma at the surface have experienced between 2 km and 3.5 km
of exhumation respectively since that time. For the higher temperature ZHe data the
same geotherm would equal approximately 5.5 km of erosion since crossing the 180°C
isotherm.

Applying this geotherm to the AHe data recorded in the vicinity of the RMT also
leads to another interesting observation: The samples acquired within or at low
elevations near the RMT (700 - 900 m) show Oligocene AHe ages and thus indicate
that 30 m.y. ago the current surface buried beneath 2 – 2.5 km of rock. However, not
only do nearby higher elevation samples (>1400) show similar ages, the surrounding
peaks also stand almost 3 km taller than the current valley bottom of the RMT. The
most reasonable explanation for these observations is that the RMT is likely an old

152 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

feature which was already in existence as a major range dividing valley during the
upper Eocene and lower Oligocene, because otherwise the rocks at low elevations
would have experience up to 1km more erosion and thus, even with a moderate
geothermal gradient would show much younger low T°C ages. This concurs with the
presence of Oligocene-Miocene sedimentary deposits in the southern RMT which
were previously used to propose a proto-RMT which existed since the beginning of
Cenozoic time (Clague 1974).

Using the above estimated burial depths we can also give a first approximation
of erosion rates which caused the exhumation of the modern surface. Using the
assumption of 2-2.5km depth for AHe and 3-3.5km depth for AFT, the cooling rates of
the vertical profiles lead to erosion rates of 0.1-0.5mm/yr for the rapid cooling period
between 35 and 28 Ma. These rates are both reasonable and comparable to modern
active orogens, such as the Alps. Erosion then would have slowed down to rates
<1mm/a, and only picked up again to an estimated 1-3 mm/yr during the past 10 m.y.
for the areas showing younger ages and later cooling.

These first estimates, however are based on three basic assumptions that do
not necessarily hold: (1) The closure temperature isotherms are parallel to the surface,
(2) The crystals have travelled along a vertical path and thus perpendicular to the
isotherms when being exhumed by erosion and uplift (i.e. no lateral rock movement
such as thrusting or faulting occurred), (3) that changes in heat flow and exhumation
rates have not affected the geothermal gradient and that it has thus been steady over
the past.

Large valleys present in the study area, such as the RMT with a maximum width
of ~ 16 km, indicate that the wavelength of the topography is large enough (>10 km) to
cause some degree of bending of particularly the low temperature isotherms (Braun
2005) and thus assumption (1) is already unlikely to be true. Resulting from this
assumption (3) is unlikely to hold true for low temperatures and thus particularly the
AHe data as well since even slight changes in exhumation rate would have had an
influence on the geothermal gradients. Assumption (3) can also be debated for ages

153 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

older than 40 Ma, as the rising of the core complexes would most certainly have had
an impact on the thermal regime and thus the heat flow of the area. These factors will
be discussed in great detail and explored using forward modelling of the vertical
profile data with numerous exhumation histories and geothermal gradients in chapter
5 before final conclusions can be drawn, and thus will not be discussed further at this
point.

Therefore, in order to briefly summarize and conclude the results presented in


this chapter we can state:

 There are 3 cooling phases that could be identified in different


elevation ranges, indicating that elevation is an important control on
observed ages:
o High elevations rapidly cooled below ZHe and AFT closure
temperatures (~180° and ~120°C) during the Cretaceous and
Palaeocene (75-45 Ma). Subsequent slow cooling rates through
the AHe Tc (70°C) and temperatures below indicate gradual
(<5°C/ m.y.) exhumation to the modern surface temperature.
o Medium elevations of the modern day topography show rapid
cooling of approximately 10°C/ m.y. below ~180° and ~120°C at
the end of the Eocene to the beginning of the Oligocene (45-30
Ma), followed by a period of much slower (<1°C/ m.y.) cooling
between 30 and 10 Ma.
o Low elevations also share in part a slow cooling phase in the
middle Cenozoic but then record a late phase of extremely rapid
exhumation (>10°C/ m.y.) from ~70-60°C to surface
temperatures over the past 10 Ma.
 Patterns observed on contour plots indicate that AHe, AFT and ZHe ages
generally increase with increasing distance to the RMT and that
particularly lower elevations additionally show a centre of youngest
ages in the northern Selkirk Mountains.

154 | P a g e
Thermochronology of the Columbia Mountains Chapter 3

 Lastly, the RMT could be an ancient valley which was already present
during the Oligocene, which is consistent with Oligocene sedimentary
deposits in the southernmost RMT (Clague 1974).

155 | P a g e
Thermochronology of the Southern Rockies Chapter 4

Chapter 4:
Thermochronology Results of the
Southern Canadian Rocky Mountains

4.1 Introduction

The southern Canadian Rocky Mountains are an eastward trending fold and
thrust belt which was formed on the Canadian continental edge as a result of the
collision of various terranes with this craton during the Cretaceous. The Rocky
Mountain Trench (RMT) which separates them from the Columbia Mountains to the
west (Figure 4.1-1) roughly marks the position of the edge of the Canadian craton and
thus the boundary to a more complicated and varied geology, which was discussed in
the previous chapters (see e.g. Chapter 1, Figure 1.2-2). In the east, the geology
comprises of eastwardly stacked, duplex faulted sediments from upper Cretaceous age
in the foothills to Cambrian and Devonian and in the north even upper Proterozoic
deposits near the Rocky Mountain Trench. Lithologically these stacks comprise of
passive continental margin sediments such as quartzites, shales and carbonates and
have – unlike the geology in the west – not undergone metamorphism above low
grade greenschist facies, with many rocks recording only diagenesis.

Because of these more clastic rock types, sampling for thermochronology


analysis using apatite and zircon is more challenging, as intrusive or highly
metamorphic rocks which are generally rich in these target minerals are absent.
Although both apatite and zircon can generally withstand sedimentary transport quite
well, the dilution occurring during transport and deposition makes it challenging to
find beds which contain enough target minerals for analysis. In this case, clastic
sediments such as the occasional quartzites are the only rocks that are suitable. For
example, of the 6 bedrock samples collected west of the RMT for this study only 2

156 | P a g e
Thermochronology of the Southern Rockies Chapter 4

yielded enough apatite for analysis. Luckily previously unpublished data of modern
detrital sediments provided by Gavin Foster and further bedrock data provided by my
Co-Author Bernard Guest from the University of Calgary could be obtained to
complete this dataset presented here. They are both thanked kindly for their
contribution.

Figure 4.1-1 Physiographic map of the south eastern Canadian Cordillera (after (Mathews 1986), showing the
main ranges. The study area of this chapter includes all ranges east of the Rocky Mountain Trench (yellow dashed
line). (Background image: Google Earth, 2015)

4.2 Results

The data results for the Southern Canadian Rocky Mountains are presented in
two sections below: The northern part which contains five bedrock samples forming

157 | P a g e
Thermochronology of the Southern Rockies Chapter 4

an east to west horizontal transect across the range and eight further bedrock samples
along a vertical transect near Jasper, as well as three modern detrital sediment
samples. The southern half of the range unfortunately only yielded two modern
detrital sediment data points and two bedrock samples.

4.2.1 The North

Figure 4.2-1: Map displaying the thermochronology results for the northern part of the Southern Canadian Rocky
Mountains. Apatite (U-Th)/He data (AHe) is shown in blue while Apatite Fission Track data (AFT) is presented in
pink. White markers indicate locations of modern detrital sediment samples and yellow shaded areas show the
extent of the corresponding catchment areas. Black markers indicate the locations of bedrock samples.
(Background image: Google Earth, 2015)

Of the three modern detrital sediment samples obtained for the northern part
of these Mountains the previously discussed GF 37 (Which was also included in
Chapter 3) is the only one draining towards the west and thus the RMT (Figure 4.2-1).
The sample was collected from the Holmes river with a catchment ranging from 706 –
2880 m elevation covering an area of approximately 690 km2 and yielded an Oligocene

158 | P a g e
Thermochronology of the Southern Rockies Chapter 4

central AFT age of 28 ± 2.4 Ma with two populations centred around 22.3 ± 1.1 and
44.9 ± 4.8 Ma. The two remaining detrital samples GF 52A and GF 49 were both
collected from the eastward draining Athabasca River, with sample GF 52A being
located further upstream and GF 49 more distal near the front of the range. This
implies that the catchment area of GF 52A (more densely shaded in Figure 4.2-1) is
included in the larger catchment of sample GF 49. Both samples yielded Eocene
central ages and contained two populations, a major Eocene group and minor
Cretaceous contributions. GF 52A with a catchment area of almost 5800km 2 and
elevations ranging from 1004 – 3700m a.s.l. produced a central AFT age of 47.8 ± 3.9
Ma, with 28 crystals contributing to the Eocene population around 38.6 ± 2.7 Ma and 4
crystals producing a second upper Cretaceous population around 70.5 ± 15.8 Ma. The
more distal sample GF 49 (catchment area 9624km2 and elevation range 969 – 3700 m
a.s.l.) yielded a central AFT age of 46.5 ± 7.4 Ma with 15 crystals contributing to the
Eocene population of 44.6 ± 4.5 Ma and one crystal yielding a lower Cretaceous age of
110.9 ± 18 Ma.

The five bedrock samples 11-AS-22, 11-AS-26, J-91-2, J-91-1 and J-91-18 which
make up the East-West horizontal transect at approximately 53° North are indicated
by black markers in Figure 4.2-1. All were dated by apatite (U-Th)/He dating and the
resulting ages ranged from 18.0 ± 1.5 Ma to 97.7 ± 30.1 Ma. Overall the ages appear to
be younger towards the RMT, which is confirmed when plotted (Figure 4.2-2) along
the elevation profile indicated Figure 4.2-1. The topography increases from NE to SW
and the sample elevation (black markers for bedrock samples, white markers for
modern detrital sampling locations) although similar, slightly decreases in the same
direction. AHe ages are shown by the blue markers with 2σ errors while the pink
markers indicate an observed modern detrital AFT central age with 1σ error. For both
datasets correlation trend lines (blue and pink linear respectively) show a younging
trend of the observed ages towards the SW and thus the RMT.

159 | P a g e
Thermochronology of the Southern Rockies Chapter 4

Figure 4.2-2: Elevation profile (as shown in Figure 4.2-1) from SW to NE, showing the increase in topographic
elevation towards the RMT (SW) with an overall decrease in bedrock AHe (blue) and modern detrital AFT (pink)
ages (trendlines for both chronometers shown in blue and pink respectively). Error bars shown are 2σ for AHe
and 1σ for AFT. Location of Jasper vertical profile is indicated by pink box.

4.2.1.1 Jasper Vertical Profile

In the centre of the horizontal transect just described lies the only available
vertical profile this side of the RMT (Figure 4.2-1). The transect lies approximately 2 km
south-west of Jasper and stretches for just over 3 km up the northern flank of Whitney
Peak, covering elevations from 1297m – 2545m a.s.l. (Figure 4.2-3, top left). All eight
samples were dated using the (U-Th)/He method on both apatite and zircon, yielding
Oligocene to upper Eocene AHe and lowest Eocene to Palaeocene ZHe ages
throughout. Apatite (U-Th)/He ages increase with elevation from 20.7 ± 2.7 Ma to 45.1
± 6.5 Ma, resulting in the steep age elevation relationship displayed in Figure 4.2-3
(blue markers, top right). A similarly constrained age elevation relationship can be
observed for the obtained zircon (U-Th)/He data (green markers) with ages ranging
from 53.3 ± 9.1 Ma to 63.3 ± 4.1 Ma.

160 | P a g e
Thermochronology of the Southern Rockies Chapter 4

2500 y = 49.237x
+ 120.33
AHe
2000

Elevation (m)
ZHe

1500
y = 107.74x
- 4203.6
1000
0 10 20 30 40 50 60 70 80
Age (Ma)

Figure 4.2-3: (Top left) Sample locations and elevation profile line of the Jasper vertical profile on the northern
flank of Whistlers Peak. The white line indicates the trace of the elevation profile; (bottom) Elevation profile with
apatite (U-Th)/He and zircon (U-Th)/He results; (top right) Age elevation plot resulting from this data including
trendlines (age-elevation relationships) for both AHe and ZHe data. The data of this profile (unpublished) was
provided by B. Guest, University of Calgary. (Background image: Google Earth, 2015)

The thermal history models of the lowest and highest elevation samples of the
vertical profile are depicted in Figure 4.2-4. Although no AFT track length distributions
for samples east of the RMT could be obtained (which would help narrow the
potential model solutions considerably), the samples along the Jasper profile with
both AHe and ZHe data available make it possible to create a somewhat useful thermal
history. An additional constraint is given by the nearby modern detrital AFT age, which
despite by nature being sampled from elevations above the sampling location, at least
represents a minimum reset temperature and age. Meaning, if the rocks at higher

161 | P a g e
Thermochronology of the Southern Rockies Chapter 4

elevations experienced temperatures higher than 120°C sometime between 80 and 20


Ma (approximate range of observed AFT ages), the lower sample elevations must have
experienced the same if not hotter temperatures. The last thermal constraint used
was the initial deposition time during the Cambrian (10 ± 10°C at 480-550 Ma, off
scale), when these sedimentary rocks were deposited. Unlike the rocks discussed in
the previous chapter the strata east of the RMT has not experienced considerable
metamorphism and thus there are no higher temperature chronometer ages available
which could help narrow the history further.

Figure 4.2-4: HeFTy thermal history inverse model results for top sample NMI37-1 (left) and bottom sample
NMI39-8 (right) of the Jasper vertical profile. Constraints imposed on the inverse models are the original
deposition age ranging from 480-550 Ma at 10 ± 10°C (box constraint used during modelling, not depicted on
scale), the AFT age range of nearby detrital AFT sample GF 52A with ages ranging from 20-80Ma at temperatures
higher than 120°C (black box) and the present day average surface temperature of 10 ± 10°C. Green envelope
marks the range of acceptable paths found (50% confidence level, so 50% of observed data (single grain ages, not
displayed) could have been produced by these paths) and magenta the range of good paths (95% confidence
level). The model was run until 100 good paths were found. The thick black line indicates the best fit cooling path
for the dataset.

Both results still allow for a large range of thermal histories to be permitted,
especially for the burial period between the Cambrian and the upper Cretaceous (500-
80Ma) but also during the last 60 m.y. where exhumation took place, which is due to
the unavailability of track lengths and further constraints during the Mesozoic.
However, the results show that there is one crucial difference between samples
NMI37-1 which was sampled at the top of the profile at 2454 m a.s.l. and the bottom

162 | P a g e
Thermochronology of the Southern Rockies Chapter 4

sample NMI39-8 at 1297 m a.s.l. While the top sample shows relatively rapid cooling
(9°C/Ma) between 60 and 40 Ma all the way to surface temperatures where it
remained since, the best fit (black line) for the lower elevation sample shows gradual
and constant cooling (3°C/Ma) since 60 Ma to the present.

4.2.2 The South

Figure 4.2-5: Map displaying the thermochronology results for the southern part of the Southern Canadian Rocky
Mountains. Apatite (U-Th)/He data (AHe) is shown in blue while Apatite Fission Track data (AFT) is presented in
pink. White markers indicate locations of modern detrital sediment samples and yellow shaded areas show the
extent of the corresponding catchment areas. Black marker indicates the locations of a bedrock sample.
(Background image: Google Earth, 2015)

Of the two modern detrital samples available in the southern part of the range
GF 61 was already presented in chapter 3. This sample was taken from the upper
Columbia River near Donald and thus drains the ranges to the west as well as the east
of the RMT, with the sampled catchment area spreading over almost 11800km2 and
covering elevations from 791m up to 3200m a.s.l. (Figure 4.2-5). The obtained central AFT

163 | P a g e
Thermochronology of the Southern Rockies Chapter 4

age was uppermost Eocene with 35.0 ± 1.8 Ma (single population). The second modern
detrital sample GF 55 is more centrally located in the Rocky Mountains and was
sampled from the Bow River in Lake Louise, draining a much smaller catchment area of
885km2 with elevations ranging from 1563m – 3220m a.s.l.. The central AFT age for
this sample, which also presented with a single population was determined to be 48.0
± 6.7 Ma and thus upper Eocene. Just like these detrital samples the two available
bedrock sample for this area, 12-AS-01 and Ice River, also yielded Eocene apatite (U-
Th)/He ages of 54.1 ± 1.4 Ma (at 1649 m a.s.l.) and 38.5 ± 3.8 Ma (at 1430 m a.s.l.).

4.3 Data Discussion

The individual aliquot data for all (U-Th)/He results in this chapter are
presented in Table 4.3-1 (Bedrock AHe data) and Table 4.3-2 (bedrock ZHe data). For
both thermochronometers most of the individual ages agree fairly well, resulting in a
reliable mean age and a reproducibility error (standard deviation) of generally less
than 15%. However, as previously discussed for the data in chapter 3, some samples
show considerably more dispersion between the individual aliquots. For the AHe data
(Table 4.3-1) samples 11-AS-22 (50% dispersion), J91-1 (31%), NMI37-4 (26%) and NMI
37-3 (27%) and for ZHe data samples NMI37-3 (19%) and NMI 39-7 (17%) show age
dispersions larger than expected for normal reproducibility variations in (U-Th)/He
data.

As before, we can assume that inclusions and implantation are not the cause,
as the careful picking procedure should eliminate the possibility of the first while the
latter was deemed to be extremely rare and unlikely to influence datasets on a
significant scale (Brown et al. 2013). However, as discussed in chapter 2 the largest
known influence on aliquot age dispersion is the content of radiation effective
uranium and thorium eU (eU= 238U(ppm)+ 0.235 232Th(ppm)). A strongly positive
correlation of age vs. eU indicates that grains with higher eU content have most likely

164 | P a g e
Thermochronology of the Southern Rockies Chapter 4

experienced more radiation damage and thus increased the retention of Helium and
increased the effective closure temperature, causing the age to be older than lower eU
crystals (as shown in chapter 2, section 2.3.8.1). As before, all individual ages were plotted
against the measured eU content and positive correlations are shown in Figure 4.3-1 for
both apatite (left) and zircon (right). (For clarity only positive correlations are shown here,
all other eU correlations are included in Appendix 3.1).

Interestingly, the only over dispersed AHe sample showing a positive correlation is
J91-1, while none of the above mentioned over dispersed ZHe samples produced positive
correlations at all. So while we can assume that for J91-1 the large spread in eU is the
most likely cause for the wide range of ages, we have to look for alternative reasons for
the other samples. Because sample 11-AS-22 only yielded two usable apatites, a large
difference between the two observed ages is not surprising and can explain part of the
large dispersion of 50%. The other samples for both AHe and ZHe all include a minimum of
4 crystals and are thus better constrained. In these cases the age dispersion could be
attributed to increased diffusion with extreme radiation damage (potentially even causing
a negative eU correlation) or zonation of radiation damage following zonation of U/Th
within the crystal. As described in both chapter 2 and 3 these effects are still barely
understood, but have been theorized to be responsible for over dispersion in (U-Th)/He
datasets (Farley 2000, Nasdala et al. 2004, Ault and Flowers 2012).

The individual count data of the modern detrital fission track ages presented in
this chapter can be found in Table 4.3-3. As before the ages presented are central ages
(weighted by precision, i.e. the total number of track counts) and where these are
based on more than one age population it is indicated in the table (all radial plots can
be found in Appendix 3.2). Aside from sample GF55, which with only 7 crystals has an
limited sample size, all samples passed the statistical χ2 test for expected natural
variability within datasets

165 | P a g e
Thermochronology of the Southern Rockies Chapter 4

AHe positive eU ZHe positive eU


correlations correlations
160.0 70.0
12-AS-1a
140.0 65.0
J91-1
120.0 NMI38-05 60.0
NMI37-02
100.0 55.0
NMI37-01
Age (Ma)

Age (Ma)
80.0 50.0

60.0 45.0

40.0 40.0
zNMI37-01
20.0 35.0 zNMI39-06
0.0 30.0
0.00 50.00 100.00 0.0 200.0 400.0 600.0
eU (ppm) eU (ppm)
Figure 4.3-1: Positive correlations of effective Uranium (eU) vs. age for AHe (left) and ZHe (right) data. Strongly
positive correlations (such as J91-1 for Ahe and zNMI37-01 for ZHe) indicate that the spread in single ages is most
likely attributed to radiation damage and thus He-retention. All other eU plots for samples showing no or
negative correlations can be found in the Appendix.

166 | P a g e
Table 4.3-1: Apatite (U-Th)/He Data of Bedrock Samples

Analytical Averaged Standard


width length U Th Raw Age Corr. Age*
Sample 4He (mol) Th/U eU+ Ft Error** sample Deviation
(µm) (µm) (ppm) (ppm) (Ma) (Ma)
(±2σ) Age (Ma) (Ma)

11-AS-22-2 114.3 119.4 5.65E-15 1.03 4.54 4.4 2.1 37.4 0.74 50.5 1.1 33.6 16.9
11-AS-22-3 129.2 187.9 1.21E-14 2.26 2.46 1.1 2.8 13.2 0.79 16.6 1.6

11-AS-26-1 222.8 342.4 5.19E-14 1.63 1.74 1.1 2.0 18.3 0.88 19.1 1.1 18.0 1.5
11-AS-26-3 91.0 104.7 7.54E-15 8.72 12.73 1.5 11.7 14.7 0.68 17.0 1.0

12-AS-1a-2 115 168 1.69E-13 12.15 10.77 0.89 14.68 40.86 0.77 53.3 0.5 54.0 1.4
12-AS-1a-3 67 93 4.08E-14 18.51 39.04 2.11 27.69 31.35 0.60 52.7 0.6
12-AS-1a-4 86 183 2.96E-13 31.86 41.53 1.30 41.62 39.86 0.71 55.9 0.5

Ice River-1 90.0 130.0 1.96E-12 99.58 1529.70 15.4 459.1 29.8 0.70 42.5 0.5 38.5 3.8
Ice River-2 79.6 139.7 1.88E-12 114.07 2452.39 21.5 690.4 22.6 0.68 33.4 0.4
Ice River-3 95.9 137.8 2.53E-12 102.52 1753.83 17.1 514.7 28.5 0.72 39.6 0.4

J91-1-1 75 78 1.09E-11 22.26 18.03 0.81 26.50 79.52 0.58 137.0 8.2 97.7 30.8
J91-1-2 63 78 3.22E-12 9.35 11.72 1.25 12.10 70.05 0.53 132.1 7.9
J91-1-3 82 124 3.84E-12 6.56 4.40 0.67 7.60 51.99 0.65 80.5 4.8
J91-1-5 65 86 1.79E-12 9.25 11.73 1.27 12.00 33.79 0.55 61.7 3.7
J91-1-6 73 110 4.22E-12 10.53 10.79 1.02 13.06 46.43 0.60 77.3 4.6

J91-2-1 65 91 5.60E-12 87.52 86.37 0.99 107.82 12.30 0.56 22.1 1.3 25.1 5.4
J91-2-2 47 106 2.65E-12 48.33 87.62 1.81 68.92 14.83 0.45 32.6 2.0
J91-2-3 86 157 2.65E-12 10.78 14.49 1.34 14.18 13.64 0.66 20.5 1.2

J91-18-1 - - 4.23E-12 3.79 6.55 1.73 5.32 31.49 0.73 43.0 2.6 41.4 5.4
J91-18-2 - - 1.00E-11 5.75 8.70 1.51 7.79 28.16 0.78 36.2 2.2
J91-18-3 - - 5.02E-12 4.14 6.77 1.63 5.73 33.32 0.73 45.9 2.8
J91-18-4 - - 9.52E-12 4.25 7.26 1.71 5.95 36.33 0.76 47.9 2.9
J91-18-5 - - 2.97E-12 4.75 7.80 1.64 6.59 23.59 0.69 34.1 2.0

167 | P a g e
NMI39-08-1 69 121 9.05E-13 11.15 1.21 0.11 11.50 12.67 0.61 20.7 1.2 20.7 1.2

NMI39-07-1 68 123 3.24E-12 26.89 14.46 0.54 30.69 16.33 0.60 27.1 1.6 30.9 3.3
NMI39-07-2 67 124 3.82E-12 31.87 24.41 0.77 37.75 16.65 0.59 28.2 1.7
NMI39-07-3 72 127 4.17E-12 20.00 1.93 0.10 20.59 21.10 0.66 31.9 1.9
NMI39-07-4 81 132 1.69E-11 86.79 76.68 0.88 104.94 22.31 0.61 36.7 2.2
NMI39-07-5 60 117 1.10E-11 108.28 107.67 0.99 133.99 17.16 0.56 30.9 1.9

NMI39-06-1 95 137 9.76E-12 19.69 77.12 3.92 37.65 18.85 0.67 28.1 1.7 34.2 6.9
NMI39-06-2 81 128 3.38E-12 18.79 2.74 0.15 19.66 18.74 0.66 28.5 1.7
NMI39-06-4 82 158 3.61E-11 92.15 37.15 0.40 101.25 30.17 0.67 45.2 2.7
NMI39-06-5 68 111 3.34E-12 23.99 19.51 0.81 28.67 20.54 0.59 35.0 2.1

NMI38-05-1 83 107 3.16E-12 15.41 7.78 0.51 17.32 22.29 0.64 34.7 2.1 37.7 8.5
NMI38-05-2 62 145 7.53E-12 22.43 76.46 3.41 40.64 29.41 0.56 52.2 3.1
NMI38-05-3 81 131 4.24E-12 14.42 26.79 1.86 20.92 20.81 0.64 32.6 2.0
NMI38-05-4 92 120 6.09E-12 20.10 23.83 1.19 25.66 20.92 0.67 31.4 1.9

NMI37-04-1 73 180 1.51E-11 56.86 13.68 0.24 60.43 23.58 0.65 36.3 2.2 39.1 10.1
NMI37-04-3 95 153 2.01E-12 2.27 3.37 1.49 3.29 38.51 0.69 56.2 3.4
NMI37-04-4 63 169 2.51E-12 15.25 4.98 0.33 16.49 20.25 0.60 33.6 2.0
NMI37-04-5 75 128 1.30E-11 90.61 26.84 0.30 97.60 16.73 0.64 26.3 1.6
NMI37-04-6 74 188 1.02E-10 276.28 217.44 0.79 327.15 27.55 0.64 42.9 2.6

NMI37-03-1 92 102 5.00E-12 18.34 10.02 0.55 20.73 25.20 0.66 38.1 2.3 41.4 11.4
NMI37-03-2 98 156 1.44E-11 29.17 7.27 0.25 31.09 27.22 0.71 38.2 2.3
NMI37-03-3 85 111 7.31E-12 19.85 64.89 3.27 35.61 22.34 0.63 35.3 2.1
NMI37-03-4 87 124 2.58E-12 4.63 2.45 0.53 5.65 41.53 0.66 62.5 3.7
NMI37-03-5 96 103 1.19E-11 29.81 19.21 0.64 34.45 32.38 0.67 48.3 2.9
NMI37-03-6 80 126 5.28E-12 33.05 6.38 0.19 34.78 17.07 0.65 26.2 1.6

NMI37-02-1 66 145 1.68E-12 8.58 10.76 1.26 11.24 20.86 0.59 35.1 2.1 40.2 6.4
NMI37-02-2 88 171 2.51E-12 5.81 9.40 1.62 8.47 19.60 0.67 29.1 1.7
NMI37-02-3 80 146 2.69E-11 65.09 79.42 1.22 83.95 30.91 0.64 47.9 2.9
NMI37-02-4 79 145 1.23E-11 18.85 100.39 5.33 42.52 28.05 0.63 44.5 2.7
NMI37-02-5 93 153 6.08E-12 13.42 3.75 0.28 14.73 27.94 0.70 40.1 2.4

168 | P a g e
NMI37-02-6 110 131 4.45E-12 6.41 5.33 0.83 7.83 31.82 0.71 44.6 2.7

NMI37-01-2 72 161 1.04E-11 30.10 24.24 0.81 36.07 31.13 0.63 49.7 3.0 45.1 4.6
NMI37-01-4 124 208 4.82E-12 3.91 1.56 0.40 4.35 31.02 0.77 40.5 2.4

+ 238 232
eU= U(ppm)+ 0.235 Th(ppm)
* Corrected for alpha ejection after Farley et al (1996)
** Analytical error for single grain analyses are 6% (2σ) and representative of the reproducibility of laboratory standards.
Samples 12-AS-1a analyzed by A.Szameitat, Samples J-91-… and NMI… provided by B.Guest, collected and analyzed by B.Guest/D.Stockli and Naomi Miles
respectively

169 | P a g e
Table 4.3-2: Zircon (U-Th)/He Data of Bedrock Samples

Analytical Average
width length U Th Raw Age Corr. Age* StDev
Sample 4He (mol) Th/U eU+ Ft Error** sample
(µm) (µm) (ppm) (ppm) (Ma) (Ma) (Ma)
(±2σ) Age (Ma)

zNMI38-05-1 73 191 4.54E-10 350.56 126.79 0.36 379.9 46.5 0.75 62.2 5.0 56.8 6.2
zNMI38-05-2 82 138 2.51E-10 277.89 91.07 0.33 298.9 35.8 0.76 47.3 3.8
zNMI38-05-3 75 243 8.56E-10 441.20 682.25 1.55 598.3 41.8 0.75 55.6 4.5
zNMI38-05-4 106 212 1.58E-10 38.68 55.47 1.43 51.5 50.5 0.81 62.6 5.0
zNMI38-05-5 152 260 1.06E-09 104.25 172.21 1.65 143.9 48.2 0.86 56.2 4.5

zNMI37-01-1 121 314 2.22E-09 330.46 191.23 0.58 374.5 51.0 0.84 60.7 4.9 63.3 4.1
zNMI37-01-3 103 288 1.92E-09 413.33 148.47 0.36 447.6 55.7 0.82 68.0 5.4
zNMI37-01-5 120 377 2.86E-09 354.23 206.60 0.58 401.9 51.7 0.84 61.3 4.9

zNMI37-03-1 109 302 4.75E-10 70.95 108.88 1.53 96.0 54.7 0.82 66.6 5.3 56.5 10.8
zNMI37-03-2 89 204 2.89E-10 118.91 131.58 1.11 149.2 47.5 0.78 61.0 4.9
zNMI37-03-4 74 173 4.23E-10 326.98 382.27 1.17 415.0 42.2 0.74 57.1 4.6
zNMI37-03-5 93 160 6.13E-10 422.60 524.05 1.24 543.3 32.2 0.78 41.4 3.3

zNMI37-02-2 77 220 8.03E-10 416.66 410.57 0.99 511.2 46.9 0.76 61.8 4.9 58.0 2.3
zNMI37-02-3 91 217 1.02E-09 408.95 356.74 0.87 491.2 45.8 0.79 58.3 4.7
zNMI37-02-4 93 228 7.82E-10 285.54 262.38 0.92 346.0 45.2 0.79 57.1 4.6
zNMI37-02-5 75 187 7.47E-10 511.34 725.43 1.42 678.4 41.8 0.74 56.2 4.5
zNMI37-02-6 82 196 9.04E-10 504.55 566.00 1.12 634.9 43.1 0.76 56.4 4.5

zNMI39-07-1 104 198 6.09E-10 203.82 220.68 1.08 254.7 44.5 0.80 55.5 4.4 53.3 9.1
zNMI39-07-2 74 156 9.48E-10 796.08 1623.42 2.04 1170.0 37.3 0.73 51.0 4.1
zNMI39-07-3 80 209 6.49E-10 424.84 913.73 2.15 635.2 30.0 0.76 39.5 3.2
zNMI39-07-4 101 204 1.36E-09 367.89 555.36 1.51 495.8 51.4 0.80 64.3 5.1
zNMI39-07-6 91 147 6.41E-10 369.72 499.84 1.35 484.8 43.2 0.77 56.3 4.5

zNMI39-06-1 90 203 3.23E-10 143.52 239.96 1.67 198.8 39.0 0.78 50.1 4.0 54.8 8.4

170 | P a g e
zNMI39-06-2 85 259 8.72E-10 312.14 395.10 1.27 403.1 45.8 0.78 58.8 4.7
zNMI39-06-4 106 229 3.85E-10 96.58 274.75 2.84 159.8 37.0 0.81 45.9 3.7
zNMI39-06-5 115 204 5.97E-10 116.92 211.76 1.81 165.7 52.6 0.82 64.4 5.2

zNMI37-04-1 88 148 4.31E-10 245.57 315.90 1.29 318.3 46.3 0.76 60.6 4.8 55.8 4.6
zNMI37-04-2 73 152 9.74E-11 72.43 193.16 2.67 116.9 40.3 0.73 55.4 4.4
zNMI37-04-3 83 194 3.88E-10 125.49 655.46 5.22 276.4 40.9 0.76 53.9 4.3
zNMI37-04-4 85 189 7.40E-10 435.43 627.63 1.44 580.0 36.9 0.77 48.1 3.8
zNMI37-04-5 96 191 6.33E-10 222.45 328.91 1.48 298.2 47.5 0.79 60.3 4.8
zNMI37-04-6 78 175 1.12E-09 764.85 992.45 1.30 993.5 42.2 0.75 56.4 4.5

zNMI39-08-3 91 154 1.04E-09 619.67 445.14 0.72 722.3 44.1 0.78 56.9 4.6 58.7 4.5
zNMI39-08-4 98 180 7.68E-10 318.28 354.57 1.11 400.0 43.1 0.79 54.4 4.4
zNMI39-08-5 88 205 5.98E-10 265.50 256.77 0.97 324.6 45.6 0.78 58.5 4.7
zNMI39-08-6 92 185 5.55E-10 230.96 190.34 0.82 274.8 50.8 0.78 64.9 5.2

+
eU= 238U(ppm)+ 0.235 232Th(ppm)
* Corrected for alpha ejection after Farley et al
(1996)
** Analytical error for single grain analyses are 6% (2σ) and representative of the reproducibility of laboratory
standards.
Data in this table provided by B.Guest, analyzed by Naomi Miles

171 | P a g e
Table 4.3-3: Modern Detrital Apatite Fission Track Data (unpublished results by A. Carter and G. Foster)

No. of Ρ(χ2), Central Age ± Age populations ± 1σ error (Ma) and Elevation range
Sample ρs* Ns ρi* Ni ρd* Nd
Grains % 1σ error (Ma) (no. of grains) (m a.s.l.)

GF 37 0.501 425 3.902 3153 1.291 5369 23 0 28.0 ± 2.4 22.3±1.1(19) 44.9±4.8(4) 706 m - 2800 m
GF 49 0.406 219 1.481 893 1.291 5369 16 0.0 46.5 ± 7.4 44.6±4.5(15) 110.9±18.0(1) 969 m - 3700 m
GF 52A 0.174 355 8.141 1620 1.291 5369 32 2.6 47.8 ± 3.9 38.6±2.7(28) 70.5±15.8(4) 1004 m - 3700 m
GF 55 0.590 113 2.629 525 1.291 5369 7 7.3 48.0 ± 6.7 Single population 1563 m - 3220 m
GF 61 0.387 563 2.357 5369 1.291 5369 45 1.5 35.0 ± 1.8 Single population 791 m - 3200 m

*Track Densities are 106 tracks cm-2

Abbreviations used are: ρs: spontaneous track density; ρi: induced track density on mica detector; ρd: dosimeter standard glass track density; Ns: number of
spontaneous tracks counted; Ni: number of induced tracks counted; Nd: number of tracks counted on dosimeter standard glass
P(χ2) is the probability for obtaining χ2 value for n degrees of freedom, where n = (no. of grains - 1)
The Central Age is an averaged age, weighted for the precision of individual crystals (after Galbraith 1992)
Analyst: A.Carter, samples collected by G. Foster and R. Parrish

172 | P a g e
Thermochronology of the Southern Rockies Chapter 4

4.4 Discussion, Summary and Conclusions

When compared to the previously presented mainly upper Eocene and lower
Oligocene ages west of the RMT (chapter 3) the ages to the east presented here are
very similar. Overall they do appear to be slightly older; however the lack of data
coverage horizontally across the map as well as vertically, makes definite conclusions
difficult. Most ages observed in the sections above are early Eocene, Palaeocene with
some older Cretaceous results. Only few AHe samples at elevations 1500m a.s.l. and
below show lower Oligocene ages, which were a distinct age group in the previous
chapter. If true, this could indicate that the ranges on either side of the trench have
experienced different cooling histories and thus slightly different phases of
exhumation. The fact that the RMT roughly marks the edge of the thick continental
craton in the east and the thinner, accreted crust in the west could thus lead to many
theories regarding the underlying cause for this disparity (to be addressed in chapters
5 and 6).

However, there are several contributing factors which could be skewing the
observed age distribution in this chapter towards older ages and cause the observed
age difference effect. Firstly, most sample elevations this side of the trench are higher
in altitude (average bedrock sample elevation 1634.7m) when compared to the data
presented in chapter 3 (average elevation: 1151.5m). In this case, assuming the same
age-elevation relationship (and thus no difference in exhumation history) applies,
older ages would be expected, as age should by definition increase with elevation as a
result of higher altitudes having passed through the partial annealing/retention zone
first. Furthermore, the topography to the east of the RMT is generally sharper, with
more glaciated areas, steeper valleys and higher peaks. This steeper gradient leads to
more active erosion of higher elevations and could cause the modern detrital ages to
appear older as well, without an actual difference in cooling history.

The lack of data coverage this side of the RMT compared to the previous
chapter makes conclusions regarding spatial trends and the creation of contour plots

173 | P a g e
Thermochronology of the Southern Rockies Chapter 4

as well as the observation of an averaged age elevation relationship difficult. The only
horizontal transect available (5 bedrock samples in Figure 4.2-1 and Figure 4.2-2)
shows a wide range in ages from 25.1 to 97.7 Ma with a slight overall younging trend
towards the RMT. This is in concordance with the observations made in the previous
chapter where the youngest ages were also found near this central range dividing
feature.

Furthermore, the young AHe ages of samples 11-AS-26 (18.0 ± 1.5 Ma at 741 m
a.s.l.) and 11-AS-22 (33.6 ± 16.9 Ma at 791 m a.s.l.) inside the RMT near the base of
Mount Robson, with 3954 m a.s.l. the highest mountain in the Rockies, also supports
the previously mentioned possibility that the Rocky Mountain Trench is an ancient
feature which has been in existence since at least the upper Eocene. If the same initial
estimated geothermal gradient of 32°C/km depth is applied, the resulting Miocene
and uppermost Eocene burial depth for these samples would be approximately 2km.
Even if Mt Robson had experienced only minor erosion over the past 33Ma (which,
based on the paleoclimate and known past and present glaciations (refer to chapter 1)
is unlikely) it would have towered more than 1km over the ‘Palaeo RMT’ at the time
the samples cooled below the AHe Tc of 70°C. However, the complicated tectonic
position of these samples with the RMT marking the transition of the thin lithosphere
of the ‘mobile belt’ (Hyndman, Currie, and Mazzotti 2005) to the thicker ancient craton
with a significantly reduced estimated heat flow of 40 mW/m2 (compared to the
previous 75 mW/m2 ) means that this might still be an underestimated burial depth.
The reduced heat flow would result in a lower geothermal gradient of approximately
25°C/km (Ehlers 2005), suggesting a potential burial depth of almost 3km. The exact
nature of this transition and how rapid the geothermal gradient changes is barely
known and will be the subject of further investigation in chapter 5 in order to improve
the estimates of the removed rock volume. Because of the sample locations directly in
the RMT we can – for now – assume that a higher gradient still applies and our
observation holds true.

174 | P a g e
Thermochronology of the Southern Rockies Chapter 4

The vertical age-elevation relationships observed in this chapter are based on


only one vertical profile near Jasper (Figure 4.2-3). This profile however shows two
clear and well defined age elevation relationships: AHe ages increasing from lower
Oligocene (20 Ma) at low elevations to mid Eocene (45 Ma) at the top and ZHe ages
are predominantly Palaeocene with a similarly steep gradient and thus slight increase
in age with elevation (53-63 Ma). For the same profile the thermal history modelling
showed a rapid cooling through the ZHe, AFT and AHe closures (180-70°C) all the way
to surface near temperatures between 70 and 40 Ma (Palaeocene and Eocene), while
the lower elevation sample indicated mostly gradual thermal exhumation through the
entire temperature range since 70 Ma to the present day at approximately 3°C/Ma.

To summarize the conclusions of this chapter we can state:

 The main cooling phase observed to the east of the Rocky Mountain Trench
indicates that cooling below temperatures of 180°C (ZHe) and 120°C (AFT) and
for high elevations 70°C (AHe) occurred during the Paleocene to lower Eocene
(70-40 Ma).
 The Oligocene exhumation (mostly 40-20 Ma) previously observed in the west
is present in AHe data of elevations below 1500 m a.s.l., however for lack of
data coverage no distinct pattern can be observed.
 A tentative observation can be made that ages increase with increasing
distance to the Rocky Mountain Trench, similar to the observed trend west of
the RMT.
 The RMT is likely to be an ancient feature which has been in existence since at
least the Eocene.

175 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

Chapter 5:
Cenozoic Exhumation, Elevation and
Uplift of the South Eastern Canadian
Cordillera
5.1. Introduction

The previous two chapters presented new low –temperature


thermochronology data from the Columbia and Rocky Mountains of the south eastern
Canadian Cordillera. However, cooling ages alone are generally not enough to allow a
complete interpretation of the exhumation history of mountain ranges. Unless a
research question is directly related to emplacement of fast cooling plutonic bodies
(such as sills) or volcanic eruptions, observed cooling ages rarely relate to distinct
geological ‘events’. In most cases data produced by thermochronological dating is the
result of complex interactions of crustal thermal processes with the diffusion kinetics
of investigated minerals – many physical properties and feedbacks of which are still
not completely understood. Large scale processes such as erosion and faulting
influence the thermal state of the crust. The resulting thermal evolution of the crust
influences the depth which the cooling ages record. For these reasons a careful
consideration and, if possible, quantification of the processes which may have
influenced the crustal thermal history has to be made before the cooling ages can be
related to major geological causes.

This chapter will first introduce some considerations necessary to assess the
thermal state of the crust in the study area and then will explore the range of thermal
and physical exhumation histories which could have produced the age-elevation
relationships observed for the vertical profiles in the previous two chapters. This is
done by forward modelling numerous geologically possible cooling and exhumation

176 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

scenarios using different time-invariant geothermal gradients and compare predicted


vs. observed age elevation relationships. The best suited exhumation history results
and geothermal gradients will be used to produce exhumation estimates for the past
45 m.y. for the vertical profile locations as well as the larger study area and eventually
lead to estimates of Cenozoic erosion and relief. These provide a basis for conclusions
on paleo-elevation and potential surface uplift.

5.2. Considerations for Interpreting Thermochronology Data

First order interpretations of cooling age datasets tend to require several


assumptions: 1) that all samples have moved vertically to the surface, 2) that the
crustal isotherms are located at a constant depth below the mean surface elevation
and 3) that the geothermal gradient, and thus the depth which each cooling age
records, has not changed over time. When interpreting cooling age data from vertical
profiles, it is also often assumed that all samples have moved in unison with similar
exhumation rates and passed through the closure temperature isotherm in intervals
equal to their altitude difference divided by the exhumation rate. However, as already
discussed in previous chapters, vertical profiles are rarely truly ‘vertical’. Due to
sampling possibilities and natural topography most profiles have a considerable lateral
distance between sampling locations. This makes careful assessment of assumptions
1) and 2) increasingly important. The first section of this chapter will discuss how these
issues are addressed and which assumptions are made for the interpretations in this
chapter.

The assumption that there was no lateral component to the sample


exhumation can be evaluated based on the geology of the area, in particular using
evidence for any faulting. While thrust faulting and folding in this part of the Cordillera
ceased around 60 Ma, late fault movements along the Lewis thrust and extensional

177 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

faulting adjacent core complexes have been recorded as late as 45 Ma [Read and
Brown, 1981; Brown and Read, 1983; Carr et al., 1987; Osadetz et al., 2004]. The
absence of significant faulting within the last 45 m.y., the period of interest to this
study, suggests that assumption 1) is reasonable for most of our study area.

Figure 5.2-1: Illustration of closure depth assumptions made when interpreting thermochronology data from
‘vertical transects’. Scenario (a) shows the simplest approach where the closure depth is assumed constant in
relation to e.g. sea level and the samples will have cooled through this temperature in descending order by
elevation (highest first > oldest, lowest last > youngest), indicating that the slope of the observed age-elevation
relationship equals the exhumation rate. (b) Closure temperature is at constant depth in relation the surface and
samples are taken from a true vertical profile. The result will be the same as in a) as the highest sample cooled
below Tc first followed by lower elevations and again the slope of the observed age-elevation relationship
indicates the exhumation rate. Scenario (c) depicts the most realistic variation: while the closure temperature
does warp in response to topography, the depth is not equal below the surface or stable in relation e.g. sea level.
Closure depth below valleys (z1) tends to be shallower than below area of high topography (z2) and the degree of
this warping depends on the wavelength of the topography (to be discussed more in text). While the samples
form a realistic vertical profile (i.e. including some lateral distance, as in a) did cool through Tc at different times,
the depth to Tc is different at each location, meaning that the slope of the age-elevation relationship is not
necessarily equal to the exhumation rate as each sample has been exhumed from a different depth since cooling.

The assumption of constant isotherm depth variation is more problematic.


Figure 5.2-1 (redrawn after [Ehlers, 2005]) illustrates two simplified end-member
scenarios used when interpreting age elevation relationships. The horizontal isotherm
end-member depicted in (a) assumes that the closure depth is constant in relation to a
horizontal surface, and that the samples taken from a mountain flank passed through
this isotherm in order of elevation (highest first, lowest last). The second scenario (b)
assumes that the closure isotherm is at a constant depth below the non-horizontal
surface topography and samples from a truly vertical profile also passed through it in
order of elevation. In both cases the slope of the best fit line through the age-elevation
relationship is equal to the exhumation rate. However, while scenario (a) may be
applicable to higher temperature chronometers, the relevant isotherms to closure of
apatites are susceptible to topographic variations. Scenario (b) is only applicable to

178 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

samples from truly vertical profiles such as cliff faces or drill cores. The reality for most
studies dealing with age-elevation relationships, lies somewhere in between. Scenario
(c) shows how closure depths of low temperature chronometers may mirror the
surface topography to some degree, but that the horizontal advection of heat causes
the depth to be shallower below valleys (z1) and greater below peaks (z2). The degree
to which this draping occurs depends mainly on the depth of the isotherm (and thus
the geothermal gradient), the wavelength of the topography and the exhumation rate.
In order to evaluate exhumation histories in this study the effect of topography on
isotherms needs to be considered (to be discussed in section 5.1.2).

5.2.1. Calculation of Modern Geothermal Gradients from Surface Heat Flow

The eastern Cordillera in the past 45 m.y. is tectonically inactive and the main
factor determining the geothermal state of the crust is likely to be conductive heat
flow. Advective heat transport (which would be significant in active orogenic belts with
exhumation rates exceeding 1 km/ m.y.) can be neglected. In the case of modest
exhumation rates that are not expected to perturb the thermal gradient significantly, a
relatively simple 1-dimensional first order calculation of the thermal gradient is
sufficient [Ehlers, 2005]. Where the vertical or horizontal gradients are affected by
rapid erosion or high relief, more complicated 2D or 3D models need to be applied
[Ehlers and Farley, 2003]. Because low temperature thermochronometers typically
deal with processes within the upper 5-10 km of the crust it is not necessary to
consider processes at greater depth and a 1-D solution to the Poisson steady-state
heat transfer equation based on the measured surface temperature T 0 and upper
mantle boundary heat flow qm can give a good approximation for the near-surface
geothermal gradient [Parrish, 1983; after Lachenbruch, 1970]:

−𝑧
𝑧 𝐴0
(1) 𝑇(𝑧) = 𝑇0 + (𝑞𝑚 𝑘) + (𝐷2 ) ∗ (1 − 𝑒 ( 𝐷 ) )
𝑘

179 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

where T(z) is the temperature at depth z, k is the thermal conductivity, A0 is the


surface radiogenic heat production and D the characteristic E-folding depth.

Figure 5.2-2: Solutions to equation (1) for geothermal gradients from upper mantle heat flow values ranging from
2 2
35 mW/m to 100 mW/m , resulting in geothermal gradients of 20°C/km to 40°C/km. The figure also shows how
depths of commonly used closure temperatures (65°C for apatite (U-Th)/He, 120°C for apatite fission track and
180°C for zircon (U-Th)/He [Farley, 2000; Ketcham et al., 1999; Reiners, 2005] vary between different geothermal
gradients. T(0) was assumed to be 0°C, thermal conductivity k= 3.32 W/mK and surface radiogenic heat
3
production A(0)=3.6 µW/m (average values documented for the SE Cordillera, after [Hyndman and Lewis, 1999])
and E-folding depth D=10km (average value for continental crust, after [Ehlers, 2005]).

The solution to this equation for the upper 10km of the crust for surface heat
flow values ranging from 40-100mW/m2 and average crustal parameters of the south
eastern Cordillera (k= 3.2 W/m*K and A0=3.6 µW/m3 [Hyndman and Lewis, 1999] and
D=10km (crustal average) [Ehlers, 2005]) are shown in Figure 5.2-2. While a surface
heat flow of 40mW/m2 results in a low geothermal gradient (~12°C/km), for the same
crustal parameters a higher heat flow of 100mW/m2 results in a gradient of ~30 °C/km.
In this case the estimated depth of commonly cited closure temperatures (65°C for
AHe [Farley, 2000], 120°C for AFT [Ketcham et al., 1999] and 180 °C for ZHe [Reiners,

180 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

2005]) vary by several kilometres, as indicated by the intersection of the geotherms


with the closure isotherms (grey bars) (Figure 5.2-2).

Table 5.2-1: Estimated modern geothermal gradients for vertical profile locations

Surface heat Thermal Crustal Surface radiogenic Resulting


Profile Location flow conductivity thickness heat generation geothermal gradient
(mW/m2) (W/mK) (km) (µW/m3) (°C/km)**

A Victor Lake 86 3.32 34.5 3.60 35


B Pettipiece Pass 86 3.32 34.5 3.60 35
C Valhalla 99 - 126 4.45 35 - 40 5 - 7.01 34 - 37
D Revelstoke 86 3.32 35 3.60 35
E Canoe Mt 86 3.32 ~40 3.60 35
F Mt Trident 86 3.32 ~40 3.60 35
G Ptarmigan Mt 86 3.32 34 3.60 35

Jasper Jasper 42 2.55* 50 0.7* 18

Surface heat flow values, thermal conductivities and radiogenic heat production from Hyndman and Lewis (1999) and Currie and
Hyndman (2006). Were no detailed values in the vicinity of samples were available the average heat flow value for the Omineca
belt (86mW/m2) was used to estimate the other values based on correlations found by Hyndman and Lewis (1999). Crustal
thicknesses estimated after Zelt and White (1995) and Buryanik et al (1997b). * Where no values could be obtained in the literature
an average values for limestone with sandstone and shale interbedded were used, as suggested by Ehlers (2005) after Lee and
Deming (1999) and Haenel et al. (1988). ** Geothermal gradients calculated after Parrish (1983) and Lachenbruch (1970).

Using equations (1) and (2) and crustal parameters (Table 5.2-1) the current
geothermal gradients for the profile locations can be estimated. In the Columbia
Mountains they range from 35°C/km (majority of profiles) to 37°C/km (Valhalla
Complex). This is slightly higher than most modern measured crustal gradients (25-30
°C/km) [Turcotte and Schubert, 2002]. The low measured surface heat flow in the
Canadian Rockies east of the RMT of 50 - 60mW/m2 [Zelt and White, 1995; Burianyk et
al., 1997; Hyndman and Lewis, 1999] would result in a low modern geothermal
gradient of only 18°C/km for the profile located in Jasper. However the presence of
hot springs [Worthington, 1991; Grasby and Hutcheon, 2001] suggests that the
geothermal gradient might be considerably higher. The underestimation could be
caused by the low radiogenic heat production of the limestone dominated lithology at

181 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

the surface (A(0) of pure limestone is 0.62 µW/m3 [Haenel et al., 1988]) or by other
hydrological factors that obscure the background heat flux. The low gradient is thus
not considered representative. These calculated gradients were used as a starting
point to test numerous exhumation scenarios described in section 5.3; both higher and
lower gradients were also evaluated.

5.2.2. The Influence of Topography on Closure Depth: Draping Correction

The second factor stemming from assumption 2) (equal depth of closure


isotherms) at the beginning of this section which has to be considered is the ‘draping’
effect of the closure depth in response to surface topography. Although the
exhumation rate of the study area is not great enough to cause significant lateral heat
advection to warrant 2D or 3D modelling of isotherm surfaces, the effect of
topography on the closure temperature depth below the surface still has to be taken
into account. The degree of curvature of isotherms in response to topography is
dependent on the wavelength of surface topography and the depth beneath the
surface (Figure 5.2.3). Equations which can estimate appropriate closure depths in this
situation can be found in [Turcotte and Schubert, 1982; Brown, 1991; Stüwe et al.,
1994; Mancktelow and Grasemann, 1997; Ehlers and Farley, 2003].

This effect can be important for thermochronometers with the lowest closure
temperatures (e.g. AHe). Despite the low expected exhumation rates, the geothermal
gradients of 25-30 °C/km make it likely that the low-T isotherms will be perturbed in
response to the first order topographic wavelength of the study area [Stüwe et al.,
1994; Brandon et al., 1998], which ranges approximately from 7-15 km. Because most
of the vertical profiles sampled in this study were taken over a distance of 3-15 km,
the mean elevation within a radius of 10 km was used for calculation of closure
depths. A digital elevation model (Figure 5.2-4, a) was filtered to calculate the mean
elevation within a 10 km radius (Figure 5.2-4, b) of each locality (for the vertical
profiles one value was used at the mid-point of each transect).

182 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

Figure 5.2-3: (a) Illustration on how the depth of the closure isotherm z’(c) in relation to the local mean elevation
h’ can help calculate the total exhumation of any sample calculated as z’(c) + Δ(sample), where Δ(sample) is the
deviation of the sample elevation (e.g. h(a) and h(b)) from the local mean elevation h’, both in relation to sea
level. Total exhumation of sample a is therefore z’(c)+Δa and the total exhumation of sample b is z’(c)-Δb. This
approach is appropriate for low temperature thermochronometry in regions of short wavelength λ (~<10km),
high geothermal gradients and/or high exhumation rates [Stüwe et al., 1994]. (b) For studies concerning higher
temperature thermochronometers or areas of longer primary wavelength λ (~>40km) , lower geothermal
gradients or extremely low exhumation rates estimating exhumation depth z(c) from the surface topography
alone can suffice [Ehlers, 2005]. Figure redrawn after [Brandon et al., 1998], equations from [Stüwe et al., 1994].

a) b)

Figure 5.2-4: (left) DEM of the study area with high resolution topography ranging from close to sea level at
distant valley locations to almost 4000m in the central areas of both the Rocky and Columbia Mountains (scale
bars show elevation in m a.s.l.). (right) DEM of the study area filtered for mean elevation within 10 km radius at
any given location. Figures produced in GMT.

183 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

5.2.3. Changes in Geothermal Gradient with Time

The variation of geothermal gradient with time needs to be evaluated in order


to produce the best fit of predicted vs. measured cooling ages from the three different
thermochronometers. In the absence of active tectonics and extensive magmatism the
variation of exhumation rate over time is likely the main factor influencing the
geothermal gradient. Assuming no significant changes in rock type and thus A(0),
erosion will generally have the effect of causing an upward migration of isotherms,
relative to the surface, causing an increase in the geothermal gradient. In the case of
negative exhumation (i.e. burial through sedimentation) the geothermal state will be
lowered and the gradient depressed. This scenario may not always be the case
however, if radiogenic heat is highly concentrated in the uppermost crust with smaller
D-values.

Brandon et al. (1998) calculated a first order 1-D steady state simplification of
this response for a one dimensional column of thickness L with the constant surface
temperature Ts, assuming a constant exhumation rate έ at the top which equals the
accretion rate at the bottom (thus keeping L the same), using the thermal diffusivity κ
and an initial linear geothermal gradient g0:

1−𝑒 −έ 𝑧/𝜅
(3) 𝑇(𝑧, έ) = 𝑇𝑠 + 𝑔0 𝐿 1−𝑒 −έ 𝐿/𝜅

This equation (which is based on equation 8 in Stüwe et al. (1994)) gives the
temperature profile after re-reaching steady state with a constant exhumation rate. A
range of solutions of equation (3) for exhumation rates from 0.1 to 1 km/m.y. is shown
in Figure 5.2-5. For a thickness L = 45 km and a thermal diffusivity of 10 km2/m.y. the
resulting geothermal gradients vary from the initial 20 °C/km up to 87 °C/km.

184 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

Figure 5.2-5: Illustration of how long term exhumation rates of ε = 0.1 to 1.0 km/ m.y. will increase the
geothermal background state of the crust from an initial gradient of 20°C/km (where ε=0) to up to 86°C/km, for a
2
crustal thickness of 45km and thermal diffusivity of 10km / m.y. (after [Brandon et al., 1998]). Once again the
common closure temperatures of AHe (65°C), AFT (120°C) and ZHe (180°C) are indicated by the grey bars to show
how the closure depth varies with variation of geothermal gradient.

The time it takes to achieve this new steady state in response to constant
exhumation varies greatly, not only with exhumation rate and duration, but also with
depth. For exhumation rates lower than 1 km/m.y. significant changes (increase of e.g.
20%) in gradient typically occur after 104-107 years and thus could be relevant for the
time scale considered in this study [Ehlers, 2005].

5.3. Obtaining Exhumation Histories from Profile Data

The objective of this section is to determine exhumation (depth-time) histories


consistent with the cooling histories (temperature-time) derived from HeFTy
modelling of thermochronological data, for the various localities studied. It already is

185 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

anticipated, from variable cooling rates as a function of time, that exhumation


histories will be characterised by variable erosion rates over time.

5.3.1. Forward Modelling Vertical Profile Thermal Histories and


Exhumation Rates

The thermal histories modelled inversely with HeFTy for individual samples in
chapters 3 and 4 were stacked and combined with the observed age-elevation
relationships to find a range of possible cooling histories for each of the 7 vertical
profiles (Figure 5.3-1). Each history of a vertical profile in which the depth dimension is
known (albeit over a limited depth range) was forward modelled using HeFTy multiple
times using different geothermal gradients, with the age of one representative grain
from each sample predicted.

The profiles located west of the Rocky Mountain Trench (RMT) have 40Ar-39Ar
mica and hornblende cooling ages placing these rocks at 300-450°C at 45 to 90 Ma
(see Table 1.3.1. in chapter 1 for detail). These constraints were used as initial T-t
conditions for the modelled exhumation histories. For Jasper, which is located east of
the RMT and thus in the sedimentary and un-metamorphosed Rocky Mountains, an
initial condition for modelling is the minimum temperature of 200°C pre-60 Ma
consistent with observed early Palaeocene ZHe ages.

The predicted age-elevation relationship was then compared to the observed


data, and assessed empirically for the goodness of fit (GOF) of all three
thermochronometers, as well as consistency with the known geological history. All
models were deemed a good fit if the following conditions were met: (1) the predicted
ages were mostly within error of the observed ages for all three chronometers; (2) the
predicted AFT track length was within 10% of the measured mean track length (if
available); (3) the histories are consistent with the single sample thermal histories

186 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

previously modelled; and (4) the histories are consistent with the known geological
history (e.g. Ar-Ar ages).

Figure 5.3-1: Location of vertical profiles which were modelled in this chapter.

To illustrate this, three potential solutions are shown for Profile D (Mt
Revelstoke) in Figure 5.3-2. The first column shows the modelled cooling histories for
all samples of the profile and the resulting exhumation rates calculated with
geothermal gradients used to produce the predicted ages. The second column shows
the predicted (hollow markers) vs. observed (filled markers) age-elevation relationship
on a pseudo-elevation plot. This plot is created by adding the temperature difference
between the three thermochronometers divided by the geothermal gradient, to the
higher temperature sample elevations. For example, a geothermal gradient of
30°C/km and the difference between AHe and ZHe closure of 115 °C (based on
assumed standard closure temperatures of 65 °C and 180°C), a sample with elevation
1000m would plot at an elevation of 1km+(115°C/(30°C/km)) = 4.83km for the ZHe
age. This plot can both help assess the quality of fit for each chronometer separately
as the plots are much clearer to read, and portray the overall exhumation history.

The three scenarios for profile D show how subtle differences such as an earlier
onset of cooling or a geothermal gradient change of as little as 6°C can produce

187 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

Figure 5.3-2: Examples of good fitting forward modelled exhumation histories for vertical profile D. Left column
shows time-temperature paths and resulting exhumation rates using different geothermal gradients. Right
column shows the measured (filled markers) vs. predicted (empty markers) ages on age-pseudoelevation plots
(see text for detail on pseudoelevation). Error bars are 1σ (AFT) and 2σ (AHe and ZHe) average sample errors.

acceptable solutions for the same age-elevation plot. All three are similar and
provided good fits with small differences in t-T paths and onset of exhumation for
geothermal gradients varying from 20-26°C/km. While the more rapid exhumation
scenarios for the early Cenozoic with rates of >0.2 km/m.y. shown in A and C provide
better fits for the ZHe data, a more steady rate for cooling below 150 °C appears to
produce better fits for the AHe data. However, the t-T paths with a period of increased

188 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

erosion between 40 – 20 Ma (A and B) at rates of 0.59 to 0.38 km/m.y. come closer to


predicting the AFT data. By varying the geothermal gradient and the onset of initial
cooling from 350°C by as little as 5 m.y. good fits can be achieved for all 3 exhumation
histories. Notably, all presented histories show a later Micoene increase in exhumation
since 20-10 Ma with exhumation rates between 0.02 and 0.18 km/m.y. producing a
better fit for the AHe data than histories without this increase. Also the higher
geothermal gradient of 26°C/km of scenarios B and C produced better fits (for AHe and
ZHe) than lower or higher gradients. All three depicted scenarios are consistent with
the thermal history of the medium elevation sample (12-AS-14) presented in chapter
3. Based on an empirical assessment of the closest predicted mean AFT track length
(predicted 13.9µm, measured: 13.3µm) and the overall closest fit for all three
chronometers, scenario C, with a gradient of 26°C/km and gradual exhumation that
then increased in the later Micoene, is deemed the ‘best fit’ for profile D.

Exhumation histories for nearby localities of profiles A and B are shown in


Figure 5.3-3. Both profiles produced the best fitting predicted vs. observed age
relationships for the a geothermal gradient of 26°C/km; however a range of 20-
26°C/km for profile A and 26-30°C/km for profile B also produced good results. The
best history for profile A shows that the location had to have experienced rapid
exhumation of 0.77 km/m.y. until approximately 70 Ma, followed by slow exhumation
of 0.01 km/m.y. while the samples were being held at temperatures between 80-60°C
and lastly a more rapid phase moving the samples to the current surface with rates of
0.23-0.32 km/Ma. For profile B two end-members of exhumation histories are shown,
both for a geothermal gradient of 26°C/km, which produced reasonable fits. Finding
suitable histories in this case was complicated because three thermochronometers
were used and because the higher elevation AHe samples display much older ages
than the AFT or even ZHe data. Although such inverted ages occur occasionally in
thermochronological studies (including this one, as mentioned and discussed in the
previous chapters), our current understanding of diffusion kinetics and forward
modelling make these observations inconsistent with the other dates and they have to
be ignored, emphasizing instead data from mid-range samples.

189 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

Figure 5.3-3: Examples of good fitting forward modelled exhumation histories for vertical profiles A (top) and B
(bottom). Left column shows time-temperature paths and resulting exhumation rates using the modern
(26°C/km) geothermal gradient, which in both cases produced the best fitting results. Right column shows the
measured (filled markers) vs. predicted (empty markers) ages on age-elevation (A) or age-pseudoelevation (B)
plots (see text for detail on pseudoelevation plotting). Error bars are 1σ (AFT) and 2σ (AHe and ZHe) average
sample errors.

190 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

The first scenario A was modelled based on the thermal history result of sample JC-60-
C (at 2220m a.s.l.) presented previously. While the rapid increase of exhumation to
0.38 km/m.y. between 35 and 25 Ma provides a good fit with the observed AFT data,
the slow rate of 0.16 to 0.21 km/m.y. prior to 35 Ma and 0.02 t0 0.12 km/m.y. within
the last 25 m.y. respectively have the opposite effect on the higher and lower
temperature chronometers. Scenario B shows that a constant Palaeocene exhumation
at 0.18 - 0.23 km/m.y. followed by slower, more gradual exhumation of 0.01 to 0.08
km/m.y. until present day produces better fits with the helium data. In this case the
slope and curvature of the predicted AHe ages resembles the observed age-elevation
relationship closely and thus suggests that slow cooling through the relevant closure
range must have occurred. Nevertheless of all the exhumation scenarios tried, A
produced the best overall fit and consistency with thermal modelling and AFT track
lengths and was thus deemed the best fit solution.

The exhumation history of the northern profiles at Mt Trident and Ptarmigan


Mountain are shown in Figure 5.3-4. For Mount Trident (profile F) a high geothermal
gradient between 30-35°C/km is required to produce fit to predicted ages and the best
fitting history has exhumation rates of 0.14 km/m.y. between 70 and 45 Ma and 0.03
km/m.y. until the highest sample reaches the surface or until 10 Ma where the lower
elevation experienced a final increase in exhumation up to 0.24 km/m.y.. The steeper
gradient (0.42 km/m.y.) before 70 Ma results from existing Ar-Ar ages which put these
samples at 350°C between 80 and 90 Ma. For Ptarmigan Mountain (profile G) best fit
histories are generated by geothermal gradients of 20-30 °C/km and gradual
exhumation at 0.28 – 0.32 km/m.y. from 350°C at 70 Ma followed by slower
exhumation (0.01 – 0.08 km/m.y.) over the past 30 m.y.. While an increase in
exhumation rate over the past 10 m.y. to up to 0.15 km/m.y. predicts ages that lie
within error of measured ages (scenario A) the absence of increase (scenario B)
produces an overall better fit.

191 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

Figure 5.3-4: Examples of good fitting forward modelled exhumation histories for vertical profiles F (top) and G
(bottom). Left column shows time-temperature paths and resulting exhumation rates using a range of
geothermal gradients, which produced the best fitting results. Right column shows the measured (filled markers)
vs. predicted (empty markers) ages on age-elevation (F) or age-pseudoelevation (G) plots (see text for detail on
pseudoelevation plotting). Error bars are 1σ (AFT) and 2σ (AHe and ZHe) average sample errors.

192 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

Figure 5.3-5: Examples of good fitting forward modelled exhumation histories for vertical profiles C (top) and
Jasper (bottom). Left column shows time-temperature paths and resulting exhumation rates using a range of
geothermal gradients, which in both cases produced the best fitting results. Right column shows the measured
(filled markers) vs. predicted (empty markers) ages on age-elevation plots. Error bars are 1σ (AFT) and 2σ (AHe
and ZHe) average sample errors.

193 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

The last two profiles remaining are the southernmost location of the Valhalla
Complex (profile C) and the only profile to the east of the RMT, in Jasper. As shown in
Table 5.2-1 the geothermal gradient of profile C was estimated to be around 35-40
°C/km based on the local surface heat flow. Forward modelling results show that
indeed only high average gradients (30-45 °C/km) are able to produce adequate fits to
the observed ages. Scenario A (35°C/km) generates exhumation rates of 0.31, 0.02 and
0.24 km/m.y. between 60-35, 35-10 and 10-0 Ma respectively. Because the measured
mid-elevation AFT age is younger than the AHe age of the same sample a scenario
fitting both is impossible to achieve, however since the predicted AFT age lies
extremely close to the error of the observed AHe age and the predicted AHe age
agrees perfectly with the measured one, this can be seen as a well-fitting plot. The
same exhumation history produced good fits to the data for geothermal gradients up
to 45°C/km, and was thus deemed the best fitting exhumation history. For the second
scenario (B) an equally high gradient is required. The main difference between A and B
is that in the second scenario the uppermost, sample which lies on the footwall of the
Slocan lake normal fault and thus on the Eocene core complex, moves independently
until about 45-50 Ma when all samples become coupled and the profile is exhumed
uniformly at similar rates as scenario A. Because the Valhalla complex rose during the
phase of Eocene extension commencing 57 Ma and the Slocan lake fault which helped
exhume the underlying Palaeocene meta-sediments was active until 45 Ma [Carr et al.,
1987], this scenario is also possible.

To the east of the Rocky Mountain Trench in Jasper a modern geothermal


gradient based on the surface heat flow and surface rock properties of 18 °C/km was
calculated at the start of this chapter. However the gradients producing the best fits
were indeed higher than the low modern predicted one of 18°C/km, ranging from 30
to 45 °C/km. The best fitting history in Figure 5.3-5 was produced for a gradient of
26°C/km and shows that gradual exhumation at a rate of 0.21 km/m.y. until 40 Ma
and subsequent slowing down to 0.02 km/m.y. produces a suitable history. As for
other profiles, a later increase of exhumation up to 0.22 km/m.y. improved the age fit

194 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

for the AHe data for the lower elevation samples which without it were predicted to
be ~10Ma older.

The range of best-fit cooling and exhumation histories for all profiles is shown in Table
5.3.1 and Figure 5.3-6. Both also summarize the range of probable geothermal
gradients. When comparing these histories from all locations two things stand out: a)
the range of suitable geothermal gradients and the best fitting solutions appear to be
higher on average than the modern calculated gradients based on modern day surface
heat flow presented in Table 5.2-1, and b) the overall shape of the best fitting cooling
histories for all profiles (first column) are similar. The best fitting geothermal gradients
ranged from 26 – 35 °C/km, while a total range of 20 - 45 °C/km produced acceptable
fits for various localities. Based on the influence of prolonged high exhumation rates
on geothermal gradients (section 5.2.3.), higher average gradients such as these are
reasonable for a time period where significant exhumation of rates up to 1 km/m.y.
occurred. Furthermore the Eocene phase of extensional tectonics caused rising of
deeper crustal rocks and thinning of the crust, which likely significantly increased the
surface heat flow at the time, a state which could have persisted well into the
Oligocene. Higher gradients are also plausible since heat producing elements in
material eroded in the past 40 m.y. would serve to increase the equilibrium gradient.

The modelled timing of changes in exhumation rate varies between profiles,


but the general pattern remains the same. Following a rapid phase of post-orogenic
cooling which ended depending on location between 60 and 25 Ma and showed
exhumation rates between 0.12 and 1.1 km/m.y., all profiles show a decrease in
exhumation rates to 0.02 – 0.14 km/m.y. (with a possible higher value for profile B) for
the middle Cenozoic from about ~35 to ~10 Ma. Although the histories shown in Table
5.3-1 and Figure 5.3-6 depict this change to lower rates as abrupt the change would
have been more gradual, as a result of a decreasing isostatically-influenced erosion
rate following the culmination of tectonic events that ended ~45-~60 Ma age,
depending on the locality.

195 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

Table 5.3-1: Acceptable range of exhumation histories for vertical profiles

Min. Max.
Fitting exhumation exhumation
Example of suitable geothermal Identified exhumation rate during rate during
history gradients phases [Ma] phase (top phase (bottom
[°C/km] sample) sample)
[km/Ma] [km/Ma]
Profile A

20 - 26 between 80 and (70 to 65) 0.72 - 1.1 0.77 - 1.1

(best: 26) (70 to 65) - (12 to 10) 0.01 - 0.02 0.01 - 0.02

(12 to 10) - present 0.18 - 0.30 0.23 - 0.32


Profile B

26 - 30 between 60 and (50 to 35) 0.1 - 0.21 0.16 - 0.21

(best: 26) (50 to 35) - (25 to 10) 0.17 - 0.38 0.17 - 0.38

(25 to 10) - present 0.02 - 0.04 0.08 - 0.12


Profile C

30 - 45 between 50 and (40 to 35) 0.31 - 0.39 0.31 - 0.46

(best: 35) (40 to 35) - (7) 0.0 - 0.03 0.02 - 0.03

(7) - present 0 - 0.04 0.0 - 0.24

25 - 30 (60 to 55) - (25 to 20) 0.12 - 0.35 0.12 - 0.35


Profile D

(best: 26) (25 to 20) - present 0.05 - 0.27 0.05 - 0.27

some: 10 - present 0.05 - 0.12 0.09-0.2

between 80 and 70 0.31 - 0.42 0.31 - 0.42


Profile F

25 - 35 70 - (55 to 40) 0.14 - 0.2 0.14 - 0.2

(best: 35) (55 to 40) - (10 to 8) 0.03 - 0.04 0.03 - 0.04

(10 to 8) - present 0 0.23 - 0.24


Profile G

20 - 30 between (65 to 60) and 30 0.28 - 0.32 0.28 - 0.32

(best: 30) 30 - present 0.01 - 0.03 0.02 - 0.08

some: 10 - present 0 - 0.06 0.1 - 0.15

30 - 45 between (80 to 70) and 40 0.14 - 0.21 0.14 - 0.21


Jasper

(best: 26) 40 - present 0.02 - 0.05 0.02 - 0.05

some: 10 - present 0 - 0.02 0.18 - 0.22

196 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

Thus rates of about 0.01-0.27 km/m.y. appear typical of the phase of Oligocene to
Early Miocene ‘hiatus’. Between 12 and 7 Ma exhumation increased, documented
mainly by lower elevation samples, to rates of up to 0.35 km/m.y., with the possible
exceptions of profiles B and G. Figure 5.3-6 illustrates the overall pattern by stacking
all exhumation histories and exhumation rates on top of one another and showing the
overall range of possible exhumation phases and erosion rates found for the entire
study area. Furthermore Figure 5.3-7 shows the exhumation rates plotted against the
geothermal gradients for these best fitting histories for all profiles, overlain by
contours of the corresponding cooling rates ΔT. Most data points fall well below the
10°C/Ma cooling rate isotherm, and thus correspond nicely to the cooling rates found
during the thermal history modelling in chapters 3 and 4. Only the earliest data pairs
(exhumation rates from pre-45Ma rapid exhumation from some profiles) show higher
cooling rates.

Figure 5.3-6: Summary Figure of all fitting exhumation histories which were tested with HeFTy and presented in
Table 5.3-1. Left: All exhumation histories overlain to show possible phases that can be identified for the wider
study area. Right: Range of exhumation rates found for the same exhumation phases.

The overall similarity of the appearance of the best fitting exhumation histories
to the observed cooling histories presented in chapter 3 is striking. The consistency of
this pattern across the whole elevation range covered by the vertical profiles and
seemingly the entire study area suggests that the underlying geological cause for

197 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

exhumation is broadly the same across the study area. This will be investigated further
in section 5.4 below.

Figure 5.3-7: Erosion rate versus geothermal gradients plotted for all 7 vertical profiles. The contours correspond
to the cooling rate ΔT, ranging from 1 °C/Ma to 30 °C/Ma.

In addition to forward modelling numerous exhumation histories for all 7


profiles with HeFTy, 3 profiles (Profile D, C and Jasper) were also modelled with QTQt
(see Chapter 2, section 2.5.2 for detail on this software or diffusion/annealing kinetics
used), using the same boundary conditions and modern temperature ranges as for
HeFTy to assess the two programs in parallel. The results for these 3 profiles are
presented in Figure 5.3-8, Figure 5.3-9 and Figure 5.3-10. These figures were reduced
with the same initial conditions as the HeFTy forward models (Ar-Ar based starting
conditions of 350±50°C at 80±10Ma for Profiles C and D and a slightly lower
temperature start of 200±50°C at 70±10Ma for the Jasper profile, which is based on
the maximum palinspastic reconstruction depth at that time. For all profiles the
current temperature was set to 0±10 degrees and the modern temperature difference

198 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

was assumed to be 0 (all samples are currently at the surface), identically to the HeFTy
models.

The results generally agree with the ones produced by the HeFTy method. The
best fitting history for profile D (Figure 5.3-8 A) is nearly identical to the good fits (A
and B) shown in Figure 5.3-2, with a short phase of rapid exhumation between 35-30
Ma followed by slow exhumation until present day. Similarly to the HeFTy solutions
the range of possible paths for the lower elevation sample (pink envelope) also
suggests a possible increase in exhumation at approximately 10 Ma. The possible
temperature offset ranges given by QTQt lead to a range of likely geothermal
gradients from 14.5 – 47.4 °C/km with the best fitting history being produced at 19.75
°C/km, and thus slightly lower than the 26°C/km gradient suggested by the HeFTy
results.

The QTQt solution for profile C is also almost identical to the solutions
presented in Figure 5.3-5, only QTQt suggest the phase of rapid exhumation to last
from 40-35 Ma instead of the slightly earlier onset of 45-40 Ma presented by the
forward model. However both solutions suggest that the upper elevation sample
reached the surface at approximately 35 Ma while the lower elevation was only
exhumed over the past 10 Ma (red line and pink envelope in Figure 5.3-9). The range
of geothermal gradients given by the QTQt solutions is 18.2 – 23.4 °C with the best
fitting history at 20.5 °C/km and thus again slightly lower than suggested before.

Like the previous two examples the vertical profile near Jasper also shows very
little difference in the thermal histories predicted with HeFTy or QTQt. A steady
exhumation until approximately 40 Ma is followed by slightly slower exhumation and
gradual movement of all samples to the surface. The range of geothermal gradients
found with QTQt is 27.5 – 44.7 °C/km, with the best suited history showing a gradient
of 27.5 and therefore the same as the one predicted with HeFTy.

The similarity of the QTQt results presented for these 3 profiles with the
forward models produced with HeFTy confirms the latter ones, with the only

199 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

Profile D: QTQt forward model

A)

B)
Figure 5.3-8: QTQt forward model results for vertical profile D. A) Best fitting exhumation history for the profile. The lowest
elevation sample best fitting history is shown by the rede line while the upper and lower range of possible histories is indicated
by the pink envelope. The same applies to the highest elevation sample in blue colours. All other sample T-t paths are shown in
grey. The red box indicates the T-t field where hard data exists. The black box indicates the boundary condition of 80±10Ma
and 350±50°C given by existing Ar-Ar ages. B) Observed vs predicted age plot created with QTQt for AFT, AHe and ZHe data as
well as Mean track lengths for the best fitting history shown in A. The X markers indicate predicted ages for other thermal
histories besides the best fitting one.

200 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

Profile C: QTQt forward model

A)

B)

Figure 5.3-9: QTQt forward model results for vertical profile C. A) Best fitting exhumation history for the profile. The lowest
elevation sample best fitting history is shown by the rede line while the upper and lower range of possible histories is indicated
by the pink envelope. The same applies to the highest elevation sample in blue colours. All other sample T-t paths are shown in
grey. The red box indicates the T-t field where hard data exists. The black box indicates the boundary condition of 80±10Ma
and 350±50°C given by existing Ar-Ar ages. B) Observed vs predicted age plot created with QTQt for AFT and AHe data as well
as Mean track lengths for the best fitting history shown in A. The X markers indicate predicted ages for other thermal histories
besides the best fitting one.

201 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

Jasper Profile: QTQt forward model

A)

B)
Figure 5.3-10: QTQt forward model results for vertical profile near Jasper. A) Best fitting exhumation history for the profile. The
lowest elevation sample best fitting history is shown by the rede line while the upper and lower range of possible histories is
indicated by the pink envelope. The same applies to the highest elevation sample in blue colours. All other sample T-t paths are
shown in grey. The red box indicates the T-t field where hard data exists. The black box indicates the boundary condition of
70±10Ma and 200±50°C which was also used for the HeFTy modelling. B) Observed vs predicted age plot created with QTQt for
ZHe and AHe data for the best fitting history shown in A. The X markers indicate predicted ages for other thermal histories
besides the best fitting one.

202 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

difference of a potentially lower temperature range of geothermal gradients.


However, most of the forward models were not tested for gradients lower than 20
°C/km as this appeared to be the lowest likely ‘realistic’ gradient based on the modern
ones calculated from the present day surface heat flow (Table 5.2-1). Therefore it
cannot be said with certainty that HeFTy would not produce suitable histories for
lower gradients.

5.3.2. Quantifying Exhumation

Using the best fitting exhumation histories the total amount of rock cover
removed at each profile can be calculated for any point in time covered by the
exhumation history. This assumes that the average geothermal gradient obtained from
fitting forward models is applicable, as changes in gradient over time would alter
exhumation volume. In order to compare exhumation spatially and in time, two time
intervals were chosen: 45 Ma to present (and thus the time after the end of the phase
of Eocene NW-SE extension) and 10 Ma to present (the period of increased incision
indicated by the youngest ages and best fitting exhumation histories). Table 5.3-2
shows the total exhumation calculated for these intervals for each sample of the seven
profiles. The uncertainties are given by the full range of geothermal gradients and
exhumation histories from Table 5.3-1.

Since 45 Ma the low elevations of the profiles have experienced between 2.65
+1.55/-0.1 km and 6.45 +0.1/-0.3 km of exhumation, while the highest elevations
experienced correspondingly less exhumation, with 1.15 +1.63/-0 km to 5.45 +0.3/-0.3
km of cover removed. For the past 10 m.y., the higher elevation samples appear to
have experienced little exhumation, between 0.1 +0.58/-0 and 0.5 +0.7/-0 km (with
the exception of profile A with 1.8 +1.2/-0 km), while the lower elevations experienced
more, between 0.6 +0.2/-0.1 km and 2.4 +1.0/-0 km.

203 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

For comparison, Table 5.3-2 also shows the modern mean elevation within a 10
km radius and the modern topography corrected estimates of closure depth for the
three thermochronometers below each sample location (calculated as described in
section 5.2.2). Modern closure depths, relative to the mean elevation, for the AHe
isotherm (assumed appropriate temperature ~65°C) range from ~0.6 km to ~3.8 km,
while the AFT depths lie between ~2.3 km and ~6.1 km and the ZHe depths between
~4.2 km and ~8.5km.

In the same way as the exhumation for the vertical profiles is determined, the
total exhumation since the cooling age for the samples which are not part of vertical
profiles can also be calculated by estimating the closure depth. Using the mean
elevation within a 10 km radius in combination with a suitable geothermal gradient
(which was chosen based on the local heat flow and the best fitting gradient of nearby
vertical profiles) the closure depths at each sample locality were calculated (Table
5.3-3). The closure depth of these samples for AHe naturally varies depending on
sample elevation and location and ranges from 1.7 to 2.75 km while the AFT closure
depth ranges from 3.8 to 5.3 km.

Ideally, similar values for the past 45 and 10 Ma should be extracted for these
locations, to enable comparison to the vertical profile sites and across the entire study
area. However, as no vertical information for these locations exists and thus no
detailed exhumation histories can be determined, the only way to obtain such
estimates is linear interpolation to an age/closure depth point and using the resulting
average exhumation rate to calculate the depth at a certain point in time. Although
AFT data has been used in the past to estimate average exhumation rates in areas of
low and steady exhumation (e.g. Brandon et al. 1998), in areas where changes in
exhumation rates through time are likely, such linear interpolation can lead to over- or
underestimation of the amount of exhumation. Similarly, extrapolating past 20 Ma, so

204 | P a g e
Table 5.3-2: Local modern closure depths, best fitting geothermal gradients and past exhumation estimates for vertical profiles

Δh AHe AFT ZHe


Estimated Estimated Exhumation
(sample) AHe AFT ZHe closure closure closure
Profile Elevation error error error exhumation since 45 exhumation since 10 between 45
to mean age age age depth depth depth
samples (m a.s.l.) (± Ma) (± Ma) (± Ma) Ma (km) and Ma (km) and and 10 Ma
elevation (Ma) (Ma) (Ma) (T=65°C) (T=120°C) (T=180°C)
[error]** [error]** (km)
(m)* (m) (m) (m)

Profile A Victor Lake local mean elevation: 1260 m a.s.l. geothermal gradient: 26 °C/km
12-AS-05 1471 -211 - - 61.2 6.5 - - 2711 5311 8211 2.15 [+1.55/-0] 1.80 [+1.2/-0] 0.35
12-AS-06 942 318 10.1 0.6 47.3 2.4 - - 2182 4782 7682 2.65 [+1.55/-0.1] 2.30 [+1.2/-0.1] 0.35

Profile B Pettipiece Pass local mean elevation: 1887 m a.s.l. geothermal gradient: 26 °C/km
JC-311-94 2320 -433 45.3 5.0 - - - - 2933 5233 7833 5.40 [+0.15/-0.8] 0.20 [+0.3/-0.1] 5.20
JC-60C-94 2220 -333 36.3 2.3 33.3 2.1 - - 2833 5133 7733 5.40 [+0.15/-0.65] 0.20 [+0.4/-0.1] 5.20
PCA-304-83 1885 2 30.1 4.7 - - 51.3 1.7 2498 4798 7398 5.70 [+0.1/-0.8] 0.40 [+0.3/-0] 5.30
JC-16-03 1860 27 27.6 6.4 - - - - 2473 4773 7373 5.70 [+0.1/-0.8] 0.40 [+0.3/-0] 5.30
12-AS-10 1490 397 26.2 2.6 28.9 2.2 - - 2103 4403 7003 5.85 [+0.45/-0.65] 0.50 [+0.4/-0.1] 5.35
12-AS-08 1100 787 19.3 2.6 32.4 1.6 - - 1713 4013 6613 6.15 [+0.4/-0.65] 0.70 [+0.4/-0] 5.45
JC-321-94 1030 857 29.8 5.0 - - - - 1643 3943 6543 6.15 [+0.4/-0.5] 0.70 [+0.5/-0] 5.45
12-AS-09 802 1085 30.2 3.5 (21.2) 1.2 52.7 3.0 1415 3715 6315 6.30 [+0.5/-0.5] 0.80 [+0.5/-0] 5.50

Profile C Valhalla Complex local mean elevation: 1074 m a.s.l. geothermal gradient: 35 °C/km
PCA-121-84 2300 -1226 - - 44.4 2.4 - - 3126 4726 6626 3.94 [+2.45/-0.04] 0.34 [+0.1/-0.34] 3.60
12-RP-AS-07 1262 -188 34.9 3.4 28.5 1.8 - - 2088 3688 5588 4.29 [+2.87/-0] 0.69 [+0.52/-0.14] 3.60
PCA-529-83 550 524 - - 26.3 1.6 - - 1376 2976 4876 5.34 [+2.87/-0] 1.74 [+0.52/-0.28] 3.60

205 | P a g e
Δh AHe AFT ZHe
Estimated Estimated Exhumation
(sample) AHe AFT ZHe closure closure closure
Profile Elevation error error error exhumation since 45 exhumation since between 45
to mean age age age depth depth depth
samples (m a.s.l.) (± Ma) (± Ma) (± Ma) Ma (km) and 10 Ma (km) and and 10 Ma
elevation (Ma) (Ma) (Ma) (T=65°C) (T=120°C) (T=180°C)
[error]** [error]** (km)
(m)* (m) (m) (m)

Profile D Mt Revelstoke local mean elevation: 1582 m a.s.l. geothermal gradient: 26 °C/km
12-AS-16 1843 -261 45.2 16.9 30.9 2.3 63.9 27.9 2961 5361 8261 5.45 [+0.3/-0.3] 0.50 [+0.7/-0] 4.95
12-AS-15 1572 10 32.9 2.7 34.5 4.0 - - 2690 5090 7990 5.65 [+0.20/-0.25] 0.70 [+0.6/-0.1] 4.95
12-AS-14 1234 348 30.7 2.5 26.1 1.6 41.5 5.2 2352 4752 7652 6.05 [+0/-0.4] 1.10 [+0.4/-0.4] 4.95
12-AS-13 857 725 25.6 8.8 28.1 1.8 - - 1975 4375 7275 6.15 [+0.2/-0.25] 1.20 [+0.6/-0.4] 4.95
11-AS-11 581 1001 19.4 12.9 - - 44.9 4.6 1699 4099 6999 6.35 [+0.1/-0.2] 1.40 [+0.5/-0.5] 4.95
80-RP-R1b 457 1125 - - 24.5 2.5 - - 1575 3975 6875 6.45 [+0.1/-0.3] 1.50 [+0.5/-0.6] 4.95

Profile F Mt Trident local mean elevation: 1732 m a.s.l. geothermal gradient: 35 °C/km
PCA-307-83 3100 -1213 - - 52.7 7.6 - - 3113 4813 6713 1.15 [+1.63/-0] 0.10 [+0.58/-0] 1.05
11-AS-12 780 1107 - - 18.3 1.1 40.8 4.6 793 2493 4393 3.35 [+1.91/-0] 2.30 [+0.86/-0] 1.05
11-AS-13 601 1286 10.1 5.1 - - - - 614 2314 4214 3.45 [+2.05/-0] 2.40 [+1.0/-0] 1.05

Profile G Ptarmigan Mt local mean elevation: 1238 m a.s.l. geothermal gradient: 30 °C/km
AH-49-1 2347 -1273 41.6 9.0 - - 50.6 22.6 3473 5473 7773 4.50 [+0.8/-2.3] 0.10 [+0.2/-0.1] 4.40
AH49-2 2042 -968 - - - - 43.5 10.5 3168 5168 7468 4.80 [+0.5/-2.2] 0.20 [+0.2/-0.2] 4.60
AH49-3 1737 -663 47.1 12.3 - - - - 2863 4863 7163 5.10 [+0.8/-2.4] 0.30 [+0.2/-0.2] 4.80
AH48-2 1478 -404 49.8 14.6 - - - - 2604 4604 6904 5.10 [+1.0/-2.3] 0.30 [+0.3/-0.1] 4.80
AH48-1 1234 -160 51.1 32.6 - - 47.2 1.6 2360 4360 6660 5.40 [+0.9/-2.5] 0.40 [+0.3/-0.1] 5.00
AH-47-1 716 358 31.8 4.6 - - 39.3 13.8 1842 3842 6142 6.00 [+0.7/-3.0] 0.60 [+0.2/-0.1] 5.40

206 | P a g e
Δh AHe AFT ZHe
Estimated Estimated Exhumation
(sample) AHe error AFT ZHe closure closure closure
Profile Elevation error error exhumation since exhumation since 10 between 45
to mean age (± age age depth depth depth
samples (m a.s.l.) (± Ma) (± Ma) 45 Ma (km) and Ma (km) and and 10 Ma
elevation (Ma) Ma) (Ma) (Ma) (T=65°C) (T=120°C) (T=180°C)
[error]** [error]** (km)
(m)* (m) (m) (m)

Profile Jasper, Rocky Mountains local mean elevation: 1629 m a.s.l. geothermal gradient: 26 °C/km
NMI 37-1 2454 -1380 45.1 6.5 - - 63.3 4.1 3880 6080 8480 2.45 [+2.65/-0.65] 0.20 [+0/-0.1] 2.25
NMI37-02 2362 -1288 40.2 7.0 - - 58.0 2.3 3788 5988 8388 2.55 [+2.65/-0.65] 0.30 [+0/-0.1] 2.25
NMI37-03 2128 -1054 41.4 12.5 - - 56.5 10.8 3554 5754 8154 2.75 [+2.65/-0.65] 0.50 [+0/-0] 2.25
NMI37-04 1996 -922 39.1 11.3 - - 55.8 4.6 3422 5622 8022 2.85 [+2.75/-0.65] 0.60 [+0.1/-0] 2.25
NMI38-05 1772 -698 37.7 9.8 - - 56.8 6.2 3198 5398 7798 3.15 [+2.65/-0.75] 0.90 [+0.0/-0.5] 2.25
NMI39-06 1706 -632 34.2 8.0 - - 54.8 8.4 3132 5332 7732 3.15 [+2.65/-0.65] 0.90 [+0/-0] 2.25
NMI39-07 1491 -417 30.9 3.7 - - 53.3 9.1 2917 5117 7517 3.35 [+2.75/-0.65] 1.10 [+0.1/-0] 2.25
NMI 39-8 1297 -223 20.7 2.7 - - 58.7 4.5 2723 4923 7323 2.60 [+3.7/-0] 1.30 [+0.1/-0] 1.30

* Mean elevations of profile locations shown in table 5.3.2. Closure depths are calculated in relation to those mean elevations (after Brandon (1998)). **Estimate for exhumation
is calculated based on the best fitting exhumation histories and geothermal gradients presented in previous tables and figures.

207 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

using the linear gradient from such a young age to estimate exhumation since e.g. 45
Ma is likely connected to an even greater uncertainty, and has therefore not been
attempted.

For these reasons the remaining samples were used only to make tentative
estimates with linear interpolation for exhumation since 10 Ma. The resulting
estimated average exhumation rates and total exhumation are shown in Table 5.3-3.
The errors given are calculated using the average sample age error, bearing in mind
the inherent uncertainty of the method. Where no AHe data age was available, the
AFT age was used, which is indicated by square brackets. In these cases the greater
depth of the AFT isotherm naturally also leads to a greater absolute uncertainty (in
km). In cases where the AFT age was younger than the AHe age (and thus inverted),
both estimates are given (AFT interpolation given in round brackets). The resulting
average exhumation rates range from 0.02 to 0.21 km/m.y. and the total exhumation
thus ranges from 0.21 ± 0.63 km to 1.98 ± 0.28 km, which is similar to the rates and
amounts observed for the vertical profiles. Although there are some exceptions, a
general correlation between low elevation samples and high exhumation, as
demonstrated for the vertical profiles before, can be observed.

In order to identify whether there is a lateral pattern of exhumation, all


exhumation estimates for the past 10 m.y., including the lowest elevation samples of
the vertical profiles, are shown in Figure 5.3-11. Here as well square brackets indicate
exhumation estimated from AFT age interpolation, rather than AHe. Although the AFT
interpolated estimates inherently come with greater uncertainty, the exhumation
values calculated from the average rates fit in well with the pattern observed from the
AHe interpolated exhumation. The main observation made from colour coding the
exhumation totals is that the highest exhumation of over 2 km is located along the
southward draining part of the Columbia River in Mica Creek (51-52°N, roughly
118.5°W), and the exhumation values appear to diminish to values of less than 0.5 km
at the southern edge near the US border and to the east, in the Rocky Mountains. In

208 | P a g e
Table 5.3-3: Wider sample exhumation estimates in relation to local mean elevation

AFT AHe
Mean Δh Average Min.
Estimated closure closure
Sample elevation (sample) recent exhumation
AHe age error AFT age error geothermal depth depth ± Error
Sample Coordinates (decimal minutes) elevation in 10 km to mean exhumation since 10
(Ma) (+/- Ma) (Ma) (+/- Ma) gradient* (T=120°C) (T=65°C) (km)
(m a.s.l.) radius elevation rate** Ma***
(°C/km) below below
(m a.s.l.) (m) (km/Ma) (km)
sample (m) sample(m)

11-AS-01 50°50'25.58"N 116°20'50.64"W 22.8 3.3 - - 929 962 33 26 5067 2467 0.11 1.08 0.36
11-AS-09 50°38'6.83"N 117°55'40.98"W 19.1 6.3 - - 454 517 63 26 5037 2437 0.13 1.27 0.81
11-AS-15 51°17'38.54"N 118°16'20.39"W 10.5 1.4 - - 605 1026 421 26 4679 2079 0.20 1.98 0.28
11-AS-22 52°44'40.56"N 119° 5'47.26"W 33.6 16.9 - - 791 1034 243 26 4857 2257 0.07 0.67 1.14
11-AS-26 53° 3'36.54"N 119°36'50.04"W 18.0 1.5 - - 741 1062 321 26 4779 2179 0.12 1.21 0.18
12-AS-01 51°27.160'N 116°17.974'W 54.0 1.4 - - 1649 1977 328 30 3872 1872 0.03 0.35 0.05
12-AS-11 50°44.322'N 116°45.570'W 24.7 6.4 29.3 1.5 1956 2312 356 26 4744 2144 0.09 0.87 0.56
12-AS-12 50°47.312'N 116°41.272'W 31.5 10.3 - - 1449 1899 450 26 4650 2050 0.07 0.65 0.67
12-AS-17 52°33.223'N 119°02.656'W 37.4 0.1 - - 1740 1488 -252 26 5352 2752 0.07 0.74 0.01
12-AS-18 52°33.337'N 119°04.176'W 32.8 12.4 - - 1278 1488 210 26 4890 2290 0.07 0.70 0.86
12-AS-19 52°37.027'N 119°07.979'W 20.3 8.8 (11.0) 0.7 875 1488 613 26 4487 1887 0.09 0.93 (4.1) 0.41 (0.29)
12-RP-AS-4 49°13.246'N 116°36.103'W 48.0 6.2 (33.8) 4.0 562 881 319 26 4781 2181 0.05 0.45 (1.4) 0.14 (0.57)
12-RP-AS-8 49°17.144'N 117°39.951'W 37.6 1.7 (34.0) 2.7 632 650 18 26 5082 2482 0.07 0.66 (1.5) 0.15 (0.40)
80-RP-K3 50°18'00''N 117°40'00''W - - 24.9 3.5 762 1444 682 26 4418 1818 0.18 [1.77] 0.28
80-RP-R2 50°58'00''N 118°22'30''W - - 20.1 2.9 549 1347 798 26 4302 1702 0.21 [2.14] 0.26
80-RP-R3 50°57'40''N 118°44'10''W - - 26.4 3.4 396 709 313 26 4787 2187 0.18 [1.81] 0.31
80-RP-R6 50°45'30''N 119°20'50''W - - 39.7 5.0 375 656 281 26 4819 2219 0.12 - -
80-RP-R9 51°02'45''N 117°57'00''W - - 21.0 2.5 610 1250 640 26 4460 1860 0.21 [2.12] 0.23
Ice River 51°7'59.52"N 116°26'4.02"W 38.5 3.8 - - 1430 1731 301 30 3899 1899 0.05 0.49 0.19
J91-1 52°55'8.33"N 118° 0'41.29"W 97.7 30.8 - - 1078 1262 184 30 4016 2016 0.02 0.21 0.63
J91-18 53°13'46.52"N 117°29'45.13"W 41.4 5.4 - - 1369 1413 44 30 4156 2156 0.05 0.52 0.28
J91-2 52°52'47.32"N 118°26'13.88"W 25.1 5.4 - - 1157 1527 370 30 3830 1830 0.07 0.73 0.39
PCA-200-83 52°28'57.50"N 118°53'21.70"W 26.8 8.9 (16.5) 1.9 2297 2009 -288 30 4488 2488 0.09 0.93 (2.7) 0.27 (0.52)
PCA-58-84 49°19'44.86"N 117°41'33.41"W - - 30.3 1.9 440 771 331 30 3869 1869 0.13 [1.28] 0.11
* Geothermal gradients estimated from best fitting results of vertical profiles in nearby locations: 26 C/km for all areas east of the Rocky Mountain Trench except the southernmost samples in the Vicinity of the Valhalla Complex, where a gradient of 30 C/km was deemed more
appropriate. The localities west of the RMT were estimated to have a similarly high (30 C/km) average Cenozoic gradient. ** Average recent exhumation rate calculated from AHe age (which is closest to the surface) and the appropriate calculated closure depth at sample location. ***
Minimum exhumation calculated based on exhumation rate from Helium age and closure depth. Where the age relationship between AHe and AFT was inverted (younger AFT ages in brackets), the older (AHe) age was used to estimate the minimum exhumation. The exhumation values
that would result from the AFT ages are shown in brackets.

209 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

Figure 5.3-11: Estimated exhumation since 10 Ma of low elevation samples of the study area, including the
lowest elevation samples from vertical profiles. Total exhumation of more than 2 km is shown in red, while
yellow, green and blue markers indicate 1-2 km, 0.5-1 km and less than 0.5 km respectively. Square brackets: AFT
interpolated data.

the north they remain slightly higher (>0.5 km) until the edge of the studied area.
Because of the lack of data in the central Rocky Mountains in the east, it is hard to say
how this pattern would continue across the trench, i.e. whether the centre of highest
exhumation is located only in the Columbia Mountains or indeed stretched across and
covers the whole central area.

5.4. Paleo-Relief and Paleo-Elevation

The main aim of this study is to characterize the spatial and temporal
development of exhumation and thus make inferences about the age of the
topography we observe today. As described in chapter 1, the prevailing view is that

210 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

the eastern Cordillera has not experienced significant changes in exhumation since the
end of the orogeny (~45 Ma) and that particularly the Rocky Mountains to the east of
the Rocky Mountain Trench have since been lowered from a high-elevation hilly
plateau to their current elevation by gradual exhumation through erosion since the
upper Cretaceous [Osborn et al., 2006], albeit rejuvenated by Pleistocene glacial
erosion. Alternatively, because the mountains are high and rugged, they may have
experienced uplift that may have led to both pre-glacial and glacial erosion to produce
the current ranges, in spite of lack of evidence for recent tectonics.

The exhumation histories determined by this study have so far shown that
exhumation rates have varied in time and space over the past 45 Ma. Following the
~45 Ma end of orogenic events, rates slowed from up to 1.1 km/m.y. to between 0.01
and 0.15 km/m.y. in the Oligo-Miocene. During the early Miocene the rates may have
been near zero and increased since ~10 Ma to up to0.32 km/m.y., giving rise to the
existing elevated and incised landscape. The spatial distribution of this most recent
exhumation was analysed in Figure 5.3-11 as far as data density permitted and found
to show greatest exhumation along the valley separating the Monashee and Selkirk
Mountains (Columbia River north of Revelstoke).

5.4.1. Paleo-Relief from Back-stacking Exhumation Estimates

To investigate paleo-relief, the exhumation estimates calculated previously


(Table 5.3-2 and Table 5.3-3) are stacked back on top of the sample localities to assess
the shape of the paleo land surface. For the vertical profile locations Figure 5.4-1
shows the reconstructed surfaces at 45 Ma and 10 Ma and in comparison to the
current topography. This visually portrays estimates of the total amount of thickness
of rock removed that is above the current elevations of localities. The black markers
indicate the amount of exhumation resulting from the chosen best fitting histories. For
the total exhumation since 45 Ma, which has large uncertainties, several alternative

211 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

well-fitting exhumation history results are indicated by white circles. In terms of the
10 Ma estimates, the reconstructed ‘surface’ is reasonably similar between higher and
lower elevation samples once the difference in present-day elevation is taken into
account. This ‘surface’ ranges mainly from 1.2 to 3.5 km in current elevation
considering all profiles; within each profile, the relief of this ‘surface’ appears to be
less than 1.5km, with most being several hundred metres. This suggests that there is
no compelling evidence for relative differences in ‘paleo-surface’ elevation exceeding
several hundred metres within a single profile; this could be interpreted to suggest a
lack of relief on the horizontal scale of the profiles (4-15 km), in contrast to the
modern relief. This conclusion is especially relevant to the ~10 Ma late Miocene
period since uncertainties are lower than the older reconstructions. Given the absence
of evidence for significant relief at ~10 Ma, this allows one to conclude that the
modern relief is relatively young, as the valleys where these samples were taken
appear to have only been created since ~10 Ma. Broadly, the samples located outside
the vertical profiles (Table 5.3-3) show a similar result. Higher exhumation estimates
for the past 10 m.y. are generally observed at modern lower elevations, while the
higher elevations display considerably less removal of cover. This suggests that this
former low relief topography on a 5-15km scale typified the study area and that the
modern topography was incised since late Miocene time.

The back-stacked ‘surface’ at 10 Ma varies from 1.2 to 3.5 km in ‘current’


elevation, as shown in the tables above. Most are between 1.5-2.7 km in ‘current’
elevation although two profiles (A, Victor Lake near Revelstoke; F, Trident Mountain)
are 3.0-3.5km. If one observes the highest summits within the vicinity of these
profiles, they are broadly similar in elevation to these reconstructed ‘palaeo-surfaces’.
This observation could be used to infer that on a larger spatial scale of several hundred
kilometres, the land surface may have had relief of less than 1 km, but with some
higher residual peaks or uplands some hundreds of metres high, with total relief being
1.0- 1.5 km about 10 m.y. ago.

212 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

Figure 5.4-1: Exhumation at vertical profile locations for two time intervals, 45 Ma and 10 Ma. Amount of cover
removed (in km) at individual sample locations (black markers) calculated from best fitting exhumation histories.
Error bars show the range of all well-fitting histories, of which examples for 45 Ma are shown by white circles.
Navy blue dashed lines indicate the total exhumation calculated from the thermal history models stacked on top
of the current topography. Modern Mean elevation (MME) and isostatically compensated mean elevation ranges
at 45Ma (45ME) and 10 Ma (10ME) (see section 5.4.2.) in relation to modern sea level are indicated by grey bars.
The modern maximum elevations within a 10 km radius are indicated by the grey dashed lines ‘max. peak’.

213 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

These conclusions are broadly consistent with the results of the study in the
Rocky Mountains by Osborn et al. (2006), who, based on their stratigraphic,
palinspastic and critical taper reconstructions concluded that the post-orogenic Rocky
Mountains resembled a hilly plateau that subsequently eroded to its current relief
much more recently. They concluded that long-term low-relief topography may have
been influenced by lithological contrasts between resistant limestone and quartzite of
Palaeozoic age and Mesozoic clastic rocks. The implication is that until resistant rocks
were extensively exposed, following stripping of the deformed but overlying Mesozoic
rocks, the surface may have had less relief. They assumed constant exhumation and
suggested that the modern topography emerged when 75% of the Mesozoic rocks (5-8
km in total, according to their palinspastic reconstruction) were removed, allowing the
steeply-dipping and resistant limestones of the front ranges to dominate the
topography. A rough estimate of timing for this was given as mid-Miocene, though
that was conjecture on their part. However, both the timing and total exhumation
estimates of Osborn et al. (2006) are consistent with the results presented here, not
only for the Rocky Mountains in the east (Jasper Profile) but crucially locations in the
Columbia Mountains as well (Profiles A-G and most remaining sample locations).

5.4.2. Paleo-Valley Inferences

Possible exceptions to the low-relief paleo-topography documented in this


dataset are the Rocky Mountain Trench and the North Thompson River valley (location
of vertical profile E, which was not modelled above but all samples are included in
Table 5.3-3). Although no vertical profile for a flank of the trench is available, the
samples located inside the RMT (11-AS-01, 11-AS-22 and 11-AS-26) show only a
relatively small difference of exhumation when compared to the nearest samples on
high elevations and peaks (12-AS-01, PCA-307-83 (top of Profile F), Ice River, and the
Bugaboo Glacier samples 12-AS-11 and -12). The exhumation for these valley samples
ranges from 0.7 ± 1.14 to 1.21 ± 0.18 km over the past 10 m.y. while the higher

214 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

elevations have experienced between 0.1 ± 0.58 and 0.87 ± 0.56 km of cover removal.
With a local relief of the RMT far exceeding 2km, even the two values at either end of
the scale (< 0.1km removed from the tops and up to 1.21 km from the valley locations)
would still require a paleo-valley near the current RMT with at least 1.5 km relief
before 10 Ma. Similarly the low and high elevation samples of vertical profile E, 12-AS-
19 and PCA-200-83 show 0.93 ±0.41 and 0.93 ± 0.27 km of estimated cover removal
respectively (based on the AHe interpolation) since 10 Ma. An elevation difference of
>1400 m suggests that this valley most likely already existed during the Miocene and
could be even older. Thus some major valleys may have existed in the past and
persisted (the RMT being one of them), despite the fact that most of the present day
topography is young.

5.4.3. Former Mean-Elevation Estimates from Isostasy

Determining the relief from exhumation is much simpler than estimating the
absolute elevation in the past. Rock cooling data is not able to provide information
regarding absolute surface uplift in relation to e.g. sea level or the geoid [England and
Molnar, 1990]. Confusion arose from some early thermochronology studies, which
determined erosion rates and amounts, but did not necessarily distinguish between
rock uplift, i.e. moving samples from greater depth to the surface, and uplift of the
land surface. The terminology was clarified by England and Molnar (1990) who stated
that

(4) Surface Uplift = Rock Uplift – Exhumation.

In an actively forming orogen, rock uplift can be dominated entirely by tectonic


forcing, which even with high erosion rates would generally result in some surface
uplift. When tectonic forcing has ended and a mountain range has entered its post-
orogenic and mainly erosional phase (such as is likely in our case), the main driving
force causing rock uplift is the isostatic response to erosion. In the absence of tectonic

215 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

uplift, erosional removal of material will gradually reduce the elevation of the land.
However, isostasy plays an important role in moderating the lowering of the land
surface by erosion.

To illustrate this, we assume there was no surface uplift (due to e.g. tectonic
forcing) and assume that the region is in isostatic equilibrium and the state of the
mantle has not changed. In this case, the isostatic compensation for the observed
erosional exhumation can be calculated. By deducting the exhumation compensated
for isostatically from the observed total exhumation we can thus estimate the former
mean elevation and compare it to the modern mean elevation. Using a simplified local
1-dimensional Airy isostatic compensation equation for post-orogenic erosion (after
[Watts, 2001], section 7.2.3 eq. 7.55) the difference in mean elevation to the current
mean elevation (ΔH) can be calculated:

𝜌𝑐
(5) ΔH = ΔE / [1 + 𝜌𝑚−𝜌𝑐]

where ΔE is the total exhumation and ρc and ρm are the average crustal and mantle
densities respectively. For a first approximation to compare all different time intervals,
common average densities of the crust and mantle of ρc = 2800 kg/m3 and the ρm =
3330 kg/m3, respectively, were chosen [Watts, 2001]. However, although the current
crustal density is estimated to be an average of ρc = 2850 kg/m3, and thus close to this
estimate, the modern mantle density is given as ρm = 3260 kg/m3 by Hyndman et al.
[2011]. This is due to the modern back-arc position and anomalously hot
asthenosphere of the Cordillera. Because it is not clear how far into the past this low
density mantle extended, the average values were chosen to give comparable
estimates. This topic will be discussed further in section 5.5.

The last phase of exhumation (< 10 Ma) caused the formerly subdued landscape
to be incised by removal of up to ~2.4 km of rock above modern day valleys, while the
exhumation at higher elevations was considerably less. As a result of this spatially
varied exhumation the overall average volume of rock removed is fairly low with ~0.1-
0.4 km. According to equation (5) this results in a similarly moderate isostatic lowering

216 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

of the mean elevation by only ~400 m (200 m in Jasper and thus the Rocky
Mountains). Once again this is assuming no change in mantle density during this
period and assuming the area has reached isostatic equilibrium. Because sea level at
10 Ma was the same as today, the 10 Ma mean elevation of the Columbia Mountains
would thus have ranged from ~1250 m to 2200 m a.s.l. and in the Rocky Mountains
from 1700 – 2200 m a.s.l., assuming the ΔH estimates can be applied to the wider
region.

Table 5.4-1 and Figure 5.4-1 show the estimated results for ΔH for all profiles at
45 and 10 Ma. For the Columbia Mountain profiles average exhumation of 2.7 [+1.55/-
0] to 6.5 [+0.2/-0.25] km (in relation to the MME) would have resulted in a decrease of
local mean elevation since the mid-Eocene (45 Ma) of ~0.4 [+0.25/-0] to ~1.0 [+0.03/-
0.04] km. In Jasper, exhumation of 2.6 [+2.65/-0.65] km would have lowered the mean
elevation by ~0.41 [+0.42/-0.1] km, while the remaining exhumation would have been
compensated for isostatically. Assuming these estimates can be extrapolated to the
wider study area we can estimate that the mean elevation of the Columbia Mountains
at 45 Ma could have been between ~1600 [+250/-0] and ~2800 [+30/-40] m a.s.l.,
since the current mean elevation ranges from 1200 – 1800 m a.s.l. (Figure 5.2-4). For
the Rocky Mountains, where the current mean elevation is slightly higher at 1500 to
2000 m a.s.l., the mid-Eocene mean elevation could have ranged from 1910 [+420/-
100] to 2410 [+420/-100] m a.s.l.. Note that these values are given in relation to the
modern sea level and that it is assumed that the state of the mantle remained
constant with a density of 3330 kg/m3. Because the average global sea level was ~50 m
higher at 45 Ma [Miller et al., 2005] this amount would have to be deducted from the
mean elevation to relate the elevation to the sea level of the past. These are not
robust conclusions to rely on however, due to the unsafe assumption that the mantle
has remained unchanged, but they illustrate the effect of isostatic adjustment as a
result of erosion of the upper crust. Because of the rapid unloading effect of

217 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

deglaciation on a ~12,000 year time scale, the area may not yet be in complete
isostatic equilibrium, though this is likely to be a minor residual effect.

The last phase of exhumation (< 10 Ma) caused the formerly subdued landscape to
be incised by removal of up to ~2.4 km of rock above modern day valleys, while the
exhumation at higher elevations was considerably less. As a result of this spatially
varied exhumation the overall average volume of rock removed is fairly low with ~0.1-
0.4 km. According to equation (5) this results in a similarly moderate isostatic lowering
of the mean elevation by only ~400 m (200 m in Jasper and thus the Rocky
Mountains). Once again this is assuming no change in mantle density during this
period and assuming the area has reached isostatic equilibrium. Because sea level at
10 Ma was the same as today, the 10 Ma mean elevation of the Columbia Mountains
would thus have ranged from ~1250 m to 2200 m a.s.l. and in the Rocky Mountains
from 1700 – 2200 m a.s.l., assuming the ΔH estimates can be applied to the wider
region.

Table 5.4-1: Total exhumation at profile locations in relation to modern mean elevation and
isostatically compensated difference to the current mean elevation.

MME [m ~ΔE (45Ma) [km] to ~ΔE (10Ma) [km] to ΔH (45ME vs MME) ΔH (10ME vs MME)
Profile
a.s.l.] MME MME [m] [m]

A 1260 2.7 [+1.55/-0] 2.3 [+1.2/-0] 422 [+247/-0] 366 [+191/-0]


B 1887 6.3 [+0.1/-0.8] 0.8 [+0.3/-0] 1003 [+16/-127] 127 [+48/-0]
C 1074 5.3 [+2.87/-0] 1.7 [+0.52/-0.14] 850 [+457/-0] 277 [+83/-22]
D 1582 6.5 [+0.2/-0.25] 1.5 [+0.6/-0.1] 1027 [+32/-40] 239 [+95/-16]
F 1732 3.5 [+1.91/-0] 2.4 [+0.86/-0] 549 [+304/-0] 382 [+137/-0]
G 1238 6.0 [+0.9/-0.25] 0.6 [+0.3/-0.1] 955 [+143/-40] 95 [+48/-16]
Jasper 1629 2.6 [+2.65/-0.65] 1.3 [+0.1/-0] 414 [+422/-103] 207 [+16/-0]

MME: Modern mean elevation at profile location, ΔE: estimated rock column removed in relation to modern mean elevation, ΔH:
Resulting estimates for difference of mean elevation in the past (45ME and 10 ME) compared to current mean elevation (MME)
after Airy isostatic compensation, calculated after Watts (2001), see equation (5) above. Positive values indicate that the mean
elevation in the past was higher by ΔH compared to the modern mean. Errors are calculated as before for the range of suitable
exhumation histories in Table 5.3 1

218 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

5.4.4. Neogene Summit Uplift and Valley Incision

As shown above, exhumation during the Neogene was localized and created
the current relief through incision of valleys, whilst removing little rock from current
high elevations. The moderate lowering of the mean elevation (assuming no mantle
changes) since the early Neogene because of moderate removal of average rock
volume, and the reconstructed flat paleo-surfaces in Figure 5.4-1 therefore imply that
considerable surface uplift of the highest elevations (summits) must have occurred
since ~10 Ma. Figure 5.4-2 illustrates how this is possible using a schematic example. If
the reconstructed surface at 10 Ma shows little relief, any given point at the surface
must have been relatively close to the 10 Ma mean elevation (10ME). Because
significant exhumation has occurred since, particularly in valleys, the modern surface
was at that time still buried. Over time, a significant incision of valleys and little
erosion at the high elevations cause the mean elevation to drop slightly, whilst
isostatic compensation for the removal of burden triggers rock uplift, bringing the
modern surface closer to the actual surface. Once the present day state of relief is
created and the modern surface exposed, the mean elevation has been lowered to the
current mean elevation. However, in response to the localized incision rock uplift has
caused the modern summits to stand above the 10 Ma mean elevation, and thus
approximately the paleo-surface. So, even if the mean elevation does not change and
the mantle state is also constant, simply the creation of relief (e.g. from increased
incision) will lift the peaks.

This principle can be applied to Figure 5.4-1, where for most locations the
current surface as well as the current maximum elevation within a 10 km radius
(indicated as ‘max. peak’) are both well below the reconstructed 10 Ma surface, yet up
to ~1 km above both the 10 Ma and present mean elevation. For those localities
where the current summits are above the reconstructed 10 Ma surface some degree
of paleo-topography (~500 – 1000 m relief) can be assumed (e.g. profiles D and G).

219 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

Figure 5.4-2: Schematic illustration of how the creation of relief can create considerable summit (surface) uplift
while lowering the mean elevation. At 10 Ma a low relief surface is located close to the the 10 Ma mean
elevation (10ME), while the modern surface is still buried under ground. Localized incision of valleys over time (5
Ma and present) triggers isostatic compensation in response to the removed burden, causing rock uplift and
bringing the theoretical modern surface closer to the actual surface. However because the exhumation is
restricted to valleys and summits experience little erosion, they rise isostatically in elevation, causing ‘summit’
surface uplift. Despite this increase of the maximum elevation the creation of pronounced relief still causes the
mean elevation to lower slightly.

Table 5.4-2: Mean elevation development and surface uplift since 10 Ma

Max.
Max
peak in MME 10 km radius peak
profile Profile peak uplift
Profile 10 km [m 10ME [m a.s.l.] uplift (estimated)
elev. [m]
[m a.s.l.] [m]
[m]
a.s.l.]

A 1471 2692 1260 1626 [+135/-48] -155 [+191/-0] 1066 [+191/-0]


B 2320 2825 1887 2014 [+64/-48] 306 [+48/-0] 811 [+48/-0]
C 2300 2506 1074 1351 [+72/-48] 949 [+83/-22] 1155 [+83/-22]
D 1843 2622 1582 1821 [+0/-80] 22 [+95/-16] 801 [+95/-16]
F 3100 3177 1732 2114 [+127/-0] 986 [+137/-0] 1063 [+137/-0]
G 2347 2367 1238 1333 [+48/-16] 1014 [+48/-16] 970 [+48/-16]
Jasper 2454 2820 1629 1836 [+0/-111] 618 [+16/-0] 984 [+16/-0]
average: 534.22 average: 978.58
MME (modern mean elevation) and Max. peak (maximum elevation) within 10 km radius of profile centre. Peak uplift calculated in
reference to Miocene mean elevation (10ME). Errors re the errors of the past mean elevation estimates, which are determined by the
highest and lowest exhumation rates found above. No correction for sea level change is needed when comparing 10 Ma elevations
to present day elevation, because the 10 Ma and present sea level are roughly the same [Miller et al., 2005].

220 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

Using the 10 Ma mean elevation estimates and the current highest profile
elevations as well as the nearby highest summit elevations, the maximum surface
uplift from this process in the vicinity of the profiles can be estimated (Table 5.4-2).
The highest elevation sample localities of the profiles have experienced between 22
[+95/-16] m and 1014 [+48/-16] m summit uplift over the past 10 m.y. in relation to
the 10 Ma mean elevation (with the exception of profile A, where the low modern
sample elevations caused the highest sample to subside by -155 [+191/-0] m). As
expected, the highest present day elevations within the 10 km radius, i.e. the actual
‘summits’, show more Neogene surface uplift. Maximum summit uplift ranges from
801 [+95/-16] m to 1155 [+83/-22] m, and averages at ~ 1 km. It should be noted that
where some degree of relief existed within the 10 km radius in the past, which cannot
be seen from the reconstructed surface at the profile locations, these estimates could
be overestimating the peak uplift by several hundred meters (e.g. profile D, where the
current peak lies above the reconstructed (back-stacked) paleo-surface, which could
indicate a paleo-high). However, since the paleo-topography (i.e. paleo peak heights)
for the wider 10 km radius is not known, there is no way of improving on these
estimates at this point.

5.5. Causes of Exhumation and Uplift

The principle outcome of this study is that there have been two phases of
exhumation across the region. The first occurred during and following orogenic events
when the higher crustal thickness may have supported a higher elevation; the second
occurred since 10 Ma following a period of stability, with amounts of total exhumation
ranging from 0.3 to almost 3 km. The causes of these will now be discussed in three
stages: 1) Orogenic events between 70 – 45 Ma, 2) Back-arc processes influencing
exhumation since 45 Ma and 3) the increase of exhumation rate at 10 Ma – present.

221 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

5.5.1. Orogenic Events and Exhumation 70-45 Ma

Many of the rocks of the study area were at upper amphibolite facies
conditions and being strongly deformed in the Cretaceous to Eocene time. This
included crustal thickening arising from the underthrusting of the craton beneath the
collage of deformed rocks to the west, including both continental margin rocks and
overlying obducted terranes. The chronology of these events has been outlined in
chapter 1, and culminated in the production of thick crust with multiple stacks of
thrusts, including repetition of the cratonic basement and its overlying
metamorphosed cover in mainly easterly verging structures. This process also
produced the imbricated foreland fold and thrust belt of the Rocky Mountains. This
contraction process ended at 55-60 Ma and was immediately followed by a period of
crustal and lithospheric extension during the Eocene until cessation at 45 Ma
[Tempelman-Kluit and Parkinson, 1986; Price and Carmichael, 1986]. The period of
extension, including rising of core complexes, dramatically thinned the crust from
perhaps >60 km to less than 40-50 km by 45 Ma within the region of extension (mainly
the Columbia Mountains and Interior Plateau, Figure 1.2-6) of southern British
Columbia. Elsewhere, for example in the Rocky Mountains, this period of extension
was absent. Erosion and tectonic exhumation took place from 70 to 45 Ma with a 5-15
km reduction of the thickness of crust as a result of tectonic exhumation and/or
erosion. However, for the purposes of this contribution, more emphasis will be placed
on the discussion of processes occurring after deformation ceased, following the end
of extension at 45 Ma.

5.5.2. Crustal and Back-arc Mantle Processes from 45 Ma to the


Present
In terms of the prevailing view of subsequent evolution, the late- and post
extensional phase was characterised in the mantle by delamination of the back-arc

222 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

lithosphere due to asthenosphere convection [Currie et al., 2008]. During the middle
Cenozoic, there was general tectonic stability and presumably some adjustment of the
mantle to a more stable, less active back arc convection situation, until the middle
Miocene.

5.5.2.1. Lithosphere Delamination

Although eastward subduction occurred underneath the North American


continent since at least ~160 Ma (which led to the accretion of the terranes),
lithosphere delamination is generally believed to have commenced at ~55 Ma
[Hardebol et al., 2012; Bao et al., 2014] when a sudden increase in arc magmatism
indicates acceleration of subduction [Haeussler et al., 2003; Madsen et al., 2006]. The
subducting Resurrection Plate was completely consumed by ~35 Ma and replaced by a
slab window beneath southern British Columbia, later followed by the northern
Farallon slab (now the Juan de Fuca plate) [Madsen et al., 2006]. While Bao et al.
(2014) suggested that this delamination occurred rapidly and most of the lithosphere
was detached as one slab within a few million years [Bao et al., 2014], most other
authors (e.g. [Currie et al., 2008; Hardebol et al., 2012]) believe that delamination
would have been a more gradual process and would have occurred in stages. Currie
(2008) suggested that delamination driven by frictional heating and stripping of the
lower layer of the lithosphere would have been significant after 5 – 10 Ma; however
based on their model complete stripping to a stable thickness that is maintained by
thermal buoyancy could take up to 60 m.y..

Based on this timeline we can assume that significant delamination would have
occurred from 45 Ma until the end of the first exhumation phase (35 – 25 Ma) and
slowed while subduction ended and the region was above the Pacific - Farallon slab
window (see Figure 1.2-9 for detail). As a result of this delamination, the modern
lithosphere thickness in the back-arc is estimated to be ~60 km, while underneath the
stable craton (east of the RMT) the mantle boundary lies at ~200 km (Figure 5.5-1).

223 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

The immediate effect of this is hot, low viscosity mantle material at shallower depths
underneath the back-arc and thus an increased geothermal gradient.

Figure 5.5-1: (A) Bouguer gravity along with observed topography and predicted topography for cold shield
geotherm (for the Columbia Mountains) from [Hyndman, 2010]. (B) Schematic illustration of the predicted
thermal situation of the Back-arc- Craton transition from tomography data, also from [Hyndman, 2010].

The high temperatures at shallower depths cause thermal expansion of both


the asthenosphere and lithosphere of the back arc and subsequent thermally
supported uplift of the back-arc surface. Hyndman et al. (2009, 2010) calculate that
the average modern temperature difference between the Cordilleran back-arc
geotherm and the cratonic geotherm in the east is ~250°C on average to a depth of
~200 km (‘reference level’) (Figure 5.5-1), where the geothermal gradients merge to
the normal mantle adiabat. This results in compositionally corrected isostatic uplift
from thermal expansion (with thermal expansion coefficient of 3.2 x 10-5°C-1, after
[Hasterok and Chapman, 2007]) of consistently ~1600 m of both the Interior Plateau

224 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

and the Columbia Mountains [Hyndman et al., 2009]. In other words, thermal
buoyancy causes the crust to the west of the RMT to be ~1600 m higher today than it
would be if it had a normal thickness lithospheric root, the current crustal thickness,
and a cratonic temperature-depth profile.

The process of delamination and the resulting surface uplift would therefore
have increased erosion at the surface by increasing elevation, in turn increasing rock
uplift and contributing to this first post-orogenic exhumation phase. Surface uplift of
up to ~1.3 km is recorded in the Interior Plateau during the upper and middle Eocene
from a drop in fluvial base level and subsequent incision of canyons ([Tribe, 2005], see
also chapter 1 section 1.3.4), supporting the prevailing view that most of the
delamination was completed by 35 Ma.

5.5.2.2. Erosion of Residual Crustal Thickness

The above described surface uplift due to lithosphere delamination from 55 Ma


was likely in part counteracted by later stage reduction of crustal thickness. By 45 Ma,
rocks of the study area which are at the surface in the Columbia Mountains had cooled
from amphibolite metamorphic conditions (as documented by existing U-Pb and Ar-Ar
data (Table 1.3-1.)) to below 300°C [Parrish, 1995; Colpron et al., 2000; Vanderhaeghe
et al., 2003; Cubley et al., 2013], and by 25 Ma most had cooled to less than 100 °C
(this study). According to the exhumation histories presented here (section 5.3.1.)
between 3 and 8 km of rock have been removed by erosion since 45 Ma. Because the
amount of erosion (1 – 2.5 km) since 10 Ma forms a significant part of this, we have to
remove this in order to evaluate the contribution of the first phase. The amount of
erosion between 45 and 10 Ma is therefore of the order of 2-6 km, depending on
locality, notwithstanding the large uncertainties in such estimates.

Therefore gradual isostatically-driven erosion and compensation to elevation


arising from the situation at 45 Ma when the crustal thickness was still some

225 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

kilometres greater than at present would further contribute to exhumation. Thicker


crust along with the hot mantle from delamination, which is supported by a large
volume of Eocene volcanism, would argue for a relatively high elevation of the land
surface. Subsequent crustal thinning from erosion encouraged by high elevation of the
land surface over time would lead to lowering of the average elevation. Although the
synchronous removal of the lithosphere likely counteracted this to some degree by the
estimated ~1300 m of the ~1600 m of current thermal support, the land surface would
still have gradually lowered in response, particularly once delamination slowed at ~35
Ma.

Some further reduction of surface elevation may have arisen from less vigorous
mantle convection activity following cessation of subduction (and delamination) at 35
Ma, the absence of volcanism in the eastern part of the British Columbia Cordillera
prior to middle Miocene time, and the consequent increase in mantle density due to
modest cooling from its former hotter state. A reduction in the erosion rate into a
period of middle Cenozoic quiescence therefore explains the general magnitude of
erosion documented by our results and seems physically reasonable.

5.5.3. Neogene Creation of Modern Topography: 10 – 0 Ma

After a hiatus from 25 to 10 Ma exhumation increased at the end of the


Miocene, incising the modern topography. Similar increases in exhumation have been
recorded for many locations in the northern hemisphere during the Miocene, such as
the Colorado Plateau and Cascades in the US, The Coast Ranges of both British
Columbia as well as Alaska, the Alaskan St Elias Mountains and even the Tibetan
Plateau [Parrish, 1983; Ruddiman and Kutzbach, 1989; Raymo and Ruddiman, 1992;
White et al., 1997; Takeuchi and Larson, 2005]. This apparent synchronicity of many
locations within the Northern Hemisphere has sparked debates on whether these
uplift rates were the result of a cooling Miocene climate, manifesting in increased

226 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

exhumation and resulting isostasy (including post-glacial isostatic rebound) [Molnar


and England, 1990; Herman et al., 2013], or are in fact the cause of climate change
through increasing range elevations and thus alteration of global wind patterns
(e.g.[Ruddiman and Kutzbach, 1989]).

Figure 5.5-2: Exhumation since 10 Ma and summit surface uplift documented in this study and further evidence
from independent studies suggesting surface uplfit. Palynology evidence from [Rouse and Mathews, 1979],
foreland basin cover removal along the foothills from [Wright et al., 1994], climatic oxygen isotope indicators just
south of the US border from [Kohn et al., 2002; Takeuchi and Larson, 2005; Kent-Corson et al., 2006]. Chilcotin
flood basalt extent and elevations from [Mathews, 1989]. See chapter 1 for more detail.

In the neighbouring western Cordillera, the Coast Mountains, increased


exhumation as a result of climate change has been suggested since ~4 Ma, where
widespread onset of mountain glaciation is thought to have caused the intensification
of incision and resulting rock erosion and ~1km of isostatic summit uplift observed
from AFT data [Farley et al., 2001]. However, exhumation along the coast has been

227 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

high since 10 Ma and is connected to up to ~3km of total surface uplift since the
Miocene, which lifted the Coast Mountains to their current height [Parrish, 1983]. This
is generally thought to be associated with arrival of the Juan-de-Fuca subduction zone
during the early Miocene (e.g. Figure 1.2-9.), the migration of the arc westwards prior
to 7 Ma, and associated changes in upper mantle temperatures and composition
above the subducting slab, causing thermally supported (and thus pre-4 Ma climate-
independent) surface uplift.

The results of this study along with further independent evidence presented in
chapter 1 (Figure 5.5-2) have already shown that up to ~1 km of surface uplift of the
summits occurred since the beginning of the Neogene. A similar amount was
suggested by the ~7 Ma Chilcotin basalt outliers at high elevations on the western
flank of the Columbia Mountains, which are today elevated by 750 – 1000m above
their flat-lying counterparts in the Interior Plateau [Mathews, 1989]. The question is,
did the eastern Cordillera experience tectonically or dynamically driven surface uplift,
was the post-Miocene exhumation caused by glacial incision and isostatic
compensation, or was it a combination of both?

5.5.3.1. Neogene Back-arc Configuration

The arrival of the northward migrating Farallon plate on the plate boundary off
the west coast of British Columbia between 15 – 10 Ma once again caused eastward
subduction underneath the Cordillera (Figure 5.5-3 and Figure 1.2-9). Since then,
ongoing subduction of young (<10 Ma old) oceanic crust and associated hydrated
sediments along the coast caused the modern back-arc asthenosphere to be hot and
wet [Hyndman and Lewis, 1999]. This resulted in low viscosity, which in combination
with the downwards stream created by subduction lead to rigorous small scale
convection of the asthenosphere and upwelling of hotter material along the craton
edge, which is likely further increased by edge driven flow [Hyndman and Lewis, 1999;
Hyndman et al., 2005; Currie and Hyndman, 2006; Hyndman, 2010].

228 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

Figure 5.5-3: Schematic illustration of the current back-arc position of the Interior Plateau and the Columbia
Mountains in relation to the Juan de Fuca subduction zone underneath the Coast Mountains.

Current edge and subduction driven flow of hotter lower asthenosphere


material into the delaminated back-arc is well documented not only by the large
difference in surface heat flow from the back-arc (75 ± 15 mW/m2 on average) to the
craton (~40 mW/m2) but also seismic tomography and mantle xenoliths (e.g.
[Hyndman and Lewis, 1999]). As described above this results in back-arc temperatures
at similar depths to be on average ~500 °C higher in comparison to the nearby craton
(Figure 5.5-1 B, from [Hyndman, 2010]). Hyndman also state that the low viscosity and
vigorous small scale convection of mantle material results in a uniform distribution of
temperature, and that the dramatic increase in surface heat flow from the plateau
from 75 ± 15 mW/m2 to locally > 120 mW/m2 near the trench is caused by differences
in lithology and an inherent increase in near-surface radiogenic heat production. A
direct result of this Neogene renewal of hot mantle and increased convection was the
eruption of the Chilcotin flood basalts (section 1.3.2). While the earlier flows date back
from ~16 – 14 Ma, the main volume erupted between 10 – 6 Ma [Mathews, 1989;
Mathews, 1991].

As described in section 5.5.2, complete delamination to the current thickness


considering the current back-arc geothermal state results in ~1600 m of thermally
supported elevation. This is the main reason for the currently relatively high elevation
of the Plateau and the Columbia Mountains. The average Columbia mean elevation is

229 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

similar to the Rocky Mountains, which in contrast are underlain by thick cratonic
lithosphere (Figure 5.5-1 A) and the difference in crustal thickness across the Rocky
Mountain Trench reaches almost 20 km, with a thickness of ~33km in the back-arc and
up to 50 km on the craton [Hyndman and Lewis, 1999]. If both regions were underlain
by a normal corresponding lithospheric thickness, the mean elevation of the back-arc
should be expected to be considerably lower. In fact, the plateau would be ~ 400 m
below sea level [Hyndman et al., 2009]. The Columbia Mountains are elevated above
the plateau in part because they are also supported by 1000 – 1500 m of isostatic rise
in mean elevation as a result of residual thicker crust from the orogeny, which
increases from ~30km in the central Plateau to 40 – 45 km towards the RMT. It is
unclear when the delamination of the back-arc lithosphere was complete and the
surface of the Interior Plateau and the Columbia mountains had been lifted by the full
~1600 m at the present time. It is possible, that the resurgence of subduction
underneath the Coast Mountains after the arrival of the Juan de Fuca plate from the
south ~15 Ma ago caused additional mantle delamination (e.g. [Madsen et al., 2006],
Figure 1.2-9). Based on our previous estimate of ~1.3 km surface uplift during the
Eocene as a result of lithosphere delamination (as documented by plateau incision),
the renewed asthenosphere upwelling could have led to a further ~300 m of surface
uplift of the whole back-arc, contributing to the Neogene exhumation phase by
increasing elevation and thus exhumation. It could even be suggested that during the
phase of no subduction the surface elevation could have lowered slightly and that it
rose again by more than 300m after subduction restarted, as any subsidence resulting
from modest cooling (due to lack of upwelling) would have been reversed by the
reheating described here.

5.5.3.2. Possible Mountain Uplift from Asthenosphere Upwelling

In addition to a general increase in elevation of the entire back-arc, the results


of this study in combination with pre-existing data (Figure 5.5-2) suggest considerable

230 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

uplift of the eastern cordilleran mountains in respect to the Interior Plateau. Because
the increase in exhumation and surface uplift of the mountains and not the plateau
since ~10 Ma pre-dates the onset of mountain glaciation (~3 Ma) and the opening of
the Nootka slab window (~4 Ma) (both to be discussed below), differential surface
uplift caused by the mantle situation is the most likely reason for onset of mountain
exhumation since 10 Ma.

Flow of hotter mantle material into the back-arc is generally thought to be


greatest along the edge of the thick cratonic lithosphere underneath the RMT (Figure
5.5-3) [e.g. Hyndman and Lewis, 1999; Hyndman and Currie, 2011]. Here the steep
gradient of the lithosphere – asthenosphere boundary encourages edge driven flow
(hot material rising up along a step) likely causing the mantle underneath the Trench
and the Columbia Mountains to be hottest and cooling towards the interior plateau
where it is mixed by the inferred small scale convection with slightly cooler material.
This could be supported by the dramatic increase in surface heat flow from the
plateau from 75 ± 15 mW/m2 to locally > 120 mW/m2 near the trench. Although
Hyndman et al. [2005] suggested that most of this increase could be due to differences
in radiogenic heat production of the crust, the large uncertainties when estimating
deep temperatures from heat flow make a local increase in temperature caused by the
focussed upwelling of edge driven flow plausible. This difference in temperature
would cause locally different mantle densities due to thermal expansion and thus
could support more topography above hotter mantle, causing surface uplift and
therefore increased exhumation.

To demonstrate this, thermal expansion of a 1 D mantle column can be


calculated using equation (6)

(6) 𝛥𝐿 = 𝛥𝑇𝐿 ∗ 𝐿0 ∗ 𝐶

where the thermal expansion ΔL of a column of initial thickness L0 can be calculated


from the average temperature increase integrated over L, ΔTL , and the thermal
expansion coefficient C = 3.2x10-5 °C-1 [Hasterok and Chapman, 2007]. Following

231 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

Hyndman [2009], ΔTL can be estimated as ΔT/2, if ΔT is the maximum difference of


temperature between the state at L0 and after heating at ΔL. So, for example, if the
temperature in the upper mantle heats up by 500°C (e.g. as a result of upwelling and
delamination, which is exactly what was described in section 5.5.2.1), then ΔTL=250°C.
For a column thickness L=200km by solving equation (6) we get the suggested 1.6 km
of expansion and consequently uplift or ‘thermal support’.

Hyndman [2009] also stated that the average estimated temperature


difference of craton vs back-arc is ~500°C, however the low horizontal resolution of
seismic tomography surveys allows for considerable variation which is undetected by
the data. Hyndman [2010] suggest that uppermost lithosphere temperatures could
range from 1000 - 1200°C in the Cordilleran back-arc, with uncertainties of 50 – 100°C.
Therefore, if we invoke a modest temperature difference of 200°C of the central
Cordilleran mantle vs. the east nearer the craton edge as a result of the suggested
edge driven flow, equation (6) results in thermal expansion of column L of ~650m of
the eastern part compared to the west. For equally reasonable slightly lower or higher
temperature differences east to west of 100°C and 300°C, thermal expansion is ~320
m and ~960 m respectively.

Heating the base of the crust will also lead to some degree of thermal
expansion of the crust, however depending on the proportion of crust heated and
considering the low thickness of the crust compared to the large mantle column this
contribution would be much smaller (likely <10%). Also, the density of the mantle for
this range of thermal expansion will only vary by ~15kg/m2 and so isostatic adjustment
as a result of modest heating under the cordillera, taking into account the crustal
thickness difference of 7 – 10 km from back-arc to mountains, is less than 40 m and
therefore negligible.

This suggests that a local increase in mantle temperature, caused e.g. by


asthenosphere upwelling and edge driven flow, of between 100°C and 300°C could
result in thermal expansion of the mantle and thus gradual surface uplift of the
mountains in respect to the slightly cooler plateau of 0.3 - 1 km. This, in addition to

232 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

the already elevated surface of the Columbia Mountains due to isostatic compensation
for 7 – 10 km thicker crust, would have caused the mountainous areas to rise several
hundred meters, potentially up to ~1km, after the increase of convection, facilitating
increased exhumation and accelerating rock uplift.

However, these estimates are based on a simple 1D calculation and are not yet
taking the potentially inhibiting effects of the flexural strength of the crust into
account. Although, as described in chapter 1, the effective elastic thickness of the
back-arc is extremely small (only 15-20km, see Figure 1.2-12), it cannot be excluded
that lateral strength of the crust might inhibit surface uplift to some degree and could
significantly decrease these estimates. It is currently highly debated amongst the
geophysicists of the dynamic topography community (dynamic topography =
topography created by dynamically lifting and depressing the earth’s crust through
asthenosphere flow changes) to what extent and over which wavelengths such surface
uplift is possible. While it is generally accepted that uplift in response to
asthenosphere upwelling is controlled by effective elastic thickness and the thermal
state of the mantle underneath and can reach several kilometres for large (>1000km)
wavelengths [Flament et al. 2013], the effects on smaller areas (such as the study area
of this research, which is only ~300km in extent) are barely studied. Generally, most
models published thus far deal with typical oceanic crust as a study medium [e.g.
Winterbourne et al, 2014; Rio and Hernandez, 2004; Wyrtki, 1975], while only few are
occupied with the more complex continental crust scenarios [e.g. Clark et al. 2005;
Lithigow-Bertollini and Silver, 1998]. Because of the thin, hot crust of the Cordilleran
back-arc and the small wavelength of the study area, the oceanic crust scenarios might
be better analogues than a conservative continental crustal stack of >100km thickness.

For these oceanic crust scenarios the more optimistic studies suggest that up
to 1km of surface uplift could occur over a wavelength of only a few hundred km in
situations of low flexural strength and effective elastic thickness below 20km (e.g. in
still hot and weak oceanic crust) [e.g. Braun, 2010], which would concur with the
estimates made by our calculations above. Other studies suggest that these models

233 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

are largely overestimating the flexibility and that no more than average amplitude of
500m of surface uplift over a minimum wavelength of 1km could be expected,
anything higher would be inhibited by lateral crustal strength [e.g. Winterbourne et al,
2014]. This would have consequences for the above made estimates, indicating that
the uplift was largely overestimated and thus the age-data has to be explained in some
other way or that a further undocumented structural component (e.g. minor thus far
undocumented movement along numerous faults which could add up to considerable
exhumation in the central area) would be required to produce the observed data.

However, even a typical oceanic crust might not be the best comparison in the
case of the study area. Because effective elastic thickness and flexural strength
particularly of continental crust are influenced by not only the thermal state of the
crust and lithosphere, but also the depth of structures such as faults and sutures and
the division of the crustal thickness into layers of differing strength [Hyndman et al.
2009], the situation of the study area may be even more complicated. All of the
simplified oceanic crust models mentioned above assume ‘normal’ and uniform, hot,
newly formed and largely undisturbed oceanic crust. However, in the case of the
Cordilleran back arc, although displaying extremely high heat flow values similar to
oceanic crust, the tectonic history of the area could further reduce the flexural
strength. Being a complex puzzle of accreted blocks and intrusions with numerous
deep crustal faults and sutures potentially reaching the lithosphere it has been
suggested that the flexural strength of this area could be extremely low with only 10-
20 N/m*1012 instead of the cratonic 120-200 N/m*1012 to the east of the RMT
[Hyndman et al. 2005]. Hyndman et al. (2005) also suggested that the inhomogeneous
nature of the Cordilleran back-arc crust requires several layers of different strength
and possibly enables detachment of these layers, offering flexibility and causing the
extremely weak lower crust which they inferred from their results. Unfortunately
there are currently no suitable analogue models or reliable parameters for comparison
in existence which investigate the flexural strength of such a complex crustal scenario.
Thus we are at this point not able to make any better comments beyond
acknowledging that some dampening effect of several hundred meters on potential

234 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

uplift may exist due to flexural strength but also that it may be negligible, as suggested
by other research in the area [e.g. Hyndman et al. 2009]. A detailed examination
involving complex crustal modelling would be necessary to determine this, which is
well beyond the scope of this study and potentially beyond the current state of
ongoing research into crustal properties and dynamic behaviour altogether.

For now, making the rather bold assumption that flexural strength can be
neglected (or is nearly negligible), this surface uplift, although likely mainly affecting
the areas west of the RMT, could have influenced exhumation of the Rocky Mountains
to some degree as well which could explain why we observe increased exhumation
since 10 Ma on both sides of the trench. As described in chapter one the Rocky
Mountain Trench is thought to ‘roughly’ mark the edge of the craton; however the
exact position and nature of this transition, i.e. the angle of the craton edge and the
thermal consequences of temperature changes for the crust directly above, remain
debated [Clowes et al., 1995; Zelt and White, 1995; Hyndman and Lewis, 1999]. It is
possible that some degree of thermal or ‘dynamic’ uplift caused by edge driven flow
could have extended a little further to the east than the trench, as for example the
estimated Moho depth ~50 km east of the trench varies between 40 and 44 km and
full lithospheric thickness is only expected 100km even further to the east (below the
foothills) [Zelt and White, 1995; Hyndman and Lewis, 1999; Burianyk et al., 1997].
However, without a better understanding of this transition any attempt to quantify
this would be speculation. Furthermore both Currie et al. [2009] and Hardebol et al.
[2011] suggested that edge driven flow underneath the trench could be causing
erosion of the thick cratonic lithosphere. Thinning of the lithospheric root of the thick
crust of the Rocky Mountains could also cause surface uplift to transgress further to
the east than the trench. However, any uplift of even slightly thicker and more rigid
continental crust to the east of the RMT would be significantly inhibited by greater
effective elastic thickness and flexural strength of this medium. In order to make more
precise estimates both the transition of crustal thickness underneath the RMT as well
as the possibilities of dynamic topography warrant further investigation, which is well

235 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

beyond the scope of this thesis. Nevertheless it has to be kept in mind when making
these interpretations.

5.5.3.3. Local Maximum Pattern: Relation to Slab Window?

The analysis of lateral age distribution and exhumation has shown that a centre
of youngest ages and thus greatest exhumation is located in the northern Selkirk
Mountains. Because this is located near the central axis of the range and near (albeit
not directly beneath) the highest topography, this could simply be attributed to a
centre of erosional force, e.g. past mountain glaciation. However, as described in
chapter 1 (section 1.2.4), the nearby Wells-Gray volcanic field (3.8 Ma to present) may
result from localized upwelling of asthenosphere material through the vertical slab
window formed by the Nootka transform fault (Figure 1.2-10 and Figure 5.5-4), which
separates the shallow Explorer Plate (unknown dip) from the steeper (45 – 65 °) Juan
de Fuca Plate since ~4 Ma [Govers and Meijer, 2001; Madsen et al., 2006].
Alternatively, it has been suggested that, although less likely, the basalt field could be
the easternmost expression of the Anahim hot-spot [Rogers and Souther, 1983; Hickson
and Souther, 1984; Bevier, 1989]. Either way, it seems plausible that a more focussed
upward flow of hotter material amongst the generally high temperature and low-
viscosity mantle could be responsible for the locally variable exhumation pattern.

The solutions to equation (6) for a locally increased mantle temperature have
already shown that even a modest increase of 100 – 300 °C amongst already hot
mantle material can cause significant thermal expansion of up to ~1km and
subsequent surface uplift. Furthermore the extremely low effective elastic thickness of
15-20km of the study area [Currie and Hyndman, 2006] could allow a pattern of much
smaller wavelength than what is generally observed in areas of higher flexural strength
by studies investigating dynamic topography (as previously discussed in section
5.5.3.2). Therefore, a further localized increase in temperature caused by a spatially
limited upward flow of ~100°C could have led to a local increase in uplift and

236 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

Figure 5.5-4: Inferred Plate tectonic situation [Madsen et al., 2006] overlain by exhumation distribution of this
study and the centre of youngest ages of chapter 3 indicated in orange.

subsequent exhumation of several hundred meters (although for such small


wavelengths there could be some effect of flexural strength of the crust, see previous
section), which is enough to produce a local drop in observed cooling ages in the order
of 5-15 Ma based on the age elevation relationships observed in chapter 3. If true, it
could be suggested that this should be observed further to the west as well and thus
the pattern of exhumation should extend into the plateau, following the trace of the
Nootka fault. However, the interpreted slab angles of 30° for the Explorer Plate and
45° - 65° of the Juan de Fuca plate [Govers and Meijer, 2001; Bostock et al. 2002],
which is steepening towards the east, will likely reduce the aperture of the vertical
slab window in the west, limiting the amount of upwelling.

237 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

5.5.3.4. Glacial Erosion, Isostasy and Summit Uplift

Following the surface uplift caused by upwelling of asthenosphere from ~10


Ma and possibly a further localised pattern commencing ~4 Ma from the Nootka fault,
onset of glaciation would have further increased exhumation of the already elevated
regions both to the east and west of the RMT.

The earliest evidence in southern British Columbia suggests onset of


widespread mountain glaciation in the eastern Cordillera at ~3 Ma. Since then the
mountainous areas have been almost permanently glaciated (although with varying ice
extent) and periodically the ice sheet extended to the Interior Plateau, where the
Cordilleran ice sheet covered most of British Columbia. Although significant evidence
for this in the form of glacial deposits is only preserved for the last two glaciations
(~40,000 and 14,000 – 11,000 yrs), it has been suggested that several more episodes
of ice sheet growth and decay could have occurred since the beginning of the
Pleistocene [Souther et al., 1984; Naeser et al., 1982; Mathews and Rouse, 1986].

The onset of mountain glaciation would have been facilitated not only by
changing climate but also because at 3 Ma the elevation of both the Rockies (created
by the initial orogeny) and the Columbia Mountains (increased and sustained by the
back-arc state and upwelling) was already high. Although valley incision likely started
in response to surface uplift since 10 Ma, a change in erosional force from
predominantly fluvial to glacial will have increased incision of valleys after 3 Ma.
Section 5.4.4 demonstrated how the total Neogene creation of relief has caused ~1km
of summit uplift from the increase in relief. This relative uplift of the maximum
elevation in relation to the mean elevation would have therefore increased and
reached its current configuration over the past 3 million years.

Post-glacial isostatic rebound can further contribute to surface uplift of


previously glaciated regions. However, although the full extent of the Cordilleran ice
sheet has been well mapped and is known to have periodically covered all of the Coast
Ranges as well as the eastern Cordillera and the Interior Plateau at its largest, its exact

238 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

thickness estimates are at best speculative and thus the amount of isostatic rebound
also remains poorly known ([Clague, 1983; Hutchinson, 1992] both in [Clague and
James, 2002], [James et al., 2000]). While most minimum estimates for the total
isostatic depression range from 200-300 m in southern BC [Mathews et al., 1970;
Hutchinson, 1992], Clague (2002) suggested that isostatic depression of maximum
500m occurred in the Interior Plateau, where the ice thickness was greatest during
peak glacial phases. In the mountains where the ice originated and was present for
longer periods, isostatic depression and subsequent rebound after ice retreat would
have been somewhat less (< 400 m). Rebound would have occurred rapidly, as studies
have shown that most of the rebound since the last glaciation (14,000-11,000 yrs)
occurred within a few thousand years and was almost complete by the end of the
Pleistocene or early Holocene. The modern remnant of post-glacial rebound barely
exceeds 3-5 mm/a in the interior plateau and is as low as 1-3 mm/a in the Coast
Mountains [Tushingham and Peltier, 1991; Clague and James, 2002], a figure that is
likely similar to what could be expected in the eastern Cordillera. Therefore, post-
glacial rebound could have contributed to the recent high exhumation and incision of
valleys of the already elevated mountains by up to ~400m.

5.6. Conclusions

This chapter has demonstrated that the south eastern Canadian Cordillera has
experienced two main exhumation phases since the end of the collisional orogeny (60
Ma) which formed the framework of these mountains. Post-orogenic exhumation
rates of 0.12 - 0.77 km/m.y. were found at all profiles until 35-25 Ma. This was
followed by relatively low exhumation (0.01 to 0.27 km/m.y.) until the mid-Miocene
(12-7 Ma), when another increase in exhumation of up to 0.32 km/m.y. incised the
present valleys and thus created the current surface topography.

239 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

The first phase of exhumation can be attributed to three overlapping geological


processes: orogenic thrusting at the very beginning (70 – 60 Ma), thinning of the crust
with exhumation of lower crustal rocks during and after the extensional phase (60 – 45
Ma), and delamination of the back-arc lithosphere through asthenospheric upwelling
(55 – 35 Ma). At ~45 Ma both the Rocky and Columbia Mountains most likely
presented as a high-elevation low-relief plateau with mean elevations approximately
0.4 – 1 km higher than present day, between 2,400 – 2,800 m a.s.l (assuming a similar
thermal state of the mantle). While it is difficult to quantify the contribution of each
process, since they all result in rock uplift and increased erosion and they all
somewhat overlap, it was suggested that ~1.3 km of surface uplift occurred as a result
of delamination in both the Interior Plateau and the Columbia Mountains.
Subsequently gradual exhumation until the Oligocene in combination with cooling of
the upper mantle after ~35 Ma would have lowered the mean elevation over time and
the low relief likely persisted during this phase.

The later Neogene phase of increased exhumation was also found to likely be
the result of a combination of processes. At 10 Ma increased subduction and
asthenosphere upwelling along the craton edge possibly led to thermal expansion of
the mantle underneath the eastern Cordillera, raising the ranges on both sides of the
RMT by a further 0.6 – 1 km, thus renewing exhumation of the eastern Cordillera and
accelerating isostatically driven rock uplift in response. The opening of the Nootka slab
window was suggested to be the cause of the local maximum of exhumation in the
northern Selkirk Mountains, which since ~4 Ma could have a similar thermally
supported effect on a more local scale and could have contributed several hundred
meters to the total uplift in the Columbia Mountains. Post-glacial isostatic rebound
and potential late stage lithosphere delamination of the back-arc possibly also
contributed to up to ~600 m in surface uplift of both the plateau and the mountains,
while increased incision and creation of the modern relief from glaciation since ~3 Ma
lifted the summits isostatically a further ~1 km while reducing the mean elevation of
the time by 0.05 – 0.4 km as a result to the total volume of rock removed. The
combination of these processes adds up to an overall rise of the highest elevation by

240 | P a g e
Cenozoic Exhumation, Elevation and Uplift Chapter 5

2.0 – 2.5 km since 10 Ma, which has caused the current elevation and relief of the
Columbia and Rocky Mountains.

241 | P a g e
Synthesis and Conclusions Chapter 6

Chapter 6:
Synthesis and Conclusions

6.1. Introduction: Aim of this Work

The North American Cordillera is a chain of mountains which span the entire
length of the west coast of North America north to south, from Alaska to Mexico. In
western Canada, the Cordillera comprises two parallel mountain chains separated by a
high elevation (~1100m) intermontane plateau. The Cenozoic cooling and exhumation
history of the western range, the Coast Mountains, has been well studied [e.g. Parrish,
1983; O'Sullivan and Parrish, 1995; Rohr and Currie, 1997; Farley et al., 2001], while
the Cenozoic history of the eastern Cordillera remained poorly constrained. The aim of
this work was to address this problem by providing a new dataset of low-temperature
thermochronology ages for the south eastern Canadian Cordillera and use it to
constrain their Cenozoic cooling, exhumation and elevation history.

The following research questions were addressed:

 Is there a systematic pattern of cooling ages that records the Cenozoic


cooling history of the south eastern Canadian Cordillera?
 What is the spatial and temporal distribution of exhumation across the
south eastern Cordillera?
 To what degree can these patterns be related to increases in paleo-relief
and potential surface uplift?
 What are the underlying geological mechanisms causing the exhumation
and surface uplift?

242 | P a g e
Synthesis and Conclusions Chapter 6

6.2. Synthesis of Results and Main Conclusions

This thesis has presented a new, large scale low temperature


Thermochronology dataset for the south eastern Canadian Cordillera. Apatite (U-
Th)/He (AHe), apatite fission track (AFT) and zircon (U-Th)/He (ZHe) ages were
obtained for 58 samples from the Columbia Mountains to the west, and the Rocky
Mountains to the east of the Rocky Mountain Trench (RMT). Samples were taken
along several horizontal and near-vertical profiles to investigate age patterns in both
dimensions.

For the vertical transects age-elevation relationships were established and the
thermal histories of numerous samples modelled using HeFty (chapters 3 and 4)
[Ketcham, 2005]. On both sides of the RMT rapid cooling (>10°C/m.y.) from peak
metamorphic temperatures (>500°C) below ZHe and AFT closure temperatures (Tc
~180° and ~120°C) was recorded during the Cretaceous to Eocene (75-40 Ma) followed
by a period of slow cooling (<1°C/m.y.), while a later phase of rapid cooling
(>10°C/m.y.) occurred below AHe closure (Tc ~70°C) since the Miocene (10-0 Ma) [Tc
from Reiners et al., 2005]. Ages increased with elevation and laterally with increasing
distance to the RMT, where the highest peaks cause the greatest relief and thus the
youngest ages are exposed at low elevations. A further distinct horizontal pattern in
cooling ages was observed to the west of the RMT, in the Columbia Mountains, where
the low elevation ages not only increased with distance to the trench, but also radially
away from the northern Selkirk Mountains (~ 52°N, 118°W). East of the trench low
data coverage made similar observations impossible.

Forward modelling of predicted vs observed age elevation relationships


(chapter 5) showed that three main exhumation phases occurred since the end of the
Laramide orogeny (60 Ma). Post-orogenic exhumation rates of 0.12 to 0.77 km/m.y.
were found, which were followed by a ‘hiatus’ or relatively low exhumation (0.01 to
0.27 km/m.y.) until the mid-Miocene (12-7 Ma). A maximum of ~10 and 7 km of cover
was removed from the Columbia and Rocky Mountains respectively since the end of

243 | P a g e
Synthesis and Conclusions Chapter 6

the orogeny. Subsequently, another increase in exhumation (up to 0.32 km/m.y.) was
found to have incised the present valleys and thus created the current surface
topography. The incision of relief over the past 10 Ma is the result of a differentiated
pattern of exhumation, where current peaks show very little removal of cover, and
valley samples indicate up to ~2.5 km of incision. Similar spatial variations were
observed from exhumation values, with the highest amount of recent (Neogene)
erosion located between the northern Selkirk and Monashee Mountains.

Estimation of the paleo-elevation and –relief showed, that at ~45 Ma both the
Rocky and Columbia Mountains most likely presented as a hilly, high elevation plateau
with mean elevations approximately 0.4 – 1 km higher than present day, between
2400 – 2800 m a.s.l.. Some major present-day valleys might have existed since the
Eocene, such as the Rocky Mountain Trench, but overall the relief was less
pronounced. Gradual exhumation until the Oligocene lowered the mean elevation
over time, and the low relief likely persisted until the Neogene. Increased localized
incision during the last exhumation phase incised the current deep valleys and
resulted in peak surface uplift of up to ~1km. A further isostatic reduction of mean
elevation of 0.05 – 0.4 km was found as a result to the total volume of rock removed.

More than one cause could be inferred for each of the two rapid exhumation
phases, making identification of individual contributions difficult. The first phase of
exhumation was attributed to three overlapping geological processes: orogenic
thrusting at the beginning (70 – 60 Ma), thinning of the crust and exhumation of lower
crustal rocks during the extensional phase (60 – 45 Ma) and delamination of the back-
arc lithosphere through asthenosphere upwelling (55 – 35 Ma). Up to ~1.3 km of
surface uplift occurred as a result of delamination in both the Interior Plateau and the
Columbia Mountains.

The increased exhumation during the Neogene was interpreted to be the result
of thermally supported uplift due to further lithosphere delamination after renewed
subduction, aided by a combination of post-glacial isostatic rebound and isostatic
compensation to incision after the onset of mountain glaciation. A possible

244 | P a g e
Synthesis and Conclusions Chapter 6

contribution of localized asthenosphere upwelling resulting from either the slab


window associated with the Nootka-transform fault or the Anahim hotspot was
discussed as the cause of the localized pattern of youngest ages and highest
exhumation in the Northern Selkirk/Monashee Mountains. Combining all of these
contributions it is likely that the maximum elevations of the Columbia Mountains
increased by up to ~2.5 km over the past 10 m.y..

In conclusion, this work has constrained the previously unknown post-orogenic


Cenozoic exhumation history of the south eastern Cordillera using low-temperature
thermochronology and forward modelling of exhumation histories. The results
described their Cenozoic evolution from a high elevation (~2.5km) but low relief
plateau, which gradually lowered in mean elevation by up to 1.5 km until a later
Neogene increase lifted the peaks a further ~1.2km to their current height. Although
the timing of major exhumation phases is similar to what is observed in the Coast
Mountains (60-45 and 10 Ma to present), the elevation in the eastern Cordillera has
been continuously high. It was always above 1000-1500m and at times reaching
maximum elevations likely >3000m since 60 Ma, whereas the Coast Mountains only
rose from ~750m to their current height since ~10 Ma. This information is important
for modern climate studies [e.g. Foster et al. 2010], who thus far only had elevation
data from the Coast Mountains to infer the height of the Cordillera as a major climate
barrier of the past.

6.3. Future Work

There are several ways in which this work could and will be continued. This study
was designed to cover a large area and thus could inherently not provide a very
detailed data coverage everywhere. Furthermore it proved particularly difficult to
obtain apatite from many samples because only ~60% of the almost 80 painstakingly

245 | P a g e
Synthesis and Conclusions Chapter 6

separated samples yielded sufficient crystals of good quality to be usable. This


interfered with the sampling strategy by creating gaps in the intended transects.
Luckily, some additional data could be obtained from my Co-author B. Guest
(University of Calgary), who is currently preparing a proposal to build on our now
combined dataset to further increase the data coverage, particularly within the Rocky
Mountains in the east. This proposed study intends to sample in greater detail along
several E-W transects across both the Rocky Mountains and the Columbia Mountains
to improve our understanding of how the observed trends and age patterns continue
to the east of the Rocky Mountain Trench which separates the two ranges. Because of
the closer location of their laboratory to the field area they will not only be able to
collect more samples, but also be able to sample much larger quantities and thus
hopefully improve on the mineral yield, minimizing the issues experienced during this
project.

Furthermore, this project initially intended to also produce exhumation histories


for the northern part of the eastern Cordillera in British Columbia, the northern Rocky
Mountains and Omineca Mountains. However, difficulties with accessibility during
field work and sample yield meant that ages for only 3 samples for the northern part
could be obtained. All ages were Eocene, ranging from ~37 [±2.2] Ma (AFT age at
1900m) to 48 [± 7.6] Ma (AHe at 800 m a.s.l.). Because these ages are from central
samples near the Rocky Mountain Trench and considerably older than their
counterparts in the south (as shown by this thesis), this could indicate a slightly
different exhumation history in the north. A further study could be undertaken to
constrain this.

The elevation histories inferred by the results of this thesis could also be used by
future climate models to improve our understanding of the impact of the North
American Cordillera as a barrier to major wind systems for the North American
continent. Although the results show that the post 10Ma history is similar in timing
and elevation to that of the Coast Mountains, which has been used in recent climate
models to evaluate the influence of the Canadian Cordillera on the Northern

246 | P a g e
Synthesis and Conclusions Chapter 6

Hemisphere climate [Foster et al, 2010], the existence of higher elevations (>1-1.5km)
in the east throughout the Cenozoic has thus far been neglected. For climate
simulations reaching back further than the Neogene this data might therefore be of
interest.

Lastly, during the later stage of this PhD a Matlab code was developed by my co-
author Andy Smye (University of Oxford), which was designed to replace the manual
forward modelling undertaken to find suitable exhumation histories. Its principle is the
same as the one applied in chapter 5, however it will have two main advantages (once
complete): Speed and geothermal variability with time. It allows for faster testing of
different exhumation histories as samples do not have to be modelled individually,
instead whole age-elevation relationships can be predicted as one. Secondly,
geothermal gradients can be varied through time either by changing constraints such
as basal heat flow or crustal thickness, or as a result of exhumation rates observed on
the surface. Currently the model does not yet contain suitable AFt kinetics, which is
why only He ages can be modelled at this point. Our plan is to complete this program
and apply it to this dataset and future data as well. Although the overall exhumation
pattern presented here will remain the same, this should allow further narrowing of
possible exhumation histories by reducing noise resulting from geotherm changes
between predicted and observed ages. Assessing the variability of the geothermal
gradient with time should also allow testing of the individual contributions related to
different geological and tectonic processes.

247 | P a g e
Appendices
Appendix 1: Sample list

This list contains all samples presented in this thesis with their geographical locations and brief description of rock type and/or sample locality. The samples
are ordered alphabetically for easier access:

Elevation
Sample Coordinates (decimal minutes) Type Method Description
(m)

11-AS-01 50°50'25.58"N 116°20'50.64"W 929 bedrock AHe Quartzite on road to Bugaboo parking lot
11-AS-09 50°38'6.83"N 117°55'40.98"W 454 bedrock AHe Shelter Bay gneiss/monzonite
11-AS-11 51° 4'20.68"N 118°10'26.04"W 581 bedrock AHe, ZHe Pegmatitic granite' 5km north of Revelstoke
11-AS-12 52° 5'23.89"N 118°32'15.97"W 780 bedrock AFT, ZHe Muscovite rich granite, west shore KBL near Mica Creek Dam
11-AS-13 51°48'45.43"N 118°38'49.96"W 601 bedrock AHe Gneiss, 35km south of Mica Cr. Dam, east shore
11-AS-15 51°17'38.54"N 118°16'20.39"W 605 bedrock AHe Quartzite 35 km North of Revelstoke into Mica Cr.
11-AS-22 52°44'40.56"N 119° 5'47.26"W 791 bedrock AHe Schist 12km south of northend Kinbasket Lake, west shore
11-AS-26 53° 3'36.54"N 119°36'50.04"W 741 bedrock AHe Schist from volcanic succession
12-AS-01 51°27.160'N 116°17.974'W 1649 bedrock AHe Gog Quartzite on road near border of Yoho National Park
12-AS-05 50°55.730'N 118°23.387'W 1471 bedrock AFT Nepheline Syenite near top of powerline maintenance road
12-AS-06 50°56.197'N 118°24.448'W 942 bedrock AHe, AFT Grt-Bt-Paragneiss, halfway up powerline maintenance road
12-AS-08 51°23.774'N 118°40.370'W 1100 bedrock AHe, AFT Paragneiss up towards Petticreek Pass
12-AS-09 51°22.103'N 118°42.377'W 802 bedrock AHe, AFT, ZHe Ultramylonitic paragneiss, bottom of Petticreek Pass road
12-AS-10 51°24.174'N 118°38.257'W 1490 bedrock AHe, AFT Orthogneiss top of Petticreek Pass road

248 | P a g e
12-AS-11 50°44.322'N 116°45.570'W 1956 bedrock AHe, AFT Bugaboo granite
12-AS-12 50°47.312'N 116°41.272'W 1449 bedrock AHe Bugaboo Falls Grit
12-AS-13 51°00.646'N 118°11.193'W 857 bedrock AHe, AFT Devonian granitic gneiss, Mt Revelstoke
12-AS-14 51°01.168'N 118°10.543'W 1234 bedrock AHe, AFT, ZHe Granitic gneiss, medium elev. Mt Revelstoke
12-AS-15 51°01.309'N 118°09.200'W 1527 bedrock AHe, AFT Granitic gneiss, Mt Revelstoke, high elevation
12-AS-16 51°02.348'N 118°08.869'W 1843 bedrock AHe, AFT, ZHe Granitic gneiss, summit of Mt Revelstoke
12-AS-17 52°33.223'N 119°02.656'W 1740 bedrock AHe, ZHe Granitic gneiss, top of Clemina Creek Forest Service Road
12-AS-18 52°33.337'N 119°04.176'W 1278 bedrock AHe Massive granitic gneiss, halfway up Clemina Dora FSR
12-AS-19 52°37.027'N 119°07.979'W 875 bedrock AHe, AFT, ZHe Malton Gneiss, foot of Canoe Mountain
12-RP-AS-4 49°13.246'N 116°36.103'W 562 bedrock AHe, AFT Granite, east shore of south end Kootenay Lake
12-RP-AS-7 49°48.003'N 117°26.261'W 1262 bedrock AHe, AFT Springer Creek Road foliated granite (Nelson Batholith)
12-RP-AS-8 49°17.144'N 117°39.951'W 632 bedrock AHe, AFT Castlegar gneiss south of Castlegar
80-RP-K3 50°18'00''N 117°40'00''W 762 bedrock AFT Kuskanak Creek
80-RP-R1b 51°00'15''N 118°14'40''W 457 bedrock AFT Revelstoke
80-RP-R2 50°58'00''N 118°22'30''W 549 bedrock AFT Eagle Pass
80-RP-R3 50°57'40''N 118°44'10''W 396 bedrock AFT near Malakwa
80-RP-R6 50°45'30''N 119°20'50''W 375 bedrock AFT Tappen
80-RP-R9 51°02'45''N 117°57'00''W 610 bedrock AFT near Albert Canyon
AH47-1 52° 1'40.54"N 119°19'57.58"W 716 bedrock AHe, ZHe Data owned and analyzed by B. Guest/A. Halverston
AH48-1 52° 2'8.54"N 119°18'25.91"W 1234 bedrock AHe, ZHe Data owned and analyzed by B. Guest/A. Halverston
AH48-2 52° 2'52.14"N 119°16'50.10"W 1478 bedrock AHe, ZHe Data owned and analyzed by B. Guest/A. Halverston
AH49-1 52° 2'39.50"N 119°14'32.51"W 2347 bedrock AHe, ZHe Data owned and analyzed by B. Guest/A. Halverston
AH49-3 52° 2'43.88"N 119°15'52.60"W 1737 bedrock AHe, ZHe Data owned and analyzed by B. Guest/A. Halverston
GF 31 53°55'47.58"N 122°39'53.82"W 564 sediment AFT Fraser river sediment, before confluence with Nechacko
GF 35 53°29'40.68"N 120°36'21.60"W 722 sediment AFT Goat river sediment
GF 37 53°15'5.94"N 120° 1'43.92"W 706 sediment AFT Holmes river sediment
GF 39 52°57'51.96"N 119°28'16.44"W 751 sediment AFT Tete Creek, garnet rich sands
GF 49 53°22'35.76"N 117°41'56.34"W 969 sediment AFT Athabasca river sediment at Hinton

249 | P a g e
GF 52 53° 2'24.84"N 118° 5'23.16"W 1004 sediment AFT Athabasca river sediment at Jasper
GF 55 51°25'25.86"N 116°10'53.16"W 1563 sediment AFT Bow river at Lake Louise, confluence with Pipestone river
GF 61 51°29'4.92"N 117°10'57.54"W 791 sediment AFT Columbia river sediment near Donald station
GF 63 51°23'0.78"N 117°27'2.04"W 839 sediment AFT Beaver river sediment
GF 66 51° 0'19.44"N 118°13'12.90"W 430 sediment AFT Bigeddy at Revelstoke, sediment
GF 67 50°56'53.40"N 118°46'24.42"W 375 sediment AFT Eagle River sediment
Ice River 51°7'59.52"N 116°26'4.02"W 1430 bedrock AHe Ice River Complex Syenite
J91-1 52°55'8.33"N 118° 0'41.29"W 1078 bedrock AHe, ZHe Data owned and analyzed by B. Guest/D. Stockli
J91-18 53°13'46.52"N 117°29'45.13"W 1369 bedrock AHe, ZHe Data owned and analyzed by B. Guest/D. Stockli
J91-2 52°52'47.32"N 118°26'13.88"W 1157 bedrock AHe, ZHe Data owned and analyzed by B. Guest/D. Stockli
JC-16-03 51°19'56.42"N 118°35'7.28"W 1860 bedrock AHe Mafic dyke - intruded basement
JC-311-94 51°19'33.45"N 118°33'37.30"W 2320 bedrock AHe Hornblende orthogneiss - basement
JC-321-94 51°22'31.59"N 118°45'35.77"W 1030 bedrock AHe Amphibolitic gneiss - unit 4 in cover sequence
JC-60C-94 51°23'17.02"N 118°35'47.11"W 2220 bedrock AHe, AFT Granitic gneiss - basement
NMI37-01 52°49'34.66"N 118° 7'51.43"W 2454 bedrock AHe, ZHe Data owned and analyzed by B. Guest/N. Miles
NMI37-02 52°49'43.50"N 118° 7'35.51"W 2363 bedrock AHe, ZHe Data owned and analyzed by B. Guest/N. Miles
NMI37-03 52°49'57.01"N 118° 7'58.55"W 2128 bedrock AHe, ZHe Data owned and analyzed by B. Guest/N. Miles
NMI37-04 52°50'13.99"N 118° 7'56.99"W 1996 bedrock AHe, ZHe Data owned and analyzed by B. Guest/N. Miles
NMI38-05 52°50'14.47"N 118° 8'6.61"W 1772 bedrock AHe, ZHe Data owned and analyzed by B. Guest/N. Miles
NMI39-06 52°50'41.28"N 118° 7'29.28"W 1706 bedrock AHe, ZHe Data owned and analyzed by B. Guest/N. Miles
NMI39-07 52°50'33.36"N 118° 6'35.28"W 1491 bedrock AHe, ZHe Data owned and analyzed by B. Guest/N. Miles
NMI39-08 52°51'5.04"N 118° 7'4.80"W 1297 bedrock AHe, ZHe Data owned and analyzed by B. Guest/N. Miles
PCA-121-84 49°47'27.89"N 117°38'44.74"W 2300 bedrock AFT Valhalla Complex granite/pegmatite
PCA-200-83 52°28'57.50"N 118°53'21.70"W 2297 bedrock AFT Malton gneiss, granitic gneiss
PCA-304-83 51°19'40.80"N 118°35'9.60"W 1885 bedrock ZHe Monashee complex leucogneiss
PCA-307-83 51°53'51.00"N 118° 7'58.80"W 3100 bedrock AFT Trident Nepheline syenite, gneiss
PCA-529-83 49°46'54.12"N 117°28'15.24"W 550 bedrock AFT Valhalla Complex augen gneiss
PCA-58-84 49°19'44.86"N 117°41'33.41"W 440 bedrock AFT Southern Valhalla Complex

250 | P a g e
As described in the main body of this thesis some samples and in some cases analysed data were provided by third parties, which is indicated by the sample
names as follows:

##-AS-## Samples collected and analysed by the author of this work (A.Szameitat)
##-RP-AS-## Samples collected by R. Parrish and analysed by the author of this work (A.Szameitat)
80-RP -## Samples collected and analysed by R. Parrish
AH##-# Samples collected and analysed by A.Halverston, data provided by B.Guest
GF-## Samples collected and analysed by G. Foster
J-91 # Samples collected by B.Guest and analysed by D.Stockli
JC-##-# Samples collected and separated by Jim Crawley, analysed by A. Szameitat
NMI##-# Samples collected and analysed by N. Miles, data provided by B.Guest
PCA-###-## and Icer River Samples collected by R. Parrish and analysed by the author of this work (A.Szameitat)

251 | P a g e
Appendices

Appendix 2: Sample images

All images of crystals picked including measurements are provided on the accompanying CD.

Appendix 3: Data
A3.1 eU plots
A3.1.1 Chapter 3

The eU plots for all samples are shown below, split into multiple graphs for clarity.

252 | P a g e
Appendices

AHe eU correlation trendlines


100.0
Age (Ma)

10.0

11-AS-1
11-AS-9
11-AS-11
11-AS-13
11-AS-15
11-AS-26

1.0
1.0 10.0 100.0
eU (ppm)

253 | P a g e
Appendices

AHe eU correlation trendlines


100.0

11-AS-22
12-AS-06
12-AS-08
12-AS-09
12-AS-10
12-AS-11
12-AS-12
12-AS-13
12-AS-14
Age (Ma)

10.0

1.0
1.0 10.0 100.0 1000.0
eU (ppm)

254 | P a g e
Appendices

AHe eU correlation trendlines


100.0

10.0
Age (Ma)

12-AS-15
12-AS-16
12-AS-17
1.0
12-AS-18
12-AS-19
12-AS-1a
12-AS-26
12-AS-27
12-RP-AS-04
12-RP-AS-07

0.1
1.0 10.0 100.0 1000.0
eU (ppm)

255 | P a g e
Appendices

AHe eU positive trendlines


1000.0

12-RP-AS-08
Ice River
JC-16-03
JC-311-94
JC-321-94
JC-60C-94
PCA-200-83
100.0
PCA-304
AH48-2
AH49-3
Age (Ma)

AH48-1
AH47-1
AH49-1

10.0

1.0
1.0 10.0 100.0
eU (ppm)

256 | P a g e
Appendices

ZHe eU correlations
1000.0
11-AS-11
11-AS-12
12-AS-09
12-AS-14
12-AS-16
12-AS-17
12-AS-19
PCA-304-83
zAH47-1
zAH48-1
zAH49-1
Age (Ma)

zAH49-2
100.0

10.0
10.0 100.0 1000.0 10000.0
eU (ppm)

257 | P a g e
Appendices

A3.1.2 Chapter 4

AHe eU correlations
160.0

140.0
11-AS-22
12-AS-1a

120.0 J91-1
J91-2
J91-18
100.0 NMI39-07
NMI39-06
Age (Ma)

NMI38-05
80.0 NMI37-04
NMI37-02
NMI37-01
60.0

40.0

20.0

0.0
0.0 50.0 100.0 150.0 200.0
eU (ppm)

258 | P a g e
Appendices

ZHe eU correlations
70.0 zNMI38-05
zNMI37-01
65.0 zNMI37-03
zNMI37-02

60.0 zNMI39-07
zNMI39-06

55.0 zNMI37-04
zNMI39-08
Age (Ma)

50.0

45.0

40.0

35.0

30.0
0.0 200.0 400.0 600.0 800.0 1000.0 1200.0 1400.0
eU (ppm)

259 | P a g e
Appendices

A3.2 Radial plots

A3.2.1 Radial plots of bedrock AFT data

260 | P a g e
Appendices

261 | P a g e
Appendices

A3.2.2 Radial plots of modern detrital AFT data of both chapter 3 and chapter 4

262 | P a g e
Appendices

263 | P a g e
Appendices

264 | P a g e
References

References
Andrews, G.D., et al, The thickness of Neogene and Quaternary cover across the central Interior Plateau,
British Columbia: analysis of water-well drill records and implications, Canadian Journal of Earth
Sciences, 48(6), 973-986, 2011.

Andrews, G.D., et al, Pleistocene reversal of the Fraser River, British Columbia, Geology, 40(2), 111-114,
2012.

Ault, A.K. and Flowers, R.M., Is apatite U–Th zonation information necessary for accurate interpretation
of apatite (U–Th)/He thermochronometry data? Geochimica et Cosmochimica Acta, 79 60-78, 2012.

Bao, X., Eaton, D.W. and Guest, B., Plateau uplift in western Canada caused by lithospheric delamination
along a craton edge, Nature Geoscience, 7(11), 830-833, 2014.

Barbarand, J., et al, Compositional and structural control of fission-track annealing in apatite, Chemical
Geology, 198(1), 107-137, 2003.

Bevier, M.L., A lead and strontium isotopic study of the Anahim volcanic belt, British Columbia:
Additional evidence for widespread suboceanic mantle beneath western North America, Geological
Society of America Bulletin, 101(7), 973-981, 1989.

Bigazzi, G., The problem of the decay constant lambda f of 238U, Nucl. Tracks, 5 35-44, 1981.

Boesen, R. and Joplin, G.A., The form of the intrusive complex at Mount Dromedary, New South Wales,
Journal of the Geological Society of Australia, 19(3), 345-349, 1972.

Brandon, M.T., Roden-Tice, M.K. and Garver, J.I., Late Cenozoic exhumation of the Cascadia accretionary
wedge in the Olympic Mountains, northwest Washington State, Geological Society of America Bulletin,
110(8), 985-1009, 1998.

Braun, J., Quantitative constraints on the rate of landform evolution derived from low-temperature
thermochronology, Reviews in Mineralogy and Geochemistry, 58(1), 351-374, 2005.

Braun, J., 2010, The many surface expressions of mantle dynamics: Nature Geoscience, v. 3, no. 12, p.
825–833, doi:10.1038/ngeo1020.
Brown, R.L. and Read, P.B., Shuswap terrane of British Columbia: A Mesozoic “core complex”, Geology,
11(3), 164-168, 1983.

Brown, R.W., Backstacking apatite fission-track" stratigraphy": A method for resolving the erosional and
isostatic rebound components of tectonic uplift histories, Geology, 19(1), 74-77, 1991.

Brown, R.W., et al, Natural age dispersion arising from the analysis of broken crystals. Part I: Theoretical
basis and implications for the apatite (U–Th)/He thermochronometer, Geochimica et Cosmochimica
Acta, 122 478-497, 2013.

Burianyk, M., Kanasewich, E. and Udey, N., Broadside wide-angle seismic studies and three-dimensional
structure of the crust in the southeast Canadian Cordillera, Canadian Journal of Earth Sciences, 34(8),
1156-1166, 1997.

Burtner, R.L., Nigrini, A. and Donelick, R.A., Thermochronology of Lower Cretaceous source rocks in the
Idaho-Wyoming thrust belt, AAPG Bulletin, 78(10), 1613-1636, 1994.

Carlson, W.D., Donelick, R.A. and Ketcham, R.A., Variability of apatite fission-track annealing kinetics: I.
Experimental results, American Mineralogist, 84 1213-1223, 1999.

Carr, S.D., Parrish, R.R. and Brown, R.L., Eocene structural development of the Valhalla complex,
southeastern British Columbia, Tectonics, 6(2), 175-196, 1987.

265 | P a g e
References

Clague, J.J., The St. Eugene formation and the development of the southern Rocky Mountain trench,
Canadian Journal of Earth Sciences, 11(7), 916-938, 1974.

Clague, J.J., Glacio-isostatic effects of the Cordilleran ice sheet, British Columbia, Canada, Shorelines and
Isostasy.Academic Press, London, 321-343, 1983.

Clague, J.J., Quaternary glaciation and sedimentation, Geology of the cordilleran orogen in Canada;
edited by: Gabrielse, H., and Yorath, C. J., Geological Survey of Canada, Ottawa, Canada, 419–434,
1991a.

Clague, J., Quaternary glaciation and sedimentation, Geology of the cordilleran orogen in Canada,
edited by: Gabrielse, H., and Yorath, CJ, Geological Survey of Canada, Ottawa, Canada, 419-434, 1991b.

Clague, J.J. and James, T.S., History and isostatic effects of the last ice sheet in southern British
Columbia, Quaternary Science Reviews, 21(1), 71-87, 2002.

Clague, J.J., Cordilleran ice sheet. Encyclopedia of Paleoclimatology and Ancient Environments, Springer,
206-211, 2009.

Clark, M. K., J.W.M. Bush, and L.H. Royden (2005) Dynamic topography produced by lower crustal flow
against rheological strength heterogeneities bordering the Tibetan Plateau Geophys. J. Int. (2005) 162
(2): 575-590 doi:10.1111/j.1365-246X.2005.02580.x

Clowes, R.M., et al, Lithospheric structure in the southern Canadian Cordillera from a network of seismic
refraction lines, Canadian Journal of Earth Sciences, 32(10), 1485-1513, 1995.

Colpron, M. and Price, R.A., Tectonic significance of the Kootenay terrane, southeastern Canadian
Cordillera: An alternative model, Geology, 23(1), 25-28, 1995.

Colpron, M., Price, R.A. and Archibald, D.A., 40Ar/39Ar thermochronometric constraints on the tectonic
evolution of the Clachnacudainn complex, southeastern British Columbia, Canadian Journal of Earth
Sciences, 36(12), 1989-2006, 2000.

Crowley, J.L. and Brown, R.L., Tectonic links between the Clachnacudainn terrane and Selkirk allochthon,
southern Omineca Belt, Canadian Cordillera, Tectonics, 13(5), 1035-1051, 1994.

Cubley, J., et al, Thermochronological constraints on the Eocene exhumation of the Grand Forks
complex, British Columbia, based on 40Ar/39Ar and apatite fission track geochronology, Canadian
Journal of Earth Sciences, 50(5), 576-598, 2013.

Currie, C.A. and Hyndman, R.D., The thermal structure of subduction zone back arcs, Journal of
Geophysical Research: Solid Earth (1978–2012), 111(B8),2006.

Currie, C.A., Huismans, R.S. and Beaumont, C., Thinning of continental backarc lithosphere by flow-
induced gravitational instability, Earth and Planetary Science Letters, 269(3), 436-447, 2008.

Dahlstrom, C.D.A., Structural geology in eastern margin of Canadian Rocky Mountains, Bulletin of the
American Association of Petroleum Geologists, 54 843, 1970.

Dawson, G.M. and McConnell, R.G., Glacial deposits of southwestern Alberta in the vicinity of the Rocky
Mountains, Geological Society of America Bulletin, 7(1), 31-66, 1895.

Denton, G.H. and Armstrong, R.L., Miocene-Pliocene glaciations in southern Alaska, American Journal of
Science, 267(10), 1121-1142, 1969.

Dodson, M.H., Closure temperature in cooling geochronological and petrological systems, Contributions
to Mineralogy and Petrology, 40(3), 259-274, 1973.

266 | P a g e
References

Donelick, R.A., Crystallographic orientation dependence of mean etchable fission track length in apatite:
An empirical model and experimental observations, American Mineralogist;(United States), 76, 1991.

Donelick, R.A., Microscopic analysis of etch figure to determine fluorine chlorien and water rich apatite,
1993.

Duddy, I., Green, P. and Laslett, G., Thermal annealing of fission tracks in apatite 3. Variable
temperature behaviour, Chemical Geology: Isotope Geoscience section, 73(1), 25-38, 1988.

Ehlers, T.A. and Farley, K.A., Apatite (U–Th)/He thermochronometry: methods and applications to
problems in tectonic and surface processes, Earth and Planetary Science Letters, 206(1), 1-14, 2003.

Ehlers, T.A., Crustal thermal processes and the interpretation of thermochronometer data, Reviews in
Mineralogy and Geochemistry, 58(1), 315-350, 2005.

England, P. and Molnar, P., Surface uplift, uplift of rocks, and exhumation of rocks, Geology, 18(12),
1173-1177, 1990.

Evenchick, C.A., Parrish, R.R. and Gabrielse, H., Precambrian gneiss and late Proterozoic sedimentation
in north-central British Columbia, Geology, 12(4), 233-237, 1984.

Ewing, R.C., Weber, W.J. and Clinard Jr, F.W., Radiation effects in nuclear waste forms for high-level
radioactive waste, Progress in Nuclear Energy, 29(2), 63-127, 1995.

Farley, K., Wolf, R. and Silver, L., The effects of long alpha-
Geochimica et Cosmochimica Acta, 60(21), 4223-4229, 1996a.

Farley, K., Wolf, R. and Silver, L., The effects of long alpha- e ages,
Geochimica et Cosmochimica Acta, 60(21), 4223-4229, 1996b.

Farley, K., Helium diffusion from apatite: General behavior as illustrated by Durango fluorapatite,
Journal of Geophysical Research: Solid Earth (1978–2012), 105(B2), 2903-2914, 2000.

Farley, K., Rusmore, M. and Bogue, S., Post–10 Ma uplift and exhumation of the northern Coast
Mountains, British Columbia, Geology, 29(2), 99-102, 2001.

Farley, K.A., (U-Th)/He dating: techniques, calibrations, and applications, Reviews in Mineralogy and
Geochemistry, 47(1), 819-844, 2002.

Farley, K.A. and Stockli, D.F., (U-Th)/He dating of phosphates: Apatite, monazite, and xenotime, Reviews
in mineralogy and geochemistry, 48(1), 559-577, 2002.

Flament, N., Gurnis, G., and Mueller, D. (2013) A review of observations and models of dynamic
topography; LITHOSPHERE; v. 5; no. 2; p. 189–210 ; doi: 10.1130/L245.

Fleischer, R., Price, P. and Walker, R., Ion explosion spike mechanism for formation of charged‐particle
tracks in solids, Journal of Applied Physics, 36(11), 3645-3652, 1965.

Fleischer, R. and Hart, H., Fission track dating: techniques and problems, Calibration of Hominoid
Evolution, 135 170, 1972.

Flowers, R., et al, Radiation damage control on apatite (U-Th)/He dates from the Grand Canyon region,
Colorado Plateau, Geology, 35(5), 447-450, 2007.

Flowers, R.M., et al, Apatite (U–Th)/He thermochronometry using a radiation damage accumulation and
annealing model, Geochimica et Cosmochimica Acta, 73(8), 2347-2365, 2009.

Flück, P., Hyndman, R. and Lowe, C., Effective elastic thickness Te of the lithosphere in western Canada,
Journal of Geophysical Research: Solid Earth (1978–2012), 108(B9),2003.

267 | P a g e
References

Foeken, J., et al, A diode laser system for heating minerals for (U‐Th)/He chronometry, Geochemistry,
Geophysics, Geosystems, 7(4),2006.

Foster, G., Lunt, D. and Parrish, R., Mountain uplift and the glaciation of North America–a sensitivity
study, Climate of the Past, 6(5), 707-717, 2010.

Fyles, J.T. and Eastwood, G.E.P., Geology of the Ferguson Area, Lardeau District British Columbia, British
Columbia Department of Mines and Petroleum Resources Bulletin, 45 91, 1962.

Galbraith, R., On statistical models for fission track counts, Journal of the International Association for
Mathematical Geology, 13(6), 471-478, 1981.

Galbraith, R.F., The radial plot: Graphical assessment of spread in ages, International Journal of
Radiation Applications and Instrumentation.Part D.Nuclear Tracks and Radiation Measurements, 17(3),
207-214, 1990.

Galbraith, R.F. and Laslett, G.M., Statistical models for mixed fission track ages, Nuclear Tracks and
Radiation Measurements, 21(4), 459-470, 1993.

Gallagher, K., Brown, R. and Johnson, C., Fission track analysis and its applications to geological
problems, Annual Review of Earth and Planetary Sciences, 26(1), 519-572, 1998.

Gallagher, K., Transdimensional inverse thermal history modeling for quantitative thermochronology,
Journal of Geophysical Research: Solid Earth (1978–2012), 117(B2),2012.

Gautheron, C. and Tassan-Got, L., A Monte Carlo approach to diffusion applied to noble gas/helium
thermochronology, Chemical Geology, 273(3), 212-224, 2010.

Gehrels, G., et al, New Departures in Structural Geology and Tectonics, Meeting in Denver, CO (white
paper). 48, 2002.

Gleadow, A. and Duddy, I., A natural long-term track annealing experiment for apatite, Nuclear Tracks,
5(1), 169-174, 1981.

Gleadow, A., Duddy, I. and Lovering, J., Fission track analysis: a new tool for the evaluation of thermal
histories and hydrocarbon potential, APEA Journal, 23(1), 93-102, 1983.

Govers, R. and Meijer, P.T., On the dynamics of the Juan de Fuca plate, Earth and Planetary Science
Letters, 189(3), 115-131, 2001.

Grasby, S.E. and Hutcheon, I., Controls on the distribution of thermal springs in the southern Canadian
Cordillera, Canadian Journal of Earth Sciences, 38(3), 427-440, 2001.

Green, P., et al, Thermal annealing of fission tracks in apatite 4. Quantitative modelling techniques and
extension to geological timescales, Chemical Geology: Isotope Geoscience Section, 79(2), 155-182,
1989.

Green, P. and Durrani, S., Annealing studies of tracks in crystals, Nuclear Track Detection, 1(1), 33-39,
1977.

Guenthner, W.R., et al, Helium diffusion in natural zircon: Radiation damage, anisotropy, and the
interpretation of zircon (U-Th)/He thermochronology, American Journal of Science, 313(3), 145-198,
2013.

Haenel, R., Rybach, L. and Stegena, L., Handbook of terrestrial heat-flow density determination, Solid
Earth Sciences Library, 1988.

268 | P a g e
References

Haeussler, P.J., et al, Life and death of the Resurrection plate: Evidence for its existence and subduction
in the northeastern Pacific in Paleocene–Eocene time, Geological Society of America Bulletin, 115(7),
867-880, 2003.

Hanson, G. and Gast, P., Kinetic studies in contact metamorphic zones, Geochimica et Cosmochimica
Acta, 31(7), 1119-1153, 1967.

Hardebol, N., Pysklywec, R. and Stephenson, R., Small‐scale convection at a continental back‐arc to
craton transition: Application to the southern Canadian Cordillera, Journal of Geophysical Research:
Solid Earth (1978–2012), 117(B1),2012.

Hasebe, N., et al, Apatite fission-track chronometry using laser ablation ICP-MS, Chemical Geology,
207(3), 135-145, 2004.

Hasterok, D. and Chapman, D.S., Continental thermal isostasy: 1. Methods and sensitivity, Journal of
Geophysical Research: Solid Earth (1978–2012), 112(B6),2007.

Herman, F., et al, Worldwide acceleration of mountain erosion under a cooling climate, Nature,
504(7480), 423-426, 2013.

Hickson, C.J. and Souther, J., Late Cenozoic volcanic rocks of the Clearwater-Wells Gray area, British
Columbia, Canadian Journal of Earth Sciences, 21(3), 267-277, 1984.

Hickson, C., Late Cenozoic rocks of the Wells Gray—Clearwater area, British Columbia, 1987.

Hurford, A.J. and Carter, A., The role of fission track dating in discrimination of provenance, Geological
Society, London, Special Publications, 57(1), 67-78, 1991.

Hurford, A.J. and Green, P.F., The zeta age calibration of fission-track dating, Chemical Geology, 41 285-
317, 1983.

Hurley, P.M., The helium age method and the distribution and migration of helium in rocks,
Massachusetts Institute of Technology, Department of Geology and Geophysics, 1955.

Hurley, P., et al, Radiogenic argon and strontium diffusion parameters in biotite at low temperatures
obtained from Alpine Fault uplift in New Zealand, Geochimica et Cosmochimica Acta, 26(1), 67-80,
1962.

Hutchinson, I., Holocene sea level change in the Pacific Northwest: a catalogue of radiocarbon dates and
an atlas of regional sea-level curves, Institute of Quaternary Research, Simon Fraser University, 1992.

Hyndman, R., The consequences of Canadian Cordillera thermal regime in recent tectonics and
elevation: a review, Canadian Journal of Earth Sciences, 47(5), 621-632, 2010.

Hyndman, R., et al, Temperature control of continental lithosphere elastic thickness, Te vs Vs, Earth and
Planetary Science Letters, 277(3), 539-548, 2009.

Hyndman, R. and Lewis, T., Geophysical consequences of the Cordillera–Craton thermal transition in
southwestern Canada, Tectonophysics, 306(3), 397-422, 1999.

Hyndman, R.D., Currie, C.A. and Mazzotti, S.P., Subduction zone backarcs, mobile belts, and orogenic
heat, GSA Today, 15(2), 4-10, 2005.

Issler, D., et al, Preliminary evidence from apatite fission-track data concerning the thermal history of
the Peace River Arch region, Western Canada Sedimentary Basin, Bulletin of Canadian Petroleum
Geology, 38(1), 250-269, 1990.

James, T.S., et al, Postglacial rebound at the northern Cascadia subduction zone, Quaternary Science
Reviews, 19(14), 1527-1541, 2000.

269 | P a g e
References

Jonckheere, R., et al, L'apatite de Durango (Mexique): analyse d'un minéral standard pour la datation
par traces de fission, Chemical Geology, 103(1), 141-154, 1993.

Kalkreuth, W. and McMechan, M., Burial history and thermal maturity, Rocky Mountain Front Ranges,
foothills, and foreland, east-central British Columbia and adjacent Alberta, Canada, AAPG Bulletin,
72(11), 1395-1410, 1988.

Kent-Corson, M.L., et al, Cenozoic topographic and climatic response to changing tectonic boundary
conditions in western North America, Earth and Planetary Science Letters, 252(3), 453-466, 2006.

Ketcham, R.A., Forward and inverse modeling of low-temperature thermochronometry data, Reviews in
Mineralogy and Geochemistry, 58(1), 275-314, 2005.

Ketcham, R.A., et al, Improved measurement of fission-track annealing in apatite using c-axis projection,
American Mineralogist, 92(5-6), 789-798, 2007a.

Ketcham, R.A., et al, Improved modeling of fission-track annealing in apatite, American Mineralogist,
92(5-6), 799-810, 2007b.

Ketcham, R.A., Donelick, R.A. and Carlson, W.D., Variability of apatite fission-track annealing kinetics: III.
Extrapolation to geological time scales, American Mineralogist, 84 1235-1255, 1999.

Ketcham, R.A., Donelick, R.A. and Donelick, M.B., AFTSolve: A program for multi-kinetic modeling of
apatite fission-track data, Geological Materials Research, 2(1), 1-32, 2000.

Ketcham, R.A., Gautheron, C. and Tassan-Got, L., Accounting for long alpha-particle stopping distances
in (U–Th–Sm)/He geochronology: refinement of the baseline case, Geochimica et Cosmochimica Acta,
75(24), 7779-7791, 2011a.

Ketcham, R.A., Gautheron, C. and Tassan-Got, L., Accounting for long alpha-particle stopping distances
in (U–Th–Sm)/He geochronology: refinement of the baseline case, Geochimica et Cosmochimica Acta,
75(24), 7779-7791, 2011b.

Kohn, M.J., Miselis, J.L. and Fremd, T.J., Oxygen isotope evidence for progressive uplift of the Cascade
Range, Oregon, Earth and Planetary Science Letters, 204(1), 151-165, 2002.

Lachenbruch, A. H. (1970), Crustal temperature and heat production: Implications of the linear heat-
flow relation, J. Geophys. Res., 75(17), 3291–3300, doi:10.1029/JB075i017p03291.

Laslett, G., et al, Thermal annealing of fission tracks in apatite 2. A quantitative analysis, Chemical
Geology: Isotope Geoscience Section, 65(1), 1-13, 1987.

Lay, D., Fraser River Tertiary drainage history in relation to placer gold deposits (part 1), Bulletin Mo, 11
30, 1940.

Lithgow-Bertelloni, Carolina, and Paul G. Silver. "Dynamic topography, plate driving forces and the
African superswell." Nature 395.6699 (1998): 269-272.
Lorencak, M., et al, Low-temperature cooling history of the Shuswap metamorphic core complex, British
Columbia: constraints from apatite and zircon fission-track ages, Canadian Journal of Earth Sciences,
38(11), 1615-1625, 2001.

Lowry, A.R. and Smith, R.B., Strength and rheology of the western US Cordillera, JOURNAL OF
GEOPHYSICAL RESEARCH-ALL SERIES-, 100 17,947-17,947, 1995.

Madsen, J., et al, Cenozoic to Recent plate configurations in the Pacific Basin: Ridge subduction and slab
window magmatism in western North America, Geosphere, 2(1), 11-34, 2006.

Mancktelow, N.S. and Grasemann, B., Time-dependent effects of heat advection and topography on
cooling histories during erosion, Tectonophysics, 270(3), 167-195, 1997.

270 | P a g e
References

Mason, B., Potassium-argon ages of metamorphic rocks and granites from Westland, New Zealand, New
Zealand Journal of Geology and Geophysics, 4(4), 352-356, 1961.

Mathews, W., Physiographic evolution of the Canadian Cordillera, Geology of the Cordilleran Orogen in
Canada.Edited by H.Gabrielse and CJ Yorath.Geological Survey of Canada, Geology of Canada, (4), 403-
418, 1991.

Mathews, W., Neogene Chilcotin basalts in south-central British Columbia: geology, ages, and
geomorphic history, Canadian Journal of Earth Sciences, 26(5), 969-982, 1989.

Mathews, W.H., Physiography of the Canadian Cordillera, Map 1701A, scale 1:5.000.000, Geological
Survey of Canada, 1986.

Mathews, W. and Rouse, G., An Early Pleistocene proglacial succession in south-central British
Columbia, Canadian Journal of Earth Sciences, 23(11), 1796-1803, 1986.

Mathews, W.H., Fyles, J. and Nasmith, H., Postglacial crustal movements in southwestern British
Columbia and adjacent Washington State, Canadian Journal of Earth Sciences, 7(2), 690-702, 1970.

McDonough, M., et al, Apatite fission track and 40Ar-39Ar analysis of the external zone of the Canadian
Rockies near Jasper, Alberta: Implications for thermal maturity, Proceedings of the Oil and Gas
Forum'95: Energy from Sediments: March 28-29, 1995, Calgary, Alberta, 1995, Natural Resources
Canada, 83, 1995.

Miller, K.G., et al, The Phanerozoic record of global sea-level change, Science (New York, N.Y.),
310(5752), 1293-1298, 2005.

Molnar, P. and England, P., Late Cenozoic uplift of mountain ranges and global climate change: chicken
or egg? Nature, 346(6279), 29-34, 1990.

Monger, J., Price, R. and Tempelman-Kluit, D., Tectonic accretion and the origin of the two major
metamorphic and plutonic welts in the Canadian Cordillera, Geology, 10(2), 70-75, 1982.

Morrow, D., et al, Paleozoic burial and organic maturation in the Liard Basin region, northern Canada,
Bulletin of Canadian Petroleum Geology, 41(1), 17-31, 1993.

Naeser, C.W., Fission Track Dating, US Geol Surv Open-File Report 76-190, 1976.

Naeser, C., The use of apatite and sphene for fission track age determinations, Geological Society of
America Bulletin, 78(12), 1523-1526, 1967.

Naeser, N.D., et al, Fission-track ages of late Cenozoic distal tephra beds in the Yukon Territory and
Alaska, Canadian Journal of Earth Sciences, 19(11), 2167-2178, 1982.

Nasdala, L., et al, Incomplete retention of radiation damage in zircon from Sri Lanka, American
Mineralogist, 89(1), 219-231, 2004.

Nurkowski, J., Coal quality and rank variation within Upper Cretaceous and Tertiary sediments, Alberta
plains region, Alberta Research Council Earth Sciences Report, 01, 1985.

Osadetz, K.G., et al, Foreland belt thermal history using apatite fission-track thermochronology:
Implications for Lewis thrust and Flathead fault in the southern Canadian Cordilleran petroleum
province, Deformation, fluid flow, and reservoir appraisal in foreland fold and thrust belts: AAPG
Hedberg Series, 1 21-48, 2004.

Osadetz, K.G., Pearson, D.E. and Stasiuk, L.D., Paleogeothermal gradients and changes in the geothermal
gradient field of the Alberta Plains, Current research part D: Geological Survey of Canada Paper, 165-
178, 1990.

271 | P a g e
References

Osborn, G., Stockmal, G. and Haspel, R., Emergence of the Canadian Rockies and adjacent plains: A
comparison of physiography between end-of-Laramide time and the present day, Geomorphology,
75(3), 450-477, 2006.

O'Sullivan, P.B. and Parrish, R.R., The importance of apatite composition and single-grain ages when
interpreting fission track data from plutonic rocks: a case study from the Coast Ranges, British Columbia,
Earth and Planetary Science Letters, 132(1), 213-224, 1995.

Parrish, R.R., Thermal evolution of the southeastern Canadian Cordillera, Canadian Journal of Earth
Sciences, 32(10), 1618-1642, 1995a.

Parrish, R.R., Thermal evolution of the southeastern Canadian Cordillera, Canadian Journal of Earth
Sciences, 32(10), 1618-1642, 1995b.

Parrish, R.R., Cenozoic thermal evolution and tectonics of the Coast Mountains of British Columbia: 1.
Fission track dating, apparent uplift rates, and patterns of uplift, Tectonics, 2(6), 601-631, 1983.

Parrish, R.R., Carr, S.D. and Parkinson, D.L., Eocene extensional tectonics and geochronology of the
southern Omineca Belt, British Columbia and Washington, Tectonics, 7(2), 181-212, 1988.

Parrish, R., Miscellaneous U-Pb zircon dates from southeast British Columbia, Radiogenic age and
isotopic studies: Report, 5 91-92, 1992.

Parrish, R. and Armstrong, R., U-Pb zircon age and tectonic significance of gneisses in structural
culminations of the Omineca crystalline belt, British Columbia: Geological Society of America Abstracts
with Programs, 15(5), 324, 1983.

Peslier, A.H., Francis, D. and Ludden, J., The lithospheric mantle beneath continental margins: melting
and melt–rock reaction in Canadian Cordillera xenoliths, Journal of Petrology, 43(11), 2013-2047, 2002.

Press, W.H., et al, Numerical recipes, 1990.

Price, R., The Cordilleran Forelarid Thrust and Fold Belt in the Southern'Canadian Rocky Mountains,
1981a.

Price, R.A., The Cordilleran foreland thrust and fold belt in the southern Canadian Rocky Mountains,
McClay, K.R., and Price, N.J., eds., Thrust and nappe tectonics: London, Geological Society of London,
427–448, 1981b.

Price, R., Cordilleran tectonics and the evolution of the Western Canada Sedimentary Basin, Geological
Atlas of the Western Canada Sedimentary Basin, 41994.

Price, R.A. and Carmichael, D.M., Geometric test for Late Cretaceous-Paleogene intracontinental
transform faulting in the Canadian Cordillera, Geology, 14(6), 468-471, 1986.

Ranalli, G., Brown, R. and Bosdachin, R., A geodynamic model for extension in the Shuswap core
complex, southeastern Canadian Cordillera, Canadian Journal of Earth Sciences, 26(8), 1647-1653, 1989.

Raymo, M. and Ruddiman, W.F., Tectonic forcing of late Cenozoic climate, Nature, 359(6391), 117-122,
1992.

Read, P.B. and Brown, R.L., Columbia River fault zone: southeastern margin of the Shuswap and
Monashee complexes, southern British Columbia, Canadian Journal of Earth Sciences, 18(7), 1127-1145,
1981.

Reiners, P.W., Zircon (U-Th)/He thermochronometry, Reviews in Mineralogy and Geochemistry, 58(1),
151-179, 2005.

272 | P a g e
References

Reiners, P.W., et al, Helium and argon thermochronometry of the Gold Butte block, south Virgin
Mountains, Nevada, Earth and Planetary Science Letters, 178(3), 315-326, 2000.

Reiners, P.W. and Brandon, M.T., Using thermochronology to understand orogenic erosion,
Annu.Rev.Earth Planet.Sci., 34 419-466, 2006.

Reiners, P.W., Farley, K.A. and Hickes, H.J., He diffusion and (U–Th)/He thermochronometry of zircon:
initial results from Fish Canyon Tuff and Gold Butte, Tectonophysics, 349(1), 297-308, 2002.

Reiners, P.W., et al, Zircon (U-Th)/He thermochronometry: He diffusion and comparisons with 40 Ar/39
Ar dating, Geochimica et Cosmochimica Acta, 68(8), 1857-1887, 2004.

Reiners, P.W., Ehlers, T.A. and Zeitler, P.K., Past, present, and future of thermochronology, Reviews in
Mineralogy and Geochemistry, 58(1), 1-18, 2005.

Reiners, P., et al, Wildfire thermochronology and the fate and transport of apatite in hillslope and fluvial
environments, Journal of Geophysical Research: Earth Surface (2003–2012), 112(F4),2007.

Riddihough, R., Recent movements of the Juan de Fuca plate system, Journal of Geophysical Research:
Solid Earth (1978–2012), 89(B8), 6980-6994, 1984.

Rio, M.-H., and F. Hernandez (2004), A mean dynamic topography computed over the world ocean from
altimetry, in situ measurements, and a geoid model, J. Geophys. Res., 109, C12032,
doi:10.1029/2003JC002226.

Rogers, G. and Souther, J., Hotspots trace plate movements, Geoscience Canada, 12 10-13, 1983.

Rohr, K.M. and Currie, L., Queen Charlotte basin and Coast Mountains: Paired belts of subsidence and
uplift caused by a low-angle normal fault, Geology, 25(9), 819-822, 1997.

Rouse, G. and Mathews, W., Tertiary geology and palynology of the Quesnel area, British Columbia,
Bulletin of Canadian Petroleum Geology, 27(4), 418-445, 1979.

Rubin, C.M., et al, Regionally extensive mid-Cretaceous west-vergent thrust system in the northwestern
Cordillera: Implications for continent-margin tectonism, Geology, 18(3), 276-280, 1990.

Ruddiman, W. and Kutzbach, J., Forcing of late Cenozoic northern hemisphere climate by plateau uplift
in southern Asia and the American West, Journal of Geophysical Research: Atmospheres (1984–2012),
94(D15), 18409-18427, 1989.

Shuster, D.L. and Farley, K.A., The influence of artificial radiation damage and thermal annealing on
helium diffusion kinetics in apatite, Geochimica et Cosmochimica Acta, 73(1), 183-196, 2009.

Smith, M.T. and Gehrels, G.E., Structural geology of the Lardeau Group near Trout Lake, British
Columbia: implications for the structural evolution of the Kootenay Arc, Canadian Journal of Earth
Sciences, 29(6), 1305-1319, 1992b.

Smith, M.T. and Gehrels, G.E., Stratigraphic comparison of the Lardeau and Covada groups: implications
for revision of stratigraphic relations in the Kootenay Arc, Canadian Journal of Earth Sciences, 29(6),
1320-1329, 1992a.

Souther, J., Armstrong, R. and Harakal, J., Chronology of the peralkaline, late Cenozoic Mount Edziza
volcanic complex, northern British Columbia, Canada, Geological Society of America Bulletin, 95(3), 337-
349, 1984.

Spear, F.S., Fast cooling and exhumation of the Valhalla metamorphic core complex, southeastern
British Columbia, International Geology Review, 46(3), 193-209, 2004.

273 | P a g e
References

Spear, F.S. and Parrish, R.R., Petrology and cooling rates of the Valhalla complex, British Columbia,
Canada, Journal of Petrology, 37(4), 733-765, 1996.

Stockli, D.F., Farley, K.A. and Dumitru, T.A., Calibration of the apatite (U-Th)/He thermochronometer on
an exhumed fault block, White Mountains, California, Geology, 28(11), 983-986, 2000.

Stüwe, K., White, L. and Brown, R., The influence of eroding topography on steady-state isotherms.
Application to fission track analysis, Earth and Planetary Science Letters, 124(1), 63-74, 1994.

Takeuchi, A. and Larson, P.B., Oxygen isotope evidence for the late Cenozoic development of an
orographic rain shadow in eastern Washington, USA, Geology, 33(4), 313-316, 2005.

Tempelman-Kluit, D.J. and Parkinson, D.L., Extension across the Eocene Okanagan crustal shear in
southern British Columbia, Geology, 14 318-318-321, 1986.

Tipper, H.W., Taseko Lakes, British Columbia, Geol. Surv. Can. Open File, Map 5341978.

Tribe, S., Eocene paleo-physiography and drainage directions, southern Interior Plateau, British
Columbia, Canadian Journal of Earth Sciences, 42(2), 215-230, 2005.

Trop, J.M., et al, Mesozoic sedimentary-basin development on the allochthonous Wrangellia composite
terrane, Wrangell Mountains basin, Alaska: A long-term record of terrane migration and arc
construction, Geological Society of America Bulletin, 114(6), 693-717, 2002.

Turcotte, D. and Schubert, G., Geodynamics: Applications of continuum physics to geological problems,
450 pp, 1982.

Tushingham, A. and Peltier, W., Ice-3 g- A new global model of late Pleistocene deglaciation based upon
geophysical predictions of post-glacial relative sea level change, Journal of Geophysical Research,
96(B3), 4497-4523, 1991.

Unterschutz, J.L., et al, North American margin origin of Quesnel terrane strata in the southern
Canadian Cordillera: Inferences from geochemical and Nd isotopic characteristics of Triassic
metasedimentary rocks, Geological Society of America Bulletin, 114(4), 462-475, 2002.

Van Der Lee, S. and Frederiksen, A., Surface wave tomography applied to the North American upper
mantle, Seismic earth: array analysis of broadband seismograms, 67-80, 2005.

Van der Pluijm, B. A., et al, Fault dating in the Canadian Rocky Mountains: Evidence for late Cretaceous
and early Eocene orogenic pulses, Geology, 34(10), 837-840, 2006.

Vanderhaeghe, O., et al, Cooling and exhumation of the Shuswap Metamorphic Core Complex
constrained by 40Ar/39Ar thermochronology, Geological Society of America Bulletin, 115(2), 200-216,
2003.

Vermeesch, P., et al, A simple method for in-situ U-Th-He dating, Geochimica et Cosmochimica Acta, 79
140-147, 2012.

Vermeesch, P. and Tian, Y., Thermal history modelling: HeFTy vs. QTQt, Earth-Science Reviews, 139 279-
290, 2014.

Vermeesch, P., RadialPlotter: a Java application for fission track, luminescence and other radial plots,
Radiation Measurements, 44(4), 409, 2009.

Vermeesch, P., et al, α-Emitting mineral inclusions in apatite, their effect on (U–Th)/He ages, and how to
reduce it, Geochimica et Cosmochimica Acta, 71(7), 1737-1746, 2007.

Wanless, R. and Reesor, J., Precambrian zircon age of orthogneiss in the Shuswap Metamorphic
Complex, British Columbia, Canadian Journal of Earth Sciences, 12(2), 326-332, 1975.

274 | P a g e
References

Watts, A.B., Isostasy and Flexure of the Lithosphere, Cambridge University Press, 2001.

Weber, W., Ewing, R. and Meldrum, A., The kinetics of alpha-decay-induced amorphization in zircon and
apatite containing weapons-grade plutonium or other actinides, Journal of Nuclear Materials, 250(2),
147-155, 1997.

Weber, W.J., Ewing, R.C. and Wang, L., The radiation-induced crystalline-to-amorphous transition in
zircon, Journal of Materials Research, 9(03), 688-698, 1994.

Wheeler, J.O., Eastern tectonic belt of western Cordillera in British Columbia Tectonic history and
mineral deposits of the western Cordillera, Special Volume 8 27–45, 1966.

Wheeler, J.O., McFeely, P. and Sigouin, M., Tectonic assemblage map of the Canadian Cordillera and
adjacent parts of the United States of America, Geological Survey of Canada, 1991.

White, J., et al, An 18 million year record of vegetation and climate change in northwestern Canada and
Alaska: tectonic and global climatic correlates, Palaeogeography, Palaeoclimatology, Palaeoecology,
130(1), 293-306, 1997.

Winterbourne, J., N. White, and A. Crosby (2014), Accurate measurements of residual topography from
theoceanic realm, Tectonics, 33, 982–1015, doi:10.1002/2013TC003372.

Willett, S.D., Inverse modeling of annealing of fission tracks in apatite; 1, A controlled random search
method, American Journal of Science, 297(10), 939-969, 1997.

Wolf, R., Farley, K. and Kass, D., Modeling of the temperature sensitivity of the apatite (U–Th)/He
thermochronometer, Chemical Geology, 148(1), 105-114, 1998.

Wolf, R., Farley, K. and Silver, L., Helium diffusion and low-temperature thermochronometry of apatite,
Geochimica et Cosmochimica Acta, 60(21), 4231-4240, 1996.

Woodsworth, G., Anderson, R. and Armstrong, R., Plutonic Regimes in Geology of the Cordilleran
Orogen in Canada, Gabrielse, H. and Yorath, CJ (eds.), Geological Survey of Canada, Geology of Canada,
(4), 9, 1991.

Worthington, R.H.S., Karst hydrogeology of the Canadian rocky mountains, 1991.

Wright, G., McMechan, M. and Potter, D., Structure and architecture of the Western Canada
sedimentary basin, Geological atlas of the Western Canada sedimentary basin, 25-40, 1994.

Wyrtki, K. (1975) Fluctuations of the Dynamic Topography in the Pacific Ocean. J. Phys. Oceanogr., 5,
450–459, doi: 10.1175/1520-0485(1975)005<0450:FOTDTI>2.0.CO;2.

Yoshioka, T., et al, Spontaneous fission decay constant of 238U determined by SSNTD method using CR-
39 and DAP plates, Nuclear Instruments and Methods in Physics Research Section A: Accelerators,
Spectrometers, Detectors and Associated Equipment, 555(1–2), 386-395, 2005.

Zeitler, P.K., et al, U-Th-He dating of apatite: A potential thermochronometer. Geochim Cosmochim
Acta 51:2865-2868, 1987.

Zelt, C. and White, D., Crustal structure and tectonics of the southeastern Canadian Cordillera, Journal
of Geophysical Research: Solid Earth (1978–2012), 100(B12), 24255-24273, 1995.

275 | P a g e

You might also like