You are on page 1of 11

Earth and Planetary Science Letters 308 (2011) 4151

Contents lists available at ScienceDirect

Earth and Planetary Science Letters


j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / e p s l

Cold conditions in Antarctica during the Little Ice Age Implications for abrupt
climate change mechanisms
N.A.N. Bertler a,, P.A. Mayewski b, L. Carter c
a
b
c

Joint Antarctic Research Institute, Victoria University and GNS Science, PO Box 600, Wellington 6012, New Zealand
Climate Change Institute, University of Maine, Orono, ME 04469-5790, USA
Antarctic Research Centre, Victoria University, PO Box 600, Wellington 6012, New Zealand

a r t i c l e

i n f o

Article history:
Received 20 October 2010
Received in revised form 5 May 2011
Accepted 8 May 2011
Editor: P. DeMenocal
Keywords:
Little Ice Age
Mediaeval Warm Period
abrupt climate change
sea-saw mechanism
Antarctica
Southern Ocean

a b s t r a c t
The Little Ice Age (LIA) is one of the most prominent climate shifts in the past 5000 yrs. It has been suggested
that the LIA might be the most recent of the DansgaardOeschger events, which are better known as abrupt,
large scale climate oscillations during the last glacial period. If the case, then according to Broecker (2000a,
2000b) Antarctica should have warmed during the LIA, when the Northern Hemisphere was cold. Here we
present new data from the Ross Sea, Antarctica, that indicates surface temperatures were ~ 2 C colder during
the LIA, with colder sea surface temperatures in the Southern Ocean and/or increased sea-ice extent, stronger
katabatic winds, and decreased snow accumulation. Whilst we nd there was large spatial and temporal
variability, overall Antarctica was cooler and stormier during the LIA. Although temperatures have warmed
since the termination of the LIA, atmospheric circulation strength has remained at the same, elevated level.
We conclude, that the LIA was either caused by alternative forcings, or that the sea-saw mechanism operates
differently during warm periods.
2011 Elsevier B.V. All rights reserved.

1. Introduction
The Little Ice Age (LIA) is a prominent climate shift dened on the
basis of glacier advances in Europe, and variable but cooler climate
conditions throughout the Northern Hemisphere (Grove, 1988).
Whilst the term LIA is conducive to a distinct climate event, occurring
over a distinct time period, reconstructions show a highly variable
climate pattern with marked regional differences, both in style and
timing of the climate signals (Mann et al., 1999).
For this reason, the causes, timing and geographical extent of the
LIA are still debated. However, three major climate modulators have
likely played a role: changes in solar output (Ammann et al., 2007;
Bard et al., 2000; Maasch et al., 2005; Mayewski et al., 1997, 2004b,
2006; O'Brien et al., 1996), increased volcanic activity (Crowley, 2000;
Robock, 2000), and changes in the thermohaline circulation
(Broecker, 2000b, 2001; Lund et al., 2006). The reason for the
complex spatial and temporal expressions is likely due to the LIA's
small amplitude (Table 1), which competes with other climate drivers
of similar or greater amplitude, such as the El Nio-Southern
Oscillation (Turner, 2004) or the Southern Annular Mode (Thompson
and Solomon, 2002).

As shown in Table 1, the lower temperature and snow line


associated with the LIA is about 10% that of the glacial/interglacial
changes, and about 2030% that of the Younger Dryas. However, over
the past 5 kyr, the LIA is one of the most prominent climate
modulations (Kreutz et al., 1997; Mayewski and Maasch, 2006).
Moreover, it has been suggested that the LIA is the most recent rapid
climate change (Bond et al., 1999) in a sequence of Dansgaard
Oeschger events (Broecker, 2000a), better known for their abrupt
occurrences during the last glacial period. For this reason, the LIA
provides an excellent opportunity to evaluate how the climate system
creates and responds to rapid change.
Here we present new data from the Ross Sea, Antarctica, an
important area of bottom water formation and contributor to the
density driven component of the global ocean circulation (Jacobs, 2004).
Our data show that LIA climate conditions were synchronous with those
in the Northern Hemisphere. We then summarise previously described
conditions in Antarctica during the LIA to discuss spatial and temporal
differences and implications for abrupt climate change mechanisms.
2. Material and methods
2.1. Study site

Corresponding author.
E-mail address: Nancy.Bertler@vuw.ac.nz (N.A.N. Bertler).
0012-821X/$ see front matter 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.epsl.2011.05.021

New Zealand contributes to the International Trans-Antarctic


Scientic Expedition (ITASE) (Mayewski et al., 2005) by collecting ice

42

N.A.N. Bertler et al. / Earth and Planetary Science Letters 308 (2011) 4151

Table 1
Comparison of approximate changes in snowline (relative to 1975 AD) and global/
hemispherical temperature (relative to preindustrial temperature) during major
climate shifts. Data for full glacial temperature change are derived from Schneider
von Diemling et al. (2006); all other data from Broecker (2000b).

Glacial to interglacial
Younger Dryas
Little Ice Age

Lowering of the
snowline (m)

Decrease in
temperature (C)

~900950
~350
~100

5.8 1.4
~34
0.6

the core, where detailed analyses with ~ 2.5 cm resolution were


carried out. The core was processed using a continuous melter
system (Osterberg et al., 2006). 1912 ice core samples were
analysed for stable isotope ratios ( 18O, D, d excess), major ions
(Na, Ca, K, Mg, Cl, NO3, SO4) and trace element (Fe, Al, Sr, P, Cu, Si)
concentrations. In addition, 144 high-resolution tritium measurements were conducted on the top 5 m of the ice core (Patterson
et al., 2005). The analytical methods and precision are described in
Appendix A.
2.3. Chronological framework

cores from coastal locations in the Ross Sea region (Fig. 1, Map 2).
Victoria Lower Glacier (VLG), in the northernmost McMurdo Dry
Valleys, is a small (5 30 km) valley glacier. It ows from its ice divide
westward into the Victoria Valley and eastward towards the coast,
where it feeds the Wilson Piedmont Glacier. The ice of VLG is locally
accumulated, and lies within 22 km of seasonally open ocean. The ice
core came from the highest point of the glacier, the ice divide, which
lies at 626 m above sea level and is underlain by over 600 m of ice.
As characteristic of the McMurdo Dry Valleys, annual snow
precipitation is low. Snow pit data indicate that VLG average annual
snow accumulation is 0.0330.013 m water equivalent per yr (w.e.a 1)
for the past ~ 40 yrs (Bertler et al., 2004a,b). An average annual
temperature of 22 C comes from 15 m-deep temperature measurements in a borehole (Bertler et al., 2004a,b). However, the
McMurdo Dry Valleys experience some of the largest seasonal
temperature amplitudes on Earth with Victoria Valley recording
summer maxima of 10 C and winter minima of 60 C (Doran et al.,
2002a,b). This exceptional range is caused by the ice free area
experiencing strong solar heating during the summer, and radiative
cooling during winter (King and Turner, 1997). The winter cooling is
particularly strong in Victoria Valley as the valley is sheltered from
katabatic winds and hence creates ideal conditions for a stable winter
stratication of the lower troposphere, which enhances effective
radiative cooling (Doran et al., 2002a,b).

where the density (p) of ice : pi = 0.917 g/cm 3, measured density at


the surface : po = 0.200 g/cm 3, density at the reference depth zr:
pr = 0.600 g/cm 3, reference depth : zr = 8 m, best t with the
measured density prole: d = 1/0.38. From this follows:

2.2. Data

p = abz + c

During the 2001/02 eld season, a 180 m-deep ice core was
recovered at the ice divide of VLG. This paper focuses on top 50 m of

where: a = pi, b = [(pi pr) d (pi p0)]/ zr, c = (pi p0), d = 1/0.38.
If the accumulation rate (mass of ice per unit area per unit time)

Due to low snow accumulation rates, annual layer counting is


not a viable option for dating the VLG ice core. The upper 4 m of the
core were dated through correlation with a 4 m-deep snow pit
record from the same site. The snow pit data were sampled with
1 cm resolution and dated using annual layer counting of seasonal
uctuations of sodium with a precision of 1 yr for the 36 yr
record (Bertler et al., 2004a,b). Average annual accumulation over
this time period is 0.033 0.013 m w.e. a 1. The tritium measurements were used to identify major peaks caused by nuclear testing
between 1957 and 1966 as well as seasonal tritium variations.
The dating error of the tritium measurements is b0.3 yrs (Patterson
et al., 2005).
An age model was extrapolated to the remaining core using a
rn decompaction model based on the density regression function:

z = zr

pi pd pi po d
pr po d pi po d

!
1

Fig. 1. Location map: 1) map of Antarctica with red square locates Map 2. Yellow numbers denote locations of ice core records: Taylor Dome, Talos Dome, and Law Dome.
Source map: NASA, Radar Image. 2) Map of the Ross Sea region. Black squares indicate core locations of the NZ ITASE programme. Square A locates Map 3. Source map: Latitudinal
Gradient Programme (Howard-Williams et al., 2006). 3) McMurdo Sound showing the drill site (red star) on Victoria Lower Glacier and location of Lake Vida automatic weather
station (yellow star). Source map: NASA Goddard Space Flight Center image from moderate-resolution imaging spectroradiometre (MODIS) sensor (J. Descloitres, MODIS Land Rapid
Response Team).

N.A.N. Bertler et al. / Earth and Planetary Science Letters 308 (2011) 4151

is a constant 1/k (Mg m 2 a 1), then the age t of the ice at


depth z is:

From this follows:



#
bd
c d + 1  c d + 1
t = k az
:
z+

d+1
d
d

a marine ratio of SO4/Na of 0.252 (in ng/g)(Legrand and Delmas,


1984):
nssSO4

t = k z xdx

"

The density regression is independent of age when constant snow


accumulation rates are assumed. Age benchmarks from the high
resolution tritium data, plus the annual snow layer count from the
snow pit data, were used in Eq. (4) to determine k. Using the
calculated value of k = 24.9 (or 1/k = 0.04 w.e. a 1), the age of the ice
at layer z can be calculated from Eq. (4).
To adjust the density regression age model for variable snow
accumulation rates, volcanic markers were used. Inventories of
volcanic eruptions around the world are constantly improved and
extended (e.g. Gao et al. (2008)). However, not all volcanic eruptions
are preserved in Antarctic ice. Here, we correlate non-sea-salt
sulphate (nss SO4) data, a commonly used proxy of volcanic eruptions
(Legrand and Mayewski, 1997) from the VLG record with the nss SO4
data from EPICA Dronning Maud Land (EDML) reconnaissance ice
cores (Traufetter et al., 2004) (Fig. 3). We chose the EDML ice core
(Traufetter et al., 2004), as the record has an acceptable resolution
owing to (i) an average annual snow accumulation of 0.05 to
0.10 m w.e. a 1, (ii) good age control (uncertainty of b5 yrs over the
studied time period), and (iii) was drilled at an elevation of 3233 m
and ~ 500 km inland from the coast, which reduces the inuence of
sea-salt sulphate (ss SO4) compared to records from coastal locations.
As the VLG ice core comes from a low elevation site that is only 22 km
from seasonally open ocean, marine aerosol input to the site is high
(Bertler et al., 2004b). We calculated the nss SO4 concentration
by subtracting proportionally ss SO4 contribution as estimated from
Na (assuming all Na is derived from sea salt) from total SO4 assuming

43

SO4 0:252Na ng=g:

The calculated VLG nss SO4 data render less than 2% negative
values, supporting the assumption that most of the Na is derived from
sea-salt and the calculated nss SO4 contribution is a reasonable
estimate.
Nss SO4 peaks are classied as volcanic eruptions if two criteria are
fullled: 1) the maximum peak concentration exceeds one standard
deviation above the mean (N1550 ppb) and 2) the nss SO 4
background increases above the median concentration of 375 ppb
for at least 10 yrs. These conditions are satised during ve events in the
record. The volcanic eruptions identied in the VLG record are: Agung
(1963 AD), Krakatau (1883 AD), Tambora (1816 AD), Huayanaputina
(1601 AD), and "Unkown" (1259 AD) (Fig. 2). The 1259 AD eruption has
been recorded in Antarctic and Arctic ice cores and is the largest
eruption in Antarctic records in the past millennium (Oppenheimer,
2003; Stothers, 2000) and hence provides a particularly well constrained age benchmark. The EPICA-DML record identied 26 eruptions
over the same time period. However most volcanic peaks, in addition to
the ve identied in the VLG record, are of lower amplitude (with
the exception of the Kuwae eruption, 1458 AD) and might be present in
the VLG record too, but are masked by the much higher ss SO4 input to
the site.
Using the identied volcanic eruptions, changes in snow accumulation can be taken into account. The peak associated with Agung
(1963 AD) was adjusted by 3 yrs, Krakatau (1883 AD) by +33 yrs,
Tambora (1816 AD) by + 29 yrs, Huayanaputina (1601 AD) by
60 yrs, and the 1259 AD eruption by 47 yrs (Fig. 2). The new
adjusted age model renders an average annual accumulation of the
whole record of 0.033 m w.e. a 1, which is in agreement with the
snow pit snow accumulation data of 0.033 0.013 m w.e. a 1 (Bertler
et al., 2004b). To estimate the upper dating error for any point
between age benchmarks, we add a 20% uncertainty to the maximum
age adjustment in the original 850 yr long record of 60 yrs, resulting
in 72 yrs. This maximum uncertainty is applied proportionally to
the length of the record between age ties. The age uncertainty is

Fig. 2. EDML and VLG nss SO4 records for the past 900 yrs, with identied volcanic eruptions used for tie points of the VLG record. Shaded areas highlight volcanic eruptions found in
both EDML and VLG nss SO4 records as dated in the EDML record.

44

N.A.N. Bertler et al. / Earth and Planetary Science Letters 308 (2011) 4151

smallest nearest an age benchmark ( 3 yrs) and largest at greatest


distance (midpoint) between age ties. The maximum dating error
within each of the now constrained sections is 3 yrs between
2000 and 1963 AD, 7 yrs between 1963 and 1883 AD, 24 yrs
between 1883 and 1601 AD, 29 yrs between 1601 and 1259 AD.
2.4. Isotopetemperature relationship
In the McMurdo Dry Valleys, most precipitation occurs during
summer (Bromwich, 1988), hence a stable isotope record from local
ice cores should provide a proxy for summer temperature (Bertler
et al., 2004b). In Fig. 3, summer temperatures (December to
February), as recorded by Scott Base and Lake Vida automatic weather
stations (AWS), are compared with VLG 18O snow pit data spanning
the past 36 yrs (Bertler et al., 2004a). The Lake Vida station provides
the longest continuous data in Victoria Valley beginning in the
1995/1996 austral summer (Doran et al., 1995). However, the overlap
between our record and the Lake Vida data is only 5 yrs. For this
reason, comparisons are also made with the more distant (~80 km),
but longer Scott Base summer temperature data. The two data sets
show signicant similarities, suggesting that the VLG isotope data
represent the regional summer temperature (Bertler et al., 2004a,b).
However, the comparison renders a low, albeit statistically signicant,
correlation coefcient of only R 2 = 0.35, (n = 30, p = 0.0006) indicating that signicant differences also exist. Whilst warmer (colder)
summers are coeveal in all three records, the magnitude of change
varies between sites. Scott Base is more exposed to southerly and/or
katabatic wind ow, which could explain some of the differences.
Accordingly, we also correlate monthly averaged NCEP/NCAR reanalysis data (Kalnay et al., 1996) for the nearest grid points (77S,
166.5E; 77.5S, 162.5E) to VLG (77.2S, 166.5E) and to Scott Base
(77.5S, 162.5E) with the VLG isotope and Scott Base DJF data,

respectively, for the 19792000 AD time period. Neither comparison


renders a statistically signicant correlations (R 2 = 0.01 and
R 2 = 0.03, respectively). The limited performance of the isotope
temperature correlation is likely due to: a) the relatively small
isotopic range of inter-annual summer temperature variability
compared to, e.g. the seasonal or glacial/interglacial amplitude,
which decreases the signal to noise ratio, b) intermittent precipitation
at Victoria Lower Glacier which biases the record towards precipitation days, c) the lack of long-term weather station data from the ice
core site, and d) the 0.5 resolution of the re-analysis data is
insufcient to adequately capture the large geographical variability
within the McMurdo Dry Valleys and the steep relief of the Trans
Antarctic Mountains.
In the absence of a suitable meteorological record to calculate the
local temperatureisotope conversion slope, no absolute temperatures
are deferred from the VLG record. Instead, we use previously published
isotopetemperature conversion slopes to estimate change of temperature (T). Masson-Delmotte et al. (2008) provided a comprehensive
review of stable isotope data from Antarctica and their interpretation.
Using the entire, quality assessed data set, they obtain a spatial
temperatureisotope conversion slope for 18O of 0.80 0.01 and
for D of 6.34 0.09, but note local slopes may vary by 20% or more.
For this reason, we discuss three isotopetemperature slopes from sites,
which share important similarities with the VLG record: Taylor Dome
and Talos Dome (Fig. 1) are the closest deep ice core records to VLG, and
Law Dome (Fig. 1, Map.1, No. 3), a low elevation, coastal ice core record
located in Wilkes Land. The Taylor Dome ice core was drilled at 2374 m
above sea level (asl) (Steig et al., 1998). The site lies about 120 km from
the Ross Sea coast and experiences an average annual temperature
(derived from 15 m rn cores) of 43 C, and has an average annual
snow accumulation of 0.050.07 m w.e.a 1 (Masson et al., 2000). The
Talos Dome ice core was drilled at 2316 m elevation, lies about 250 km

Fig. 3. VLG 18O snow pit data (red), Scott Base summer temperature (green), and Lake Vida summer temperature (blue) are shown since 1970 AD. Top left inset: Correlation
between 18O snow pit data and Scott Base summer temperature with correlation coefcient R2 = 0.35, n = 35, p = 0.0006. Top right inset: Correlation between VLG ice core 18O
and D data.

N.A.N. Bertler et al. / Earth and Planetary Science Letters 308 (2011) 4151

45

Fig. 4. Raw VLG Deuterium isotope (blue) and deuterium excess (green) data for the past 900 yrs. Values above average are shown in dark blue or dark green, values below average in
light blue or light green. Grey shaded area highlights LIA time period (1288 to 1807 AD) as identied by deuterium isotope record, preceded by the last 150 yrs of the Mediaeval
Warm Period (MWP) (1140 to 1287 AD), and followed by the Modern Era (ME) (since 1808 AD).

from the Ross Sea. The site experiences an average annual temperature
of 41 C, with an average annual snow accumulation of 0.05
0.11 m w.e.a 1 as measured from snow stakes and a long-term average
as obtained from the Talos Dome ice core of 0.08 m w.e. a 1 (Stenni
et al., 2002). The Law Dome record is situated in the Indian Ocean sector
of Antarctica. The ice core was drilled at 1390 m asl, has an average
annual temperature of 22 C, and an average annual snow accumulation of 0.66 m w.e. a 1 (Delmotte et al., 2000). To convert isotope data
into temperature changes from the Taylor Dome ice core record, a
temporal temperatureisotope slope was calculated (Steig et al., 1998).
The conversion is based on a 1.5 C cooling that occurred over the past
4000 yrs as observed from Taylor Dome borehole temperature
measurements and an associated decrease of D = 0.6 in the ice
core record (Steig et al., 1998), which suggests a temperature
conversion slope of D = 4.0 per C. The isotopetemperature
conversion slope for the Talos Dome record was established using the
spatial isotopetemperature slope between two sites (Talos Dome and
ST556), 50 km apart, with an elevation difference of 70 m, and an
average annual temperature difference of 2.9 C (Stenni et al., 2002).
Stenni and colleagues calculated from this relationship a local
conversion slope of 18O = 0.6 per C (converted for D = 4.8 per
C). The exceptionally high snow accumulation at Law Dome allowed for
a temporal isotopetemperature slope to be calculated based on the
regression over a ve year monthly mean temperature record from a
nearby automatic weather station and the mean seasonal isotope cycle
(Van Ommen and Morgan, 1997), which provided an isotope
temperature slope of 18O = 0.44 per C (converted for D = 3.5
per C). The difference between conversion slopes from the three sites is
D = 1.5 per C. None of the three sites matches ideally the
characteristics of VLG. Whilst the Law Dome record provides the most
coastal record of the sites and was drilled at a low elevation, a
comparison of Holocene climate variability of 11 ice cores from around
Antarctica (Masson et al., 2000) identied that the Ross Sea Sector
isotope records (e.g. Taylor Dome; Talos Dome was not included in this
study) clustered separately from East Antarctic sites (e.g. Law Dome).
Both Taylor and Talos Dome records are from the Ross Sea region, but
were drilled at much higher elevations (N2000 m) than VLG (626 m).
Annual average snow accumulation rates of 0.68 m w.e. a 1 at Law
Dome allow for exceptionally high resolution data. However, average
annual accumulation at Talos and Taylor Dome are more similar (0.05

0.10 m w.e.a 1) to VLG (0.03 m w.e. a 1) and hence are more likely be
sensitive to post-depositional processes that are likely to occur at VLG.
For the above reasons, we chose the most local conversion slope from
Taylor Dome D (4.0 per C) to provide an estimate for T.
3. Results
In Fig. 4 deuterium data (D) are used to reconstruct changes
summer temperature in the McMurdo Dry Valleys for the past 900 yrs.
Large multi-decadal and centennial variability is observed throughout
the record. Abrupt changes from cooler to warmer conditions (and
vice versa) of up to 50 occur in less than a decade. Whilst periods of
warmer than average summers (above 224.0) appear to reach
similar temperatures throughout the record, periods of colder than
average summers (below 224.0) are more frequent and reach
colder conditions during 1288 AD to 1807 AD. This period coincides
approximately with LIA time of 1300 to 1850 AD (Crowley, 2000;
Jones and Mann, 2004). Based on the stable isotope data, we dene
three distinct time periods in our record: the last 150 yrs of the
Mediaeval Warm Period (MWP, 1140 to 1287 AD), LIA (1288 to
1807 AD), and the Modern Era (ME, 1808 to 2000 AD) (Table 2).
During the LIA, 63% (37%) of the summers experienced below (above)
average temperatures compared to the ME and MWP with 28% (72%)

Table 2
Averages of characteristic parameters of VLG ice core data for ME, the LIA, and the last
150 yrs of the MWP.

No. of yrs
Warm summers % (no. of yrs)
Cold summers % (no. of yrs)
Average D ()
Summer temp (C) relative to ME
Average annual snow accumulation
(m w.e. a 1)
d excess()
Na/Cl
Na(ppb)
Fe (ppb)

ME 2000
1808 AD

LIA 1807
1288 AD

MWP 1287
1140 AD

192
72% (139)
28% (53)
220.1
0
0.03

519
37% (190)
63% (328)
228.0
2.0
0.04

147
71% (106)
29% (42)
218.7
+ 0.35
0.06

0.1
0.59
749
5.8

0.5
0.57
572
5.5

4.4
0.52
620
3.7

46

N.A.N. Bertler et al. / Earth and Planetary Science Letters 308 (2011) 4151

and 29% (71%), respectively (Table 2). In our characterisation, the ME


experiences one of the coldest episodes, which occurs around
1880 AD. Arguably, this could alternatively be interpreted as the
conclusion of the LIA. However, a prolonged episode of warming over
64 yrs from 1806 to 1870 precedes the cooling from 1870 to 1890. For
this reason, we argue that it is appropriate to conclude the LIA cooling
in the VLG record at 1807 AD. Using the Taylor Dome conversion for
ice cores from coastal sites of 4 per C (Mayewski et al., 2004a; Steig
et al., 1998), reconstructed summer temperatures are on average
~ 2 C cooler during the LIA (average D = 228.0) relative to the
past 200 yrs of the ME (D = 220.1). In addition, the nal 150 yrs
of the MWP were 2.3 C warmer (D = 218.7) than the LIA and
about 0.35 C warmer than the ME. The transitions from the MWP to
the LIA and from the LIA to ME occur rapidly in less than a decade.
However, the isotopic signature of snow and ice can be inuenced
by (i) changes occurring in the source region, (ii) a different source
area, (iii) modulations in the seasonality of snow precipitation, and
(iv) post-depositional processes (Masson-Delmotte et al., 2003;
Schlosser et al., 2008). Deuterium excess or dexcess (Dansgaard,
1964) of precipitated snow is sensitive to humidity (negative
correlation) and sea surface temperatures (SST) (positive correlation)
at the source region (Jouzel et al., 1982) and hence can be used to
reconstruct these parameters at a site with one dominant source
region, or to trace changes in air mass trajectory between different
source regions (Masson-Delmotte et al., 2008; Schlosser et al., 2008;
Sodemann et al., 2008; Vimeux et al., 2001). In Fig. 5, the correlation
between the VLG dexcess record and NCEP/NCAR reanalysis summer
(DecFeb) values for a) SSTs and b) relative humidity (Kalnay et al.,
1996) are shown from 1991 to 2000 AD. Correlation values 0.60 are
statistically signicant at the 95% level. The correlation with SSTs
shows statistically signicant, positive correlations between 50 and
70S for the Southern Ocean, with the exception of the Atlantic Sector,
where the correlation is negative. In addition, the positive correlation
extends north of the Ross Sea into the Pacic to 35S. The opposing
trend in the Atlantic Sector to East Antarctica has been attributed to
the inuence of the Southern Annular Mode and the El Nio-Southern
Oscillation (Kwok and Comiso, 2002; Thompson and Solomon, 2002;
Turner, 2004). The emerging correlation pattern demonstrates that
warmer SSTs in the Southern Ocean between 0E and 90W co-vary
with higher VLG dexcess data. The correlation with relative humidity

shows overall a more complex relationship but identies an expected


strong, statistically signicant anti-correlation with dexcess data in the
Pacic Sector. The correlation patterns for SST and relative humidity
support the assumption that snow precipitating at VLG over recent
times originates from the Southern Ocean via the Ross Sea as
suggested by Sinclair et al. (2010) based on air mass back trajectory
analysis, which suggests that synoptic and meso-scale cyclones
typically enter the eastern Ross Sea region and move northward in
the western Ross Sea (Fig. 5a, black arrows).
The VLG dexcess record exhibits high positive values during the nal
150 yrs of the MWP, which transitions sharply to negative values
during the LIA (Fig. 4). This could reect decreased humidity, warmer
SSTs in the Ross Sea and the Pacic Sector of the Southern Ocean, or a
signicant change in air mass trajectory between the MWP and LIA.
Air mass trajectory reconstructions from NCEP/NCAR and ERA40 reanalysis data (Sinclair et al., 2010) and from snow pit data (Bertler
et al., 2004b, 2005; Patterson et al., 2005), identied two distinct
source regions for precipitation in the McMurdo region: a) relatively
warm, humid, marine air masses that are transported by synoptic and
mesoscale cyclones from the Ross Sea and b) relatively cold, dry,
terrestrial air masses arriving as katabatic wind ow from the interior
of Antarctica. Sinclair et al. (2010) showed that in McMurdo Sound
only ~10% of the snow precipitation occurred during katabatic events,
whilst the remaining 90% was derived from cyclones originating in the
Ross Sea and beyond. As the VLG site lies near the McMurdo Dry
Valleys the largest ice-free region in Antarctica katabatic ow will
carry terrestrial aerosols and particulates. Thus geochemical tracers
for these terrestrial (e.g. Fe) and marine (e.g. Na) components can be
used to identify signicant changes in the relative contribution of
ocean and katabatic air masses (Legrand and Mayewski, 1997).
Sodium (Na) and chloride (Cl) are associated with marine air mass
sources with a marine Na/Cl ratio of 0.56 (Legrand and Delmas, 1988;
Legrand and Mayewski, 1997; Wolff et al., 1998a,b). Whilst the VLG
Na/Cl data vary, the average is 0.56 (s.d. = 0.17) (Fig. 6). This
suggests that marine air masses are important throughout the record
and their relative contribution remains stable.
Moreover, Na concentrations are lower by 24% during the LIA
compared to ME, whilst Fe concentrations are higher by 34% during the
LIA and ME compared to the MWP. In contract to records from the
Antarctic interior, the VLG snow pit data show maximum Na peaks

Fig. 5. Correlation between summer (December, January, February) VLG d excess data and NCEP/NCAR reanalysis data for a) SSTs and b) Relative Humidity for 1991 to 2000 AD.
Correlation values 0.60 are statistically signicant at the 95% level. Black arrows indicates idealised trajectory for synoptic and meso-scale cycloes after Sinclair et al. (2010). Images
provided by the NOAA/ESRL Physical Sciences Division, Boulder Colorado from their Web site at http://www.esrl.noaa.gov/psd/.

N.A.N. Bertler et al. / Earth and Planetary Science Letters 308 (2011) 4151

47

Fig. 6. Geochemical data from VLG, including A) deuterium data (blue), B) Sodium/Chloride ratio (Na/Cl, purple), C) Sodium (Na, green), and D) Iron (Fe, grey). The grey shaded area
highlights the LIA time period as dened by D.

associated with summer precipitation (Bertler et al., 2004b). This likely


reects VLG's proximity to seasonally open ocean (Fig. 1). For this reason,
the reduction in Na during the LIA time points to reduced open ocean,
either spatially (a smaller region of sea-ice break-out), and/or temporally
(fewer weeks with open water). The increase in Fe during the LIA also
supports cooler and windier conditions. However, increased Fe continues
through into the ME. For the past 40 yrs we observe a precipitation bias
towards summer precipitation, it is therefore unlikely that the Fe increase
indicates a seasonal shift in precipitation towards winter precipitation.
Alternatively, it could reect a response to a more vigorous atmospheric
circulation, which is capable of transporting more terrestrial particulates
in McMurdo Sound (Dunbar et al., 2009). In Section 4 we note that a
strengthening of the Southern Hemisphere westerly winds and a
deepening of the Amundsen Sea Low have indeed been reported
(Mayewski et al., 2009). Overall, the geochemical records suggest that
throughout the past millennium, the VLG site continued to receive
precipitation predominantly from the Southern Ocean, via the Ross Sea.
For this reason, we interpret the abrupt shift in dexcess not as a change in
air mass source but a signicant change in the Southern Ocean. Since an
abrupt change in SST within less than a decade is unlikely, we propose

that the sharp decrease in dexcess at the onset of the LIA was caused by a
signicant increase in sea ice (reduced Na) along with an increase in
katabatic wind strength (increased Fe) over the Ross Sea, and cooler SST
during summer (reduced open ocean, spatially and/or temporally).
Average annual snow accumulation is highest during the MWP
(0.06 m w.e.a 1) and lower during ME (0.03 m w.e.a 1) and LIA
(0.04 m w.e.a 1) (Table 2). In addition, the McMurdo Dry Valleys
were 0.35 C warmer during the MWP than during ME, accompanied
by warmer conditions in the Ross Sea, and increased snow
precipitation is perhaps indicative of less sea-ice.
Whether the changes observed in the McMurdo Dry Valleys were
initiated predominantly by changes in the Ross Sea/Southern Ocean
and/or atmosphere may highlight the mechanisms that initiate
DansgaardOeschger events. As discussed above, dexcess captures
changes in the Ross Sea/Southern Ocean, whilst D allows reconstruction of summer temperature changes in the McMurdo Dry
Valleys. An evolutionary power spectrum (Fig. 7) graphically displays
the changes of statistical signicant frequencies (non-black) contained in the dexcess and D data throughout the VLG record. The dexcess
frequency spectrum shows a clear change at the MWPLIA transition

Fig. 7. Evolutionary spectrum for annually resampled A) VLG d excess and B) VLG D data. Colours associated with values above 10 (non-black) are statistically signicant at the 95%
condence level. The chosen window width is 300 yrs, with an implemented step of 2 yrs.

48

N.A.N. Bertler et al. / Earth and Planetary Science Letters 308 (2011) 4151

from dominant frequencies in the ~ 60 yr to ~ 50 and ~ 40 yr


frequencies, which strengthen as they converge to a ~40 yr frequency
from 1600 to 1800 AD. During the LIAME transition, the frequency
weakens and shifts towards a ~ 45 yr rhythm. In contrast, the
evolutionary power spectrum of D changes gradually. Around
13001400 AD a ~ 40 yr frequency changes to ~ 30 yr and the
frequency band intensies and narrows. At 1450 AD, the frequency
splits into a narrow 30 yr band and a developing ~ 40 to 50 yr band.
Overall the changes around the MWP to LIA and LIA to ME are abrupt
in dexcess and gradual in the D data. Whilst the dating errors
associated with these data (329 yrs) limits the use of the actual
frequencies observed, D and dexcess data were measured on the same
sample and are therefore co-registered. For this reason, the abrupt
change in the dexcess frequency spectrum might suggest that initiation
for the changes observed in the VLG record reside in the ocean via an
increase in sea-ice (dexcess), initated by stronger katabatic wind (Fe),
leading to atmospheric cooling (D).
4. Climate conditions during the past millennium across Antarctica
To put our results into a wider context, we review previously
reported climate conditions across Antarctica. For this comparison,
the MWP is broadly dened as occupying 800 to 1300 AD, the LIA
1300 to 1850 AD, and the ME since 1850 AD, these time frames
capturing a range of published records afliated with the MWP and
LIA (e.g. Crowley, 2000). The focus is on ice core records, but includes,
where available, ice borehole measurements, marine and terrestrial
paleoenvironmental records, as well as modern data (Fig. 8). The
records only includes those where the published literature offers an
interpretation with regards to ME, MWP, and LIA trends, thus taking
advantage of the authors' insight regarding any limitations of the data.
In making this review, we recognise it is limited by dating uncertainties, regional and hemispherical variability of climate and
ocean, and imperfect knowledge of the range of climate drivers and
their interactions (Mayewski et al., 2009; Turner et al., 2009).
4.1. Mediaeval Warm Period
Our results suggest the McMurdo Dry Valleys experienced warm
summers, with increased snow accumulation and higher SSTs in the
Ross Sea/Southern Ocean, perhaps accompanied by less sea-ice (this
paper) (Fig. 8a).
Elsewhere, a magnetic susceptibility record from Palmer Deep
marine core (PD92 30MS) also supports warmer MWP conditions, this
time in Drake Passage (Domack and Mayewski, 1999). However, the
temperature amplitude of this warming is comparable to 10 other
warm events throughout the 2500 yr Palmer Deep record. In addition,
Masson et al. (2000) conducted a thorough review of the Holocene
records contained in seven Antarctic ice cores and identied nine
climate oscillations. The latest oscillation occurred around 1000 yrs
ago with a warm event followed by a cooling event. The Ca record
from the Siple Dome ice core suggests the MWP was windy following
a 4000 yr period of declining wind strength (Yan et al., 2005). Na
concentrations at Siple Dome also highlight the MWP. Na is a sensitive
recorder of the strength of the Amundsen Sea Low (ASL) (Kreutz et al.,
2000), a dominant feature of Antarctic meteorology (Mayewski et al.,
2009). Kreutz et al. (2000) showed that the ASL gradually intensied
over the past 10,000 yrs, but was weakest during the MWP.
4.2. Little Ice Age
In the Ross Sea region, our data reveal that the LIA (Fig. 8b) started
abruptly, in less than a decade, with colder SSTs in the Ross Sea,
perhaps accompanied with more extensive sea-ice, and 67% reduced
snow accumulation. In addition, McMurdo Dry Valleys' summer
temperatures were 2.3 C cooler than during the MWP, whilst higher

Fe concentrations in the VLG ice core suggest stronger katabatic ow


to the site. Lower Na concentrations are interpreted to reect perhaps
increased sea-ice and/or a shorter sea-ice break-out season.
As noted earlier, the last of the Holocene climatic oscillations
identied by Masson et al. (2000) comprised a cooling phase
coincident with LIA time. Likewise, a review by Stenni et al. (2002)
suggests a prolonged cooler climate from the 16th century to the
beginning of the 19th century. They note that the cooling varied from
place to place and is not temporally synchronous between sites.
Stenni et al. (2002) also note that snow accumulation at Talos Dome
decreased by 11% during the LIA compared to ME. They suggest that
the cooling in East Antarctic and the Ross Sea could lead to stronger
katabatic ow, which increased the efciency of polynyas to produce
more sea-ice as supported by the diatom assemblages (Leventer and
Dunbar, 1988). This implies more persistent katabatic winds in the
southwestern Ross Sea during 1600 to 1875 AD. Leventer and Dunbar
(1988) also argue that larger and/or more persistent polynas would
enhance production of High Salinity Shelf Water, a key component in
the production of Antarctic Bottom Water. Broecker et al. (1999) also
suggest that the LIA was likely a time of increased Antarctic Bottom
Water formation that may explain a ~ 10 ppm decrease in CO2
concentrations between 1550 AD and 1800 AD, as seen in the Law
Dome (Etheridge et al., 1996) and Taylor Dome (Indermuhle et al.,
1999) ice core records. At Palmer Deep, the PD92 30MS magnetic
susceptibility record also conrms a prolonged and distinctly cold LIA,
superimposed on a longer term cooling trend with 8 similar but less
cool climate oscillations (Domack and Mayewski, 1999). The LIA is
also accompanied by enhanced westerly winds (Yan et al., 2005),
whilst the ASL deepens (Kreutz et al., 2000); trends that continue
today. Furthermore, the East Antarctic High (EAH) decreases from
1200 to 1700 AD, out of phase with the ASL (Mayewski et al., 2004a).
Mayewski et al. (2004a) relate this trend to the poleward migration
of the circumpolar trough, which can be expressed as a positive
trend in the Southern Annular Mode. Between 1700 and 1850 AD,
however, the ASL and EAH are correlated, with large abrupt changes,
before they resume their anti-phase behaviour in modern times
(Mayewski et al., 2004a). The Atlantic sector of Antarctica has little
evidence of the LIA with perhaps the exception of ice core data from
Dronning Maud Land that suggest an 8% decrease in snow accumulation (Karlf et al., 2000).
Overall, it appears that Antarctica during the LIA experienced
cooler and drier conditions with higher wind speeds, cooler SSTs,
more extensive sea-ice and inferred bottom water increase.
4.3. Modern Era
For the past 200 yrs (Fig. 8c), temperatures in the McMurdo Dry
Valleys were on average 2 C warmer than conditions during the LIA, but
0.35 C cooler than during the nal 150 yrs of MWP. Our dexcess data
suggest that SST in the Ross Sea remained relatively cold, which is
accompanied by a 50% reduction in snow accumulation in comparison
to the MWP (and 24% reduction compared with the LIA). Increased Na
concentrations suggest that sea-ice might be reduced in contrast to the
LIA; Fe concentrations similar to LIA data suggest continuing strong
katabatic ow. Overall, whilst temperatures in McMurdo Sound indicate
milder conditions since the termination of the LIA, atmospheric
circulation and oceanic conditions appear to be still in LIA-mode.
With regard to other studies, Schneider et al. (2006) compiled ice
core temperatures for the last 200 yrs, which show that Antarctica as a
whole has warmed by 0.2 C. However, the warming is not uniform.
Law Dome, for example, experienced cooler conditions since 1750 AD
possibly in response to enhanced downwelling of stratospheric air
over the site (Mayewski et al., 2004a). In contrast, Siple Dome has
warmed markedly since about 1800 AD, which reect increased
intrusion of (warm) marine air masses into West Antarctica
(Mayewski et al., 2004a). At Talos Dome and in Dronning Maud

N.A.N. Bertler et al. / Earth and Planetary Science Letters 308 (2011) 4151

Fig. 8. Summary of conditions in Antarctica during the MWP, LIA, and ME. A) MWP: data for McMurdo Dry Valleys and Ross Sea (this paper); data for Law Dome, Dome C, Vostok, Dominion Range, Byrd, and Plateau Remote by Masson et al.
(2000); data for strength of the Amundsen Low by Kreutz et al. (2000); data for Southern Hemisphere Westerlies by Yan et al. (2005); data for Palmer Deep by Domack and Mayewski (1999). B) LIA: data for Ross Sea SST, humidity, and sea-ice
extent from geochemical ice core records (this paper); temperature data for McMurdo Dry Valleys from stable isotope records (this paper), temperature data for Law Dome, Dome C, Vostok, Dominion Range, Byrd, and Plateau Remote from
stable isotope records by Masson et al. (2000); temperature data for Talos Dome, EPICA Dome C, South Pole, and Taylor Dome from stable isotope records by Stenni et al. (2002); snow accumulation data for Talos Dome from ice core record by
Stenni et al. (2002); temperature data for Taylor Dome from borehole measurements by Broecker (2000b); data for Antarctic Bottom Water formation from ux calculations by Broecker (2000b); data for katabatic ow, diatom plume, size of
the polyna, and sea-ice from diatom data by Leventer and Dunbar (1988); data for strength of the Amundsen Low from geochemical ice core data by Kreutz et al. (2000); data for East Antarctic High from geochemical ice core data by Mayewski
et al. (2004a); data for Southern Hemisphere Westerlies from geochemical ice core data by Yan et al. (2005); data for SST from magnetic susceptibility record for Palmer Deep by Domack and Mayewski (1999); snow accumulation data for
Dronning Maud Land from ice core data by Karlf et al. (2000); data for atmospheric CO2 concentration by Etheridge et al. (1996) and Indermuhle et al. (1999). C) ME, which is dened here as the past 150 yrs: data for Ross Sea SST, humidity,
and sea-ice extent from geochemical ice core records (this paper); temperature data for McMurdo Dry Valleys from stable isotope records (this paper), temperature data for Talos Dome, EPICA Dome C, and Taylor Dome from stable isotope
records by Stenni et al. (2002); temperature data for Law Dome from stable isotope data by Mayewski et al. (2004a); snow accumulation data for Talos Dome from ice core record by Stenni et al. (2002); data for Antarctic Bottom Water
formation from ux calculations by Broecker (2000a); data for katabatic ow, diatom plume, size of the polyna, and sea-ice from diatom data by Leventer and Dunbar (1988); data for strength of the Amundsen Low from geochemical ice core
data by Kreutz et al. (2000); data for East Antarctic High from geochemical ice core data by Mayewski et al. (2004a); data for Southern Hemisphere Westerlies from geochemical ice core data by Yan et al. (2005); snow accumulation data for
Dronning Maud Land from ice core data by Karlf et al. (2000); data for snow accumulation at Gomez by Thomas et al. (2008); data for atmospheric CO2 concentration by Etheridge et al. (1996) and Indermuhle et al. (1999).

49

50

N.A.N. Bertler et al. / Earth and Planetary Science Letters 308 (2011) 4151

Land, snow accumulation increased by 11% (Stenni et al., 2002) and


8% (Karlf et al., 2000), respectively, whereas the Gomez ice core in
the Antarctic Peninsula recorded a doubling in snow accumulation
from 1855 to 2006 AD (Thomas et al., 2008). Broecker et al. (1999)
note that Antarctic Bottom Water formation decreased, which is
consistent with the longer term, glacialinterglacial cyclicity of
bottom water formation documented by Hall et al. (2001), and with
observed reductions in modern bottom water density (Aoki et al.,
2005). Diatom abundance also decreases in the southwestern Ross
Sea (Leventer and Dunbar, 1988), suggesting an easing of katabatic
ow, decreased size and duration of polynyas and hence a decrease in
sea-ice. The ASL is stronger than at any time over the past 10,000 yrs,
but it may be weakening as the EAH strengthens (Mayewski et al.,
2004a).

5. Concluding comments
Broecker (2000b, 2006) and Lund et al. (2006) suggested that
meridional overturning in the North Atlantic decreased during the LIA
thus amplifying cooling in Europe. According to the see-saw hypothesis
(Broecker, 1998; Broecker and Denton, 1989; Severinghaus, 2009) this
should have led to warmer conditions in Antarctica as observed for
DansgaardOeschger events (EPICA Community Members, 2006).
However, our data from the McMurdo Dry Valleys reveal the Ross Sea
region experienced colder conditions, stronger katabatic ow and
perhaps increased sea-ice. These data are generally consistent with an
Antarctic-wide assessment of LIA climate although there are temporal
and spatial differences in climatic responses. Casting the net wider, New
Zealand, which has direct climatic (Kidston et al., 2009; Ummenhofer
and England, 2007) and oceanic (Carter et al., 2008; Orsi and
Whitworth, 2005) links with Antarctica, also displays a cooling that
began around 12501350 AD and peaked ~1500 to 1650 AD before
recovering at the end of the 19th Century (Lorrey et al., 2008). Similarly,
southernmost South America underwent a LIA climate shift (Koch and
Kilian, 2005; Lamy et al., 2001).
In summary it appears, that the timing of the Antarctic LIA was
concomitant with that in the Northern Hemisphere, inferring that
the thermohaline circulation changes, as invoked to explain the
bipolar see-saw associated with DansgaardOeschger events, were
unlikely to be the prime driver of the LIA climate modulation. For
this reason we conclude that the LIA was caused a) by alternative
forcings (e.g. solar variability exacerbated by volcanic eruptions),
b) a see-saw mechanism that operated differently during warm
periods including perhaps non-linear thresholds (Capron et al.,
2010) or inherent lags (Goosse et al., 2004), or c) changes in
regional winds that affected oceanic circulation and heat transport
(Lozier, 2010; Toggweiler et al., 2006).

Acknowledgements
We would like to thank Antarctica New Zealand and Scott Base
for the logistical support. We are grateful for useful suggestions and
comments by two anonymous reviewers. This project was funded
the Foundation of Research Science and Technology via contracts
awarded to Victoria University of Wellington and GNS Science
(contracts VICX0704, CO5X0202, and CO5X0902). Monthly mean
NCEP re-analysis time series was obtained from the NOAA, Earth
System Research Laboratory (http://www.esrl.noaa.gov/psd/cgibin/data/timeseries/timeseries1.pl), hourly temperature data for
Scott Base were obtained from the National Institute for Water and
Atmospheric Research (http://clio.niwa.co.nz/index.html), and the
monthly mean temperature data for Lake Vida were obtained from
the Long-Term Ecological Research Programme (http://huey.
colorado.edu/LTER/datasets/meteorology/vida.html).

Appendix A. Supplementary data


Supplementary data to this article can be found online at doi:10.
1016/j.epsl.2011.05.021.
References
Ammann, C.M., Joos, F., Schimel, D.S., Otto-Bliesner, B.L., Tomas, R.A., 2007. Solar
inuence on climate during the past millennium: results from transient
simulations with the NCAR Climate System Model. Proc. Natl. Acad. Sci. 104,
37133718.
Aoki, S., Rintoul, S.R., Ushio, S., Watanabe, S., 2005. Freshening of the Adelie Land
bottom water near 140oE. Geophys. Res. Lett. 32, L23601.
Bard, E., Raisbeck, G., Yiou, F., Jouzel, J., 2000. Solar irradiance during the last 1200 years
based on cosmogenic nuclides. Tellus B 52, 985992.
Bertler, N.A.N., Barrett, P.J., Mayewski, P.A., Fogt, R.L., Kreutz, K.J., Shulmeister, J., 2004a.
El Nio suppresses Antarctic warming. Geophys. Res. Lett. 31.
Bertler, N.A.N., Mayewski, P.A., Barrett, P.J., Sneed, S.B., Handley, M.J., Kreutz, K.J., 2004b.
Monsoonal circulation of the McMurdo Dry Valleys signal from the snow
chemistry. Ann. Glaciol. 39, 139145.
Bertler, N.A.N., Barrett, P.J., Mayewski, P.A., Sneed, S.B., Naish, T.R., Morgenstern, U.,
2005. Solar forcing recorded by aerosol concentrations in coastal Antarctic glacier
ice, McMurdo Dry Valleys. Ann. Glaciol. 41, 5256.
Bond, G.C., Showers, W., Elliot, M., Evans, M., Lotti, R., Hajdas, I., Bonani, G., Johnson, S.,
1999. The North Atlantic's 12 kyr climate rhythm: relation to Heinrich events,
Dansgaard/Oeschger cycles and the Little Ice Age. In: Clark, P.U., Webb, R.S.,
Keigwin, L.D. (Eds.), Mechanisms of Global Climate Change at Millennial Time
Scales. AGU, Washington, D.C., pp. 3558.
Broecker, W.S., 1998. Paleocean circulation during the last deglaciation: a bipolar
seesaw? Paleoceanography 13, 119121.
Broecker, W.S., 2000a. Abrupt climate change: causal constraints provided by the
paleoclimate record. Earth Sci. Rev. 51, 137154.
Broecker, W.S., 2000b. Was a change in thermohaline circulation responsible for the
Little Ice Age? Proc. Natl. Acad. Sci. U. S. A. 97, 13391342.
Broecker, W.S., 2001. Paleoclimate: was the medieval warm period global? Science 291,
14971499.
Broecker, W.S., 2006. Abrupt climate change revisited. Global and Planetary Change
54, 211215.
Broecker, W.S., Denton, G., 1989. The role of oceanatmosphere reorganizations in
glacial cycles. Geochim. Cosmochim. Acta 53, 24652501.
Broecker, W.S., Sutherland, S., Peng, T.-H., 1999. A possible 20th-century slowdown of
southern ocean deep water formation. Science 286, 11321135.
Bromwich, D.H., 1988. Snowfall in the high southern latitude. Rev. Geophys. 26,
149168.
Capron, E., Landais, A., Chappellaz, J., Schilt, A., Buiron, D., Dahl-Jensen, D., Johnsen, S.J.,
Jouzel, J., Lemieux-Dudon, B., Loulergue, L., Leuenberger, M., Masson-Delmotte, V.,
Meyer, H., Oerter, H., Stenni, B., 2010. Millennial and sub-millennial scale climatic
variations recorded in polar ice cores over the last glacial period. Clim. Past 6,
345365.
Carter, L., Manighetti, B., Ganssen, G., Northcote, L., 2008. SW Pacic modulation of
abrupt climate change during the Antarctic Cold Reversal-Younger Dryas.
Palaeogeogr. Palaeoclimatol. Palaeoecol. 260, 284298.
Crowley, T.J., 2000. Causes of climate change over the past 1000 years. Science 289,
270277.
Dansgaard, W., 1964. Stable isotopes in precipitation. Tellus 16, 436468.
Delmotte, M., Masson, V., Jouzel, J., 2000. A seasonal deuterium excess signal at Law
Dome, coastal Antarctica: a southern ocean signature. J. Geophys. Res. 105,
71877197.
Domack, E.W., Mayewski, P.A., 1999. Bi-polar ocean linkages: evidence from lateHolocene Antarctic marine and Greenland ice-core records. Holocene 9, 247251.
Doran, P.T., Dana, G.L., Hastings, J.T., Wharton, R.A.J., 1995. McMurdo Dry Valleys LongTerm Ecological Research (LTER): LTER automatic weather network (LAWN).
Antarct. J. U.S. 30, 276280.
Doran, P.T., McKay, C.P., Clow, G.D., Dana, G.L., Fountain, A.G., Nylen, T., Lyons, W.B.,
2002a. Valley oor climate observations from the McMurdo dry valleys, Antarctica,
19862000. J. Geophys. Res. 107, 4772.
Doran, P.T., Priscu, J.C., Lyons, W.B., Walsh, J.E., Fountain, A.G., McKnight, D.M.,
Moorhead, D.L., Virginia, R.A., Wall, D.H., Clow, G.D., Fritsen, C., McKay, C.P., Parsons,
A.N., 2002b. Antarctic climate cooling and terrestrial ecosystem response. Nature
415, 517520.
Dunbar, G.B., Bertler, N.A.N., Mackay, R.M., 2009. Sediment ux through the McMurdo
Ice Shelf in Windless Bight. Antarct. Global Planet. Change 69, 8793.
EPICA Community Members, 2006. One-to-one coupling of glacial climate variability in
Greenland and Antarctica. Nature 444, 195198.
Etheridge, D.M., Steele, L.P., Langenfelds, R.L., Francey, R.J., Barnola, J.M., Morgan, V.I., 1996.
Natural and anthropogenic changes in atmospheric CO2 over the last 1000 years from
air in Antarctic ice and rn. J. Geophys. Res. 101, 41154128.
Gao, C., Robock, A., Ammann, C., 2008. Volcanic forcing of climate over the past
1500 years: an improved ice core-based index for climate models. J. Geophys. Res.
113, D23111.
Goosse, H., Masson-Delmotte, V., Renssen, H., Delmotte, M., Fichefet, T., Morgan, V., van
Ommen, T., Khim, B.K., Stenni, B., 2004. A late medieval warm period in the
Southern Ocean as a delayed response to external forcing? Geophys. Res. Lett. 31,
L06203.

N.A.N. Bertler et al. / Earth and Planetary Science Letters 308 (2011) 4151
Grove, J.M., 1988. The Little Ice Age. Methuen, London.
Hall, I.R., McCave, I.N., Shackleton, N.J., Weldon, G.P., Harris, S.E., 2001. Glacial
intensication of deep Pacic inow and ventilation. Nature 412, 809812.
Howard-Williams, C., Peterson, D., Lyons, W.B., Cattaneo-Vietti, R., Gordon, S., 2006.
Measuring ecosystem response in a rapidly changing environment: the Latitudinal
Gradient Project. Antarct. Sci. 18, 465471.
Indermuhle, A., Stocker, T.F., Joos, F., Fischer, H., Smith, H.J., Wahlen, M., Deck, B.,
Mastroianni, D., Tschumi, J., Blunier, T., Meyer, R., Stauffer, B., 1999. Holocene
carbon-cycle dynamics based on CO2 trapped in ice at Taylor Dome, Antarctica.
Nature 398, 121126.
Jacobs, S.S., 2004. Bottom water production and its links with the thermohaline
circulation. Antarct. Sci. 16, 427437.
Jones, P.D., Mann, M.E., 2004. Climate over past millennia. Rev. Geophys. 42 RG2002.
Jouzel, J., Merlivat, L., Lorius, C., 1982. Deuterium excess in an East Antarctic ice core
suggests higher relative humidity at the oceanic surface during the last glacial
maximum. Nature 299, 688691.
Kalnay, E., Kanamitsu, M., Kistler, R., Collins, W., Deaven, D., Gandin, L., Iredell, M., Saha, S.,
White, G., Woollen, J., Zhu, Y., Leetmaa, A., Reynolds, R., Chelliah, M., Ebisuzaki, W.,
Higgins, W., Janowiak, J., Mo, K.C., Ropelewski, C., Wang, J., Jenne, R., Joseph, D., 1996.
The NCEP/NCAR 40-year reanalysis project. Bull. Am. Meteorol. Soc. 77, 437471.
Karlf, L., Winther, J.G., Isaksson, E., Kohler, J., Pinglot, J.F., Wilhelms, F., Hansson, M.,
Holmlund, P., Nyman, M., Pettersson, R., Stenberg, M., Thomassen, M.P.A., van der Veen,
C., van de Wal, R.S.W., 2000. A 1500 year record of accumulation at Amundsenisen
western Dronning Maud Land, Antarctica, derived from electrical and radioactive
measurements on a 120 m ice core. J. Geophys. Res. 105, 1247112483.
Kidston, J., Renwick, J.A., McGregor, J., 2009. Hemispheric scale seasonality of the southern
annular mode and impacts on the climate of New Zealand. J. Climate 22, 47594770.
King, J.C., Turner, J., 1997. Antarctic Meteorology and Climatology, 1 ed. University Press
Cambridge, Cambridge.
Koch, J., Kilian, R., 2005. Little Ice Age glacier uctuations, Gran Campo Nevado,
southernmost Chile. Holocene 15, 2028.
Kreutz, K.J., Mayewski, P.A., Meeker, L.D., Twickler, M.S., Whitlow, S.I., Pittalwala, I.I.,
1997. Bipolar changes in atmospheric circulation during the Little Ice Age. Science
277, 12941296.
Kreutz, K.J., Mayewski, P.A., Pittalwala, I.I., Meeker, L.D., Twickler, M.S., Whitlow, S.I.,
2000. Sea level pressure variability in the Amundsen Sea region inferred from a
West Antarctic glaciochemical record. J. Geophys. Res. 105, 40474059.
Kwok, R., Comiso, J.C., 2002. Spatial patterns of variability in Antarctic surface
temperature: connections to the Southern hemisphere annular mode and the
Southern oscillation. Geophys. Res. Lett. 29 5051-5054.
Lamy, F., Hebbein, D., Rohl, U., Wefer, G., 2001. Holocene rainfall variability in southern
Chile: a marine record of latitudinal shifts of the Southern Westerlies. Earth Planet.
Sci. Lett. 185, 369382.
Legrand, M., Delmas, R., 1984. The ionic balance of Antarctic snow: a 10-year detailed
record. Atmos. Environ. 18, 18671874.
Legrand, M.R., Delmas, R.J., 1988. Formation of HCl in the Antarctic Atmosphere.
J. Geophys. Res. 93, 71537168.
Legrand, M., Mayewski, P., 1997. Glaciochemistry of polar ice cores: a review. Rev.
Geophys. 35, 219243.
Leventer, A., Dunbar, R.B., 1988. Recent Diatom Record of McMurdo Sound, Antarctica:
implications for history of sea ice extent. Paleoceanography 3, 259274.
Lorrey, A., Williams, P., Salinger, J., Martin, T., Palmer, J., Fowler, A., Zhao, J., Neil, H.,
2008. Speleothem stable isotope records interpreted within a multi-proxy
framework and implications for New Zealand palaeoclimate reconstruction. Quat.
Int. 187, 5275.
Lozier, M.S., 2010. Deconstructing the conveyor belt. Science 326, 15071511.
Lund, D.C., Lynch-Stieglitz, J., Curry, W.B., 2006. Gulf Stream density structure and
transport during the past millennium. Nature 444, 601604.
Maasch, K., Mayewski, P.A., Rohling, E., Stager, C., Karln, K., Meeker, L.D., Meyerson, E.,
2005. Climate of the past 2000 years. Geogr. Ann. 87A, 715.
Mann, M.E., Bradley, R.S., Hughes, M.K., 1999. Northern hemisphere temperatures during
the past millennium: inferences, uncertainties, and limitations. Geophys. Res. Lett. 26,
759762.
Masson, V., Vimeux, F., Jouzel, J., Morgan, V., Delmotte, M., Ciais, P., Hammer, C.,
Johnsen, S., Lipenkov, V.Y., Mosley-Thompson, E., Petit, J.-R., Steig, E.J., Stievenard,
M., Vaikmae, R., 2000. Holocene climate variability in antarctica based on 11 icecore isotopic records. Quat. Res. 54, 348358.
Masson-Delmotte, V., Delmotte, M., Morgan, V., Etheridge, D., van Ommen, T., Tartarin,
S., Hoffmann, G., 2003. Recent southern Indian Ocean climate variability inferred
from a Law Dome ice core: new insights for the interpretation of coastal Antarctic
isotopic records. Clim. Dyn. 21, 153166.
Masson-Delmotte, V., Hou, S., Ekaykin, A., Jouzel, J., Aristarain, A., Bernardo, R.T.,
Bromwich, D., Cattani, O., Delmotte, M., Falourd, S., Frezzotti, M., Gallee, H., Genoni,
L., Isaksson, E., Landais, A., Helsen, M.M., Hoffmann, G., Lopez, J., Morgan, V.,
Motoyama, H., Noone, D., Oerter, H., Petit, J.R., Royer, A., Uemura, R., Schmidt, G.A.,
Schlosser, E., Simoes, J.C., Steig, E.J., Stenni, B., Stievenard, D., Broeke, M.R.v.d., Wal,
R.S.W.v.d., Berg, W.J.v.d., Vimeux, F., White, J.W.C., 2008. A review of Antarctic
surface snow isotopic composition: observations, atmospheric circulation, and
isotopic modelling. J. Climate 21, 33593387.
Mayewski, P.A., Maasch, K.A., 2006. Recent warming inconsistent with natural
association between temperature and atmospheric circulation over the last
2000 years. Clim. Past Discuss. 2, 327355.
Mayewski, P.A., Meeker, L.D., Twickler, M.S., Whitlow, S.I., Yang, Q., Lyons, W.B.,
Prentice, M., 1997. Major features and forcing of high latitude northern hemisphere
atmospheric circulation over the last 110,000 years. J. Geophys. Res. 102 (26) 345326, 366.

51

Mayewski, P.A., Maasch, K.A., White, J.W.C., Steig, E.J., Meyerson, E., Goodwin, I.,
Morgan, V.I., van Ommen, T., Curran, M.A.J., Souney, J., Kreutz, K., 2004a. A 700 year
record of Southern Hemisphere extratropical climate variability. Ann. Glaciol. 39,
127132.
Mayewski, P.A., Rohling, E., Stager, C., Karln, K., Maasch, K., Meeker, L.D., Meyerson, E.,
Gasse, F., van Kreveld, S., Holmgren, K., Lee-Thorp, J., Rosqvist, G., Rack, F.,
Staubwasser, M., Schneider, R., 2004b. Holocene climate variability, quaternary
research. Quat. Res. 62, 243255.
Mayewski, P.A., Frezzotti, M., Bertler, N.A.N., van Ommen, T., Hamilton, G.S., Jacka, T.H.,
Welch, B., Frey, M., 2005. The International Trans-Antarctic Scientifc Expedition
(ITASE) an overview. Ann. Glaciol. 41, 180185.
Mayewski, P.A., Maasch, K., Yan, Y., Kang, S., Meyerson, E., Sneed, S., Kaspari, S., Dixon,
D., Morgan, V., van Ommen, T., Curran, M., 2006. Solar forcing of the polar
atmosphere. Ann. Glaciol. 41, 147154.
Mayewski, P.A., Meredith, M.P., Summerhayes, C.P., Turner, J., Worby, A., Barrett, P.J.,
Casassa, G., Bertler, N.A.N., Bracegirdle, T., Naveira-Garabato, A.C., Bromwich, D.,
Campbell, H., Hamilton, G.H., Lyons, W.B., Maasch, K.A., Aoki, S., Xiao, C., van
Ommen, T., 2009. State of the Antarctic and Southern Ocean Climate System
(SASOCS). Rev. Geophys. 47, 138.
O'Brien, S.R., Mayewski, P.A., Meeker, L.D., Meese, D.A., Twickler, M.S., Whitlow, S.I.,
1996. Complexity of Holocene climate as reconstructed from a Greenland ice core.
Science 270, 19621964.
Oppenheimer, C., 2003. Ice core and palaeoclimatic evidence for the timing and nature
of the great mid-13th century volcanic eruption. Int. J. Climatol. 23, 417426.
Orsi, A.H., Whitworth III, T., 2005. In: Sparrow, M., Chapman, P., Gould, J. (Eds.),
Hydrographic Atlas of the World Ocean Circulation Experiment (WOCE). : Southern
Ocean, vol. 1. International WOCE Project Ofce, Southampton, U.K.
Osterberg, E.C., Handley, M.J., Sneed, S.B., Mayewski, P.A., Kreutz, K.J., 2006. Continuous
ice core melter system with discrete sampling for major ion, trace element, and
stable isotope analyses. Environ. Sci. Technol. 40, 33553361.
Patterson, N.G., Bertler, N.A.N., Naish, T.R., Morgenstern, U., Rogers, K., 2005. ENSO
variability in the deuterium excess record of a coastal Antarctic ice core from the
McMurdo Dry Valleys, Victoria Land. Ann. Glaciol. 41, 140146.
Robock, A., 2000. Volcanic eruptions and climate. Rev. Geophys. 38, 191219.
Schlosser, E., Oerter, H., Masson-Delmotte, V., Reijmer, C., 2008. Atmospheric inuence
on the deuterium excess signal in polar rn: implications for ice-core interpretation. J. Glaciol. 54, 117124.
Schneider von Diemling, T., Ganopolski, A., Held, H., Rahmstorf, S., 2006. How cold was
the Last Glacial Maximum? Geophys. Res. Lett. 33, L14709.
Schneider, D.P., Steig, E.J., van Ommen, T.D., Dixon, D.A., Mayewski, P.A., Jones, J.M., Bitz,
C.M., 2006. Antarctic temperatures over the past two centuries from ice cores.
Geophys. Res. Lett. 33, L16707.
Severinghaus, J.P., 2009. Monsoons and meltdowns. Science 326, 240241.
Sinclair, K., Bertler, N.A.N., Trompetter, W.J., 2010. Synoptic controls on precipitation
pathways and snow delivery to high-accumulation ice core sites in the Ross Sea
region, Antarctica. J. Geophys. Res. 115, D22112.
Sodemann, H., Masson-Delmotte, V., Schwierz, C., Vinther, B.M., Wernli, H., 2008.
Interannual variability of Greenland winter precipitation sources: 2. Effects of
North Atlantic Oscillation variability on stable isotopes in precipitation. J. Geophys.
Res. 113, D12111.
Steig, E.J., Brook, E.J., White, J.W.C., Sucher, C.M., Bender, M.L., Lehman, S.J., Morse, D.L.,
Waddington, E.D., Clow, G.D., 1998. Synchronous climate changes in Antarctica and
the North Atlantic. Science 282, 9295.
Stenni, B., Proposito, M., Gragnani, R., Flora, O., Jouzel, J., Falourd, S., Frezzotti, M., 2002.
Eight centuries of volcanic signal and climate change at Talos Dome (East
Antarctica). J. Geophys. Res. 107, 4076.
Stothers, R.B., 2000. Climatic and demographic consequences of the massive volcanic
eruption of 1258. Clim. Change 45, 361374.
Thomas, E.R., Marshall, G.J., McConnell, J.R., 2008. A doubling in snow accumulation in
the western Antarctic Peninsula since 1850. Geophys. Res. Lett. 35, L01706.
Thompson, D.W.J., Solomon, S., 2002. Interpretation of recent Southern Hemisphere
climate change. Science 296, 895899.
Toggweiler, J.R., Russell, J.L., Carson, S.R., 2006. Midlatitude westerlies, atmospheric
CO2, and climate change. Paleoceanography 21 PA2005.
Traufetter, F., Oerter, H., Fischer, H., Weller, R., Miller, H., 2004. Spatio-temporal
variability in volcanic sulphate deposition over the past 2 kyr in snow pits and rn
cores from Amundsenisen, Antarctica. J. Glaciol. 50, 137146.
Turner, J., 2004. The El Nio-southern oscillation and Antarctica. Int. J. Climatol. 24, 131.
Turner, J., Bindschandler, R., Convey, P., di Prisco, G., Fahrbach, E., Gutt, J., Hodgson, D.,
Mayewski, P.A., Summerhayes, C.P., 2009. Antarctic Climate Change and the
Environment. UK, Scientic Committee on Antarctic Research, Cambridge, p. 526.
Ummenhofer, C.C., England, M.H., 2007. Interannual extremes in New Zealand
precipitation linked to modes of Southern Hemisphere climate variability.
J. Climate 20, 54185440.
Van Ommen, T., Morgan, V., 1997. Calibrating the ice core paleothermometer using
seasonality. J. Geophys. Res. 102, 93519357.
Vimeux, F., Masson, V., Jouzel, J., Petit, J.R., Steig, E.J., Stievenard, M., Vaikmae, R., White,
J.W.C., 2001. Holocene hydrological cycle changes in the Southern Hemisphere
documented in East Antarctic deuterium excess records. Clim. Dyn. 17, 503513.
Wolff, E.W., Hall, J.S., Mulvaney, R., Pasteur, E.C., Wagenbach, D., Legrand, M.R., 1998a.
Relationships between chemistry of air, fresh snow and rn cores for aerosol
species in coastal Antarctica. J. Geophys. Res. 103 11,057-011,070.
Wolff, E.W., Legrand, M.R., Wagenbach, D., 1998b. Coastal Antarctic aerosol and
snowfall chemistry. J. Geophys. Res. 103 10,927-910,934.
Yan, Y., Mayewski, P.A., Kang, S., Meyerson, E., 2005. An ice-core proxy for Antarctic
circumpolar zonal wind intensity. Ann. Glaciol. 41, 121130.

You might also like