Crown's Motion To Strike Zickfeld and Kniewasser Affidavits PNW LNG July 2017

Court File No.
T-1836-16
FEDERAL COURT
BETWEEN
SKEENAWILD CONSERVATION TRUST
APPLICANT
AND
HER MAJESTY THE QUEEN, ATTORNEY GENERAL OF CANADA,

MINISTER OF ENVIRONMENT AND CLIMATE CHANGE, CANADIAN
ENVIRONMENTAL ASSESSMENT AGENCY, and PACIFIC NORTHWEST
LNG LIMITED PARTNERSHIP
RESPONDENTS
MOTION RECORD OF THE RESPONDENTS, HER MAJESTY THE

QUEEN, THE ATTORNEY GENERAL OF CANADA, MINISTER OF
ENVIRONMENT AND CLIMATE CHANGE, and THE CANADIAN
ENVIRONMENTAL ASSESSMENT AGENCY
Re: Motion to Strike Affidavit Evidence of the Applicant
Volume 1 of 2
Counsel for the Respondent, Her Counsel for the Applicant,

Majesty the Queen, Attorney General SkeenaWild Conservation Trust
of Canada, Minister of Environment
and Climate Change, Canadian Chris Tollefson and Anthony Ho
Environmental Assessment Agency Suite 16 Shoal Point
21 Dallas Road
Judith Hoffman and Lisa Laird Victoria, BC V8V 4Z9
Department of Justice Phone: (250) 888-6074
900 840 Howe Street Email: ctollefson@pacificcell.ca
Vancouver, BC V6Z 2S9 anho@pacificcell.ca
Phone: (604) 775-7421
Fax: (604) 666-2710
Email: Judith.Hoffman@justice.gc.ca
Lisa.Laird@justice.gc.ca
Counsel for the Respondent, Pacific
Northwest LNG Partnership
Mark Andrews, QC and Bridget Gilbride

Fasken Martineau DuMoulin LLP
2900 - 550 Burrard Street
Vancouver, BC V6C 0A3
Phone: (604) 631-3115/(604) 631-4891
Fax: (604) 632-3115
E-mail:mandrews@fasken.com
bgilbride@fasken.com
i
MOTION RECORD
TABLE OF CONTENTS
TAB DOCUMENT PAGE
1. Notice of Motion 1
2. Affidavit #1 of Jane Tyler (affirmed June 12, 2017) 5
3. Written Representations 419
4. SkeenaWild Notice of Application (dated October 27, 2016) 439
5. Affidavit #1 of Kirsten Zickfeld (affirmed April 26, 2017) 455
Affidavit #1 of Maximilian Kniewasser (affirmed April 27,

6. 542
2017)
7. Book of Authorities (Index) 601

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
Court File No. T-1836-16
FEDERAL COURT
BETWEEN
APPLICANT
AND

MINISTER OF ENVIRONMENT AND CLIMATE CHANGE, CANADIAN
ENVIRONMENTAL ASSESSMENT AGENCY, and PACIFIC NORTHWEST
LNG LIMITED PARTNERSHIP
RESPONDENTS
MOTION RECORD OF THE RESPONDENTS, HER MAJESTY THE

QUEEN, THE ATTORNEY GENERAL OF CANADA, MINISTER OF
ENVIRONMENT AND CLIMATE CHANGE, and THE CANADIAN
ENVIRONMENTAL ASSESSMENT AGENCY
Re: Motion to Strike Affidavit Evidence of the Applicant
Volume 2 of 2
Counsel for the Respondent, Her Counsel for the Applicant,

Majesty the Queen, Attorney General SkeenaWild Conservation Trust
of Canada, Minister of Environment
and Climate Change, Canadian Chris Tollefson and Anthony Ho
Environmental Assessment Agency Suite 16 Shoal Point
21 Dallas Road
Judith Hoffman and Lisa Laird Victoria, BC V8V 4Z9
Department of Justice Phone: (250) 888-6074
900 840 Howe Street Email: ctollefson@pacificcell.ca
Vancouver, BC V6Z 2S9 anho@pacificcell.ca
Phone: (604) 775-7421
Fax: (604) 666-2710
Email: Judith.Hoffman@justice.gc.ca
Lisa.Laird@justice.gc.ca
Counsel for the Respondent, Pacific
Northwest LNG Partnership
Mark Andrews, QC and Bridget Gilbride

Fasken Martineau DuMoulin LLP
2900 - 550 Burrard Street
Vancouver, BC V6C 0A3
Phone: (604) 631-3115/(604) 631-4891
Fax: (604) 632-3115
E-mail:mandrews@fasken.com
bgilbride@fasken.com
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
This is the 1st affidavit of

Kirsten Zickfeld
made April 26, 2017
Court File No.: T-1836-16
FEDERAL COURT
BETWEEN:
Applicant
and

MINISTER OF ENVIRONMENT AND CLIMATE CHANGE,
CANADIAN ENVIRONMENTAL ASSESSMENT AGENCY, and
PACIFIC NORTHWEST LNG LIMITED PARTNERSHIP
Respondents
APPLICATION UNDER SECTIONS 18 & 18.1 OF THE FEDERAL COURTS ACT,

R.S.C. 1985, c. F-7
AFFIDAVIT
I, Kirsten Zickfeld, Associate Professor in the Department of Geography, Simon

Fraser University, of 2244 Ferndale Street, Vancouver, BC, V5L1Y5, AFFIRM
THAT:
1. I am a climate scientist who has reviewed the Canadian Environmental

Assessment Agencys environmental assessment report respecting the Pacific
NorthWest LNG Project (the CEAA Report), and as such I have personal
knowledge of the facts and matters hereinafter deposed to save and except
where the same are stated to be based upon information or belief, and where so
stated I verily believe the same to be true.
456
2
Qualification
2. I have been retained by SkeenaWild Conservation Trust as an expert in these

proceedings. The retainer between SkeenaWild Conservation Trust and myself
is attached to this affidavit as Exhibit A.
3. I am a climate scientist and Associate Professor in the Geography Department

at Simon Fraser University, which I joined in September 2010. From 2008 to
2010 I was a Research Scientist at the Canadian Centre for Climate Modelling
and Analysis in Victoria, BC. I hold a PhD in Physics (2004) from the
University of Potsdam, Germany. My research addresses questions pertaining
to the effects of human activities on climate. Specifically, I am interested in
exploring the long-term climate impacts of anthropogenic greenhouse gas
emissions and interactions between climate change and the carbon cycle. I am
internationally recognized for my work on carbon budgets and climate change
irreversibility. I have published forty scientific articles in peer-reviewed
journals, including high-impact journals such as Nature, Science, Nature
Geoscience, Nature Climate Change. I serve as Lead Author of the Special
Report of the Intergovernmental Panel on Climate Change (IPCC) on the 1.5
degree target and was Contributing Author to the Third (2001) and Fifth (2013)
Assessment Reports of the IPCC. My current curriculum vitae is attached to
this affidavit as Exhibit B.
What is the role and importance of cumulative effects in assessing the climate
impact of greenhouse gas emissions?
4. Scientific evidence suggests that the climate effects of emissions of long-lived

greenhouse gases (GHGs), such as carbon dioxide (CO2), are proportional to
the cumulative emissions of these gases 1, i.e. the total emissions added up over
1
IPCC (2013), Summary for Policymakers. In: Climate Change 2013: The Physical Science Basis.
Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on
Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA.
http://www.ipcc.ch/pdf/assessment-report/ar5/wg1/WG1AR5_SPM_FINAL.pdf
457
3
the time period over which they occur. This proportional relationship arises due
to the long atmospheric lifetime of these GHGs.
5. Human activities, such as combustion and production of fossil fuels,

deforestation and agriculture, release large amounts of GHGs into Earths
atmosphere. The principal GHGs released by human activities are carbon
dioxide (CO2), methane (CH4), nitrous oxide (N2O) and halocarbons (HCs).
These gases have different atmospheric lifetimes (i.e. remain in the atmosphere
for different amounts of time), depending on the natural processes that absorb
(sinks) or release (sources) these gases from/into the atmosphere.
6. CO2 stands out among these gases because its removal from the atmosphere
involves processes that operate on extremely long timescales. CO2 is
continuously exchanged between the atmosphere and its reservoirs on land
(vegetation, soils, freshwater), in the ocean and the Earths crust. Several
natural processes (collectively referred to as the carbon cycle) contribute to
this exchange: on land, photosynthesis by plants removes CO2 from the
atmosphere; CO2 fixed into plants is then cycled through plant tissues, dead
plant material and soil carbon and is released back to the atmosphere by plant
respiration, microbial decomposition of plant material in soils and disturbances
(e.g. fire). In the ocean, CO2 dissolves in surface waters and is transported to
deeper waters by ocean circulation and by the biological pump. Tiny
organisms (phytoplankton) at the ocean surface take up CO2 through
photosynthesis producing plant tissue that is consumed by organisms higher up
in the food chain. Litter (dead organisms, fecal matter) then sinks to deeper
ocean waters, where it is decomposed and releases CO2. These processes result
in an effective pumping of CO2 from the sea surface to the deep ocean,
removing it from contact with the atmosphere. CO2 is also exchanged between
the atmosphere and geologic reservoirs (carbonate sediments at the sea floor
and silicate rocks).
7. The processes involved in the cycling of carbon through the different reservoirs
operate on a wide range of timescales. While CO2 absorption and release by
458
4
plants and soils and CO2 dissolution in the ocean surface layer occur on
relatively fast timescales (seconds to centuries), transport of CO2 to the deep
ocean takes centuries to millennia and cycling of carbon through the geologic
reservoirs takes tens of thousands to millions of years. As a result, removal of
human-emitted CO2 from the atmosphere by natural processes will take
thousands of years. 15 to 40% of CO2 emissions will remain in the atmosphere
longer than 1000 years, depending on the amount of CO2 released (the
percentage of CO2 remaining in the atmosphere increases with larger amounts
of emissions) 2.
8. Due to its long atmospheric lifetime, CO2 accumulates in the atmosphere when
it is continuously emitted, leading to a rise in the atmospheric CO2
concentration. Due to this accumulation in the atmosphere, the long-term
climate effects of CO2 are controlled by the cumulative amount of CO2 emitted
over a specific period of time. For instance, it is now well established that the
global surface air temperature change in a specific year is proportional to the
cumulative CO2 emissions emitted up to that year1,3,4. In other words, the long-
term temperature effects of CO2 emissions depend only on the cumulative
amount of emissions and not on the time when they are emitted.
9. The other major GHGs emitted by human activities (CH4, N2O, HFCs) have
shorter atmospheric lifetimes than CO2 (about a decade for CH4, 120 years for
N2O and 1 to 100 years for most HFCs). Therefore, a smaller amount of these
gases accumulates in the atmosphere, and the climate effects of these gases are
2
Eby M, et al. (2009) Lifetime of anthropogenic climate change: Millennial time-scales of potential
CO2 and surface temperature perturbations. J Climate, 22:25012511. A copy of this article is attached
to this affidavit as Exhibit C.
3
Matthews, HD, N Gillett, PA Stott, and K Zickfeld (2009) The proportionality of global warming to
cumulative carbon emissions, Nature, 459: 829-833. A copy of this article is attached to this affidavit
as Exhibit D.
4
Zickfeld, K, M. Eby, HD Matthews, and AJ Weaver (2009) Setting cumulative emissions targets to
reduce the risk of dangerous climate change, Proceedings of the National Academy of Science,
106(38): 16129-16134. A copy of this article is attached to this affidavit as Exhibit E.
459
5
more closely related to the rate of emissions leading up to the time when the
impacts are assessed, than to cumulative emissions 5.
In its final report, did the CEA Agency assess the cumulative effects of the
greenhouse gas emissions associated with the Project?
10. In its report the CEA Agency acknowledges the importance of cumulative GHG
emissions in determining climate change: "...the Agency concurs with
Environment and Climate Change Canada that the Project would be one of the
largest greenhouse gas emitters in Canada and that the accepted science links
environmental effects globally and in Canada to cumulative greenhouse gas
emissions." 6
11. However, the CEA Agency assesses the significance of GHG emissions
associated with the project in terms of annual emission rates. It does so by
comparing Project emissions to 2014 provincial and national GHG emissions
levels 7. It estimates that [u]pstream greenhouse gas emissions associated with
the Project of 8.8 to 9.3 million tonnes CO2e per year would represent 1.2 to
1.3 percent of national emissions based on 2014 levels8. Considering direct
emissions of 4.5 million tonnes CO2e per year in addition to upstream
emissions, GHG emissions from the project would represent 1.8 to 1.9 percent
of national 2014 emissions levels 9. In its report, the CEA Agency does not
provide an assessment of the cumulative GHG emissions over the lifetime of
the Project (>30 years 10).
12. In a cumulative GHG emissions framework, cumulative Project GHG

emissions should be assessed, especially in relation to the long-lived GHG CO2,
5
Smith MS, et al. (2012) Equivalence of greenhouse-gas emissions for peak temperature limits, Nature
Climate Change 2:535538, doi:10.1038/nclimate1496. A copy of this article is attached to this
affidavit as Exhibit F.
6
CEAA Report, p. 43.
7
CEAA Report, p. 41.
8
CEAA Report, p. 41.
9
National Inventory Report 1990-2014: Greenhouse Gas Sources and Sinks in Canada - Executive
Summary https://www.ec.gc.ca/ges-ghg/default.asp?lang=En&n=662F9C56-1.
10
CEAA Report, p. 5.
460
6
in terms of their share of a provincial or national carbon budget. A carbon

budget defines the total CO2 emissions that can be emitted over all times in
order to limit warming to a given global mean temperature target.
13. Under the Paris Agreement Canada and other nations committed to limit global
temperature increase to ... well below 2C above pre-industrial levels, and to
pursue efforts to limit the temperature increase to 1.5C 11. To achieve a
climate target of below 2C of global warming, total global CO2 emissions need
to remain below 2,550 to 3,150 billion metric tonnes of CO2 12. Given that
approximately 2000 billion tonnes of CO2 have already been emitted, total
future global emissions of CO2 must not exceed 590 to 1,240 billion tonnes so
as to stay below 2C 13. At current global CO2 emissions of 40 billion metric
tons of CO2 per year this budget will last about 15 to 30 years, which illustrates
the urgency to reduce CO2 emissions to stay within a 2 degree Celsius-
compatible budget.
14. There is considerable uncertainty surrounding 2C carbon budget estimates due

to uncertainties in the climate system response to human-induced CO2
emissions and in our ability to mitigate emissions of non-CO2 GHGs. Carbon
budget estimates are therefore usually associated with a probability of meeting
a given target. The carbon budget estimate given in the previous paragraph
corresponds to a probability of limiting global warming to below 2C with a
larger than 66% likelihood (i.e. odds of two in three), a likelihood level that is
commonly used in cumulative CO2 emissions analysis (e.g. references in
footnotes 12, 13).
15. Different sharing principles can be applied for allocating a global carbon budget
11
Paris Agreement (2015), http://unfccc.int/paris_agreement/items/9485.php
12
IPCC (2014): Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III
to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. IPCC, Geneva,
Switzerland, 151 pp. (http://ar5-syr.ipcc.ch/). Table 2.2; the budget for Simple model, WGIII
scenarios and 66% of simulations meeting goal is given here.
13
Rogelj J, et al. (2016), Differences between carbon budget estimates unravelled, Nature Climate
Change, 6:245252, doi:10.1038/nclimate2868. A copy of this article is attached to this affidavit as
Exhibit G.
461
7
quota to nations. These range from allocating global carbon budget quotas
based on current emissions (emissions based) to equity-based principles that
require an equal-per-capita allocation (equity based) of the global carbon
budget. An approach that blends these two principles is considered the most
viable 14, but here the two sharing principles at opposite ends of the spectrum
are used to assess the full range of outcomes.
16. These two sharing principles (emissions-based and equity-based) can readily be
applied to calculate the remaining carbon budget for Canada that is consistent
with a >66% likelihood of limiting warming below 2C. This calculation yields
that Canadas remaining carbon budget is 3.6 billion metric tons of CO2 if it is
allocated according to Canadas fraction of the worlds population (equity-
based budget), or 14.4 billion metric tons of CO2 if the global carbon budget is
allocated according to Canadas fraction of the worlds CO2 emissions
(emissions-based budget) 15.
17. The CEAA Report estimates GHG emissions from the Pacific Northwest LNG
project at 13.3-13.8 million tons CO2 equivalent (CO2e) per year (4.5 million
tons direct, and 8.8-9.3 million tons from upstream CO2e emissions 16). A
cumulative GHG emissions analysis is especially helpful in understanding the
climate effects of long-lived GHGs such as CO2. Therefore only the CO2
emissions from the Project GHG emissions arising from CO2 are considered 17.
Direct Project CO2e emissions are generated almost exclusively by combustion,
and are therefore assumed to be emitted entirely as CO2. Upstream Project
14
Raupach R, et al. (2014), Sharing a quota on cumulative carbon emissions, Nature Climate Change,
4:873879, doi:10.1038/nclimate2384. A copy of this article is attached to this affidavit as Exhibit
H.
15
Donner S, and K Zickfeld (2016), Canadas contribution to meeting the temperature limits in the
Paris climate agreement. http://blogs.ubc.ca/sdonner/files/2016/02/Donner-and-Zickfeld-Canada-and-
the-Paris-Climate-Agreement.pdf; the calculation is based on a central global carbon budget estimate
of 850 billion metric tons of CO2 from 2015 for a >66% probability of meeting the target (reference in
footnote 12, Table 2.2, Complex models, RCP scenarios only). A copy of this article is attached to
this affidavit as Exhibit I.
16
CEAA Report, p. 41.
17
Non-CO2 GHG emissions are considered implicitly by using a global CO2 budget that has been
adjusted downwards to allow for about 20% of the temperature changes leading to 2C to occur as a
result of additional non- CO2 GHG emissions.
462
8
COze emissions include methane emissions from leaks, venting and other
sources. Informed by BC's Inventory Reportls, it is assumed that 160/o of
upstream COze emissions arise from methane emissions.
18. Applying the above assumptions it is estirnated that direct and upstream COz-
only emissions from the Project are 11.9-12.4 million tons COz. Over the
lifetirne of the project (here assumed to be 30 years), these annual emissions
add up to about 360 million metric tons of COz cumulative emissions. These
cumulative COz emissions correspond to l\Yo of Canada's equity-based carbon
budget and2.5o/o of Canada's emissions-based budget. This means that between
2.5o/o and II%o of Canada's all-tirne carbon budget consistent with limiting
warming to 2"C would be used up by the Pacific Nolthwest LNG project alone.
AFFIRMED BEFORE ME
in the City of Vancouver,
British Columbia
on Apri ,2017.
A commissioner for taki Kirsten Zickfeld

for British Columbia
Pak Yan Anthony HO

Barrister & Solicitor
Suite 16 Shoal Point
21 Dallas Road
Victoria, BC V8Y 429
Tel: (778) 678-3818
Email : anho @p acificcell. ca
r8
BC Provincial Greenhouse Gas Inventory Report.
(last
accessed April262017)
463
April 19,2017 Our file: 2016-09-0002
VIA ELECTRONIC MAIL

This is Exhibit "fl." refemed to in the
I affidavit of Kirsten Zickfeld,
Kirsten Zickfeld, Ph.D. before me on the
Associate Professor of April,2017.
Department of Geography
Simon Fraser University
8888 University Drive
Burnaby, BC V5A 156 A commissioner for ta.fting affidavits in
British Columbia
kzickfel@sfu.ca
Dear Dr. Zickfeld:
FfEz SkeenalYild Conservation Trust v. Her Majesty The Queen et aL

Expert Retainer Letter
We are counsel to SkeenaWild Conservation Trust ("SkeenaWild") in the above-referenced legal

proceeding before the Federal Court.
We confirm that you have agreed to provide an affidavit containing your expert opinion for this
proceeding. We are writing to set out the questions that we would like you to address in your
affidavit.
Material tr'acts
SkeenaWild is a charitable purpose trust whose goal is to make the Skeena River watershed and
nearby coastal communities a global model of sustainability. A significant portion of its work
involves the protection and conservation of salmon and salmon habitats within the Skeena River
watershed.
Pacific NorthWest LNG Limited Partnership (the "Proponent") is proposing to construct the
Pacific NorthWest LNG Project (the "Project"), which consist of a liquefied natural gas ("LNG")
liquefaction facility and export terminal on Lelu Island at the mouth of the Skeena River,
approximately 15 km southwest of the Prince Rupert, BC. The purpose of the Project is to
conveft processed natural gas from Progress Energy Canada Ltd. reserves into LNG for export to
the Pacific Rim markets in Asia.
The Project was subject to an environmental assessment under the Canadian Environmental
AssessmentAct,2012,S.C.2012, c. 19, s. 52(the "Act"). On September27,20l6,the following
occurred:
Suite 16 Shoal Point,2l Dallas Rd., Victoria BC, V8V 429

(25O) 7 2'l-81 70 | htto ://www. pac if iccell.ca I i nfo@oacif iccel l.ca
464
1. The Canadian Environmental Assessment Agency (the Agency), which was the
responsible authority for the environmental assessment of the Project under the Act,
released its final environmental assessment report (the Report) pursuant to subsection
25(2) of the Act. In the Report, the Agency concluded that the Project is likely to cause
significant adverse environmental effects as a result of greenhouse gas (GHG) emissions.
2. The Minister of Environment and Climate Change (the Minister) released a decision
statement pursuant to section 54 of the Act. Among other things, this decision statement
contained the following:
a. The Ministers decision under subsection 52(1) of the Act that the Project is likely to
cause significant adverse environmental effects, after considering the Report and the
implementation of mitigation measures that she had considered appropriate; and,
b. The Ministers decision under subsection 52(2) of the Act to refer to the Governor in
Council the matter of whether those significant adverse environmental effects were
justified in the circumstances.
3. The Governor in Council issued Order in Council P.C. 2016-0838 pursuant to subsection
52(4) of the Act, in which the Governor in Council determined that the significant adverse
environmental effects associated with the Project are justified in the circumstances.
On October 27, 2016, SkeenaWild filed a notice of application for judicial review at the Federal
Court challenging the Report, the Ministers decisions, and the Order in Council.
Relevant Documents
In connection with your report, we will provide copies of the following documents, which are
expected to be introduced as evidence in these proceedings:
1. The Report dated September 27, 2016; and,

2. A document by Environment and Climate Change Canada entitled Review of Related
Upstream Greenhouse Gas Emissions Estimates dated September 27, 2016, which was
submitted to the CEA Agency.
We may provide you with additional documents and other materials to review.
Code of Conduct for Expert Witnesses*****
As an expert witness, you owe a duty to the Court that overrides any duty to a party to these
proceedings. These duties are set out in the Federal Courts Code of Conduct for Expert
Witnesses. You must review this document, which is available here: http://www.laws-
lois.justice.gc.ca/eng/regulations/SOR-98-106/page-127.html#h-254.
We will ask you to accompany your report with Form 52.2, in which you will certify that you
have read the Code of Conduct for Expert Witnesses and agree to be bound by it.
Suite 16 Shoal Point, 21 Dallas Rd., Victoria BC, V8V 4Z9

(250) 721-8170 | http://www.pacificcell.ca | info@pacificcell.ca
Page 2 of 4
465
Questions
Based on the facts set out above, and your review of the documents we have provided and other
materials, please provide your professional opinion on the following:
1. What is the role and importance of cumulative effects in assessing the climate impact of
greenhouse gas emissions?
2. In its final report, did the CEA Agency assess the cumulative effects of the greenhouse gas
emissions associated with the Project?
In preparing your expert opinion, you should rely on any source that you consider to be reliable
in support of your opinion and on your own knowledge and experience.
Please feel free to offer any addition expert opinion beyond answers to the above questions that
you believe is necessary to provide a full and comprehensive expert opinion.
Form of Affidavit
In your affidavit, please:
1. State your full name and address;

2. Describe your areas of expertise and qualifications in relation to the issues addressed in
your affidavit;
3. State the facts and assumptions on which your opinions are based;
4. Provide your answers and opinions to the questions set out above, and your reasons for
those answers and opinions;
5. List any literature or other materials specifically relied on in support of your answers and
opinions;
6. If applicable, describe the methodology that you used in providing your answers and
opinions, including any examinations, tests or other investigations on which you have
relied, including (if applicable) details of the qualifications of the person who carried them
out, and whether a representative of any other party was present;
7. State any caveats or qualifications necessary to render your expert opinion complete and
accurate, including those relating to any insufficiency of data or research and an indication
of any matters that fall outside your field of expertise;
8. State, if applicable, the particulars of any aspect of your relationship with a party to the
proceeding or the subject matter of your proposed evidence that might affect your duty to
the Court;
9. Attach, as Exhibit A to your affidavit, this retainer letter;
10. Attach, as Exhibit B to your report, your most up-to-date curriculum vitae; and,
11. Attach, in successive Exhibits to your affidavit as needed, any other material you deem
relevant or necessary to render your expert opinion complete and accurate.

Page 3 of 4
466
Timeline
You must complete your report by April 28, 2017.
Privilege and Confidentiality
Please be advised that all communications between us, including this letter, are confidential and
privileged. However, if we introduce into evidence any report that you prepare, privilege is
waived. At that time, all correspondence between us, and any drafts of reports and related notes
may become available to other parties in the proceedings.
You understand that your work product may, at our discretion on the instructions of our client, be
shared with common interest parties, any other intervenors, and their legal representatives who
may ultimately be granted standing to join in these proceedings.
Yours truly,
_________________________ _________________________
Chris Tollefson Anthony Ho
Barrister & Solicitor Barrister & Solicitor
Counsel for the applicant Counsel for the applicant
Tel: (250) 888-6074 Tel: (778) 678-3818
Email: ctollefson@pacificcell.ca Email: anho@pacificcell.ca

Page 4 of 4
467
This is Exhibit ".ll"refened to in the
!,t affidavitof Kirsten Zickfeld,
before me on the Kirsten Zickfeld
day of Aoril, 2017.
Depaltment of Geography, Simon Frasel University
8888 University Drive, Bulnaby BC, V5A 156, Canada
Phone: 118-1829047 Faxl. 118-1825841
E-mail: kzickfel @ sfu.ca
A commissioner fo frg affidavits in
Educational Background
2004 Ph.D. Physics, Potsdam University, Germany
Modeling lalge-scale singular climate events for integrated assessment (Supervisor: H.J.
Schellnhuber)
1998 Diplom (M.Sc. equivalent)
Physics, Freie Universitiit Berlin, Germany
Femtosecond relaxation dynamics of small silicon clustels: structural change, fragmentation, and
metallization (Supervisor: K.-H. Bennemann)
1993 Vordiplom (undergraduate degree)
Physics, Freie Universitiit Berlin, Germany
Employment History at Academic Institutions

September 2015 Associate Professor', Department of Geography, Simon Fraser University,
Burnaby, BC
August 2010 - August 2015 Assistant Professor, Department of Geoglaphy, Simon Fraser University,
Burnaby, BC
July 2009 - June 2014 Adjunct Professor, School of Earth and Ocean Sciences, University of Victoria,
Victoria, BC
November 2008 - July 2010 Research Scientist, Canadian Centre fol Climate Modelling and Analysis,
Environment Canada, Victoria, BC
January 2006 - October 2008 Postdoctoral Research Associate, School ofEarth and Ocean Sciences,
University of Victoria, Victolia, BC
June 2005 - December 2005 Postdoctoral Research Associate (part time), Potsdam Institute for Climate
Impact Research, Potsdam, Germany
October' 2004 - 2005
June Maternity Leave
April 2004 - September 2004 Postdoctoral Resealch Associate, Potsdam Institute for Climate Impact
Research, Potsdam, Germany
January 1999 - March 2004 Research Associate, Potsdam Institute for Climate Impact Research, Potsdam,
Germany
January 1996 - June 1998 Research Assistant (palt time), Physics, Freie Universidt Berlin, Department of
Physics, Berlin, Germany
Updated: Oct27,2016
468
Kirsten Zickfeld Curriculum Vitae
____________________________________________________________
Other - Education and Employment Information

1993 - 1994
Exchange student, Faculty of Science and Engineering, University of Edinburgh, Scotland.
____________________________________________________________
Awards, Honors and Scholarships

2010 Title: SFU Endowed Research Fellowship Type: Fellowship
Organization: Dean's Office, Faculty of Environment, SFU
Details: Competitive research fellowship awarded to new faculty
2006 Title: Elf der Wissenschaft Type: Research
Organization: Sponsor Association for German Science and science magazine Bild der
Wissenschaft
Details: Elected among top eleven young scientists in Germany across all disciplines
2005 Title: Shortlisted for the Young Scientist Award of the Leibniz Association of Science
Type: Research
Organization: Gottfried Wilhelm Leibniz Association of Science (Germany)
Details: Nominated for the award by the Environmental Science Section of the Leibniz Association
2004 Title: Michelson Prize Type: Research
Organization: Potsdam University, Faculty of Mathematics and Science
Details: Awarded for best Ph.D. dissertation of academic year 2003/2004
2004 Title: Prize of the Society of Friends and Promoters of the Potsdam Institute for Climate Impact
Research Type: Research
Organization: Society of Friends and Promoters of the Potsdam Institute for Climate Impact Research
Details: Awarded for outstanding PhD thesis
____________________________________________________________
Current Research Interests
My primary research interests are in the dynamics of the Earth System under anthropogenic forcing. In particular, I
am interested in nonlinear phenomena such as feedback mechanisms and tipping points, the long-term response of
the climate system to anthropogenic emissions, carbon budgets compatible with climate targets such as the 1.5C
and 2C targets and the climate response to negative carbon emissions. My research is largely model- based. I use a
spectrum of models, ranging from conceptual models to complex Earth system models and integrated assessment
models, and methodologies including numerical simulation, stability theory, bifurcation analysis, inverse methods,
proportional control theory, and expert elicitation.
Keywords: Climate Change, Earth System Modelling, Climate-Carbon Cycle Feedbacks, Climate Change
Commitment and Reversibility, Cumulative Carbon Emissions, Climate Targets, Negative Emissions, Greenhouse
Gas Emission Pathways, Integrated Assessment Modelling
Updated: May 11, 2016 Page: 2

469
____________________________________________________________
Completed Works
*Student under my supervision
Refereed Journal Articles

35. Zickfeld, K., A.H. MacDougall and H.D. Matthews, 2016, On the proportionality between global temperature
change and cumulative CO2 emissions during periods of net negative CO2 emissions, Environmental Research
Letters, 11, 055006.
34. MacDougall, A.H., K. Zickfeld, R. Knutti and H.D. Matthews, 2015, Sensitivity of carbon budgets to
permafrost carbon feedbacks and non-CO2 forcings, Environmental Research Letters, 10, 125003,
doi:10.1088/1748-9326/10/12/125003.
33. Tokarska, K.B.*, and Zickfeld, K., 2015, The effectiveness of net negative carbon dioxide emissions in
reversing anthropogenic climate change, Environmental Research Letters, 10, 094013,
doi:10.1088/1748-9326/10/9/094013.
32. Zickfeld, K., and T. Herrington*, 2015, The time lag between a carbon dioxide emission and maximum
warming increases with the size of the emission, Environmental Research Letters, 10, 031001,
doi:10.1088/1748-9326/10/3/031001.
31. Herrington, T.*, and K. Zickfeld, 2014, Path independence of climate and carbon cycle response over a broad
range of cumulative carbon emissions, Earth System Dynamics, 5: 409-422.
30. Eby, M., A.J. Weaver, K. Alexander, K. Zickfeld, A, Abe-Ouchi et al., 2013, Historical and idealized climate
model experiments: An intercomparison of Earth system models of intermediate complexity, Climate of the Past
9(3): 1111-1140, doi: 10.5194/cp-9-1111-2013.
29. Zickfeld, K., M. Eby, K. Alexander, A.J. Weaver, E. Crespin et al., 2013, Long-term climate change
commitment and reversibility: An EMIC intercomparison, Journal of Climate, 26(16): 5782-5809, doi:
10.1175/JCLI-D-12-00584.1.
28. Zickfeld, K., V.K Arora, and N.P. Gillett, 2012, Is the climate response to carbon emissions path dependent?
Geophysical Research Letters, 39, L05703, doi:10.1029/2011GL050205.
27. *Kvale, K.F., K. Zickfeld, T. Bruckner, K.J. Meissner, K. Tanaka, and A.J. Weaver, 2012, Carbon dioxide
emissions pathways avoiding dangerous ocean impacts, Weather, Climate and Society, 4, 212-292, doi:
10.1175/WCAS-D-11-00030.1.
26. Weaver, A.J., Jan Sedlcek, M. Eby, K. Alexander, E. Crespin, T. Fichefet, G. Philippon-Berthier, F. Joos, M.
Kawamiya, K. Matsumoto, M. Steinacher, K. Tachiiri, K. Tokos, M. Yoshimori, K. Zickfeld, 2012, Stability of the
Atlantic Meridional Overturning Circulation: A Model Intercomparison. Geophysical Research Letters, 39, L20709,
doi:10.1029/2012GL053763.
25. Matthews, H.D., and K. Zickfeld, 2012, Climate response to zeroed emissions of greenhouse gases and aerosols,
Nature Climate Change 2, 338-341.
24. Zickfeld, K. , M. Eby, H.D. Matthews, A. Montenegro, A. Schmittner, and A.J. Weaver, 2011, Nonlinearity of
carbon cycle feedbacks, Journal of Climate, 24(6): 4254-4274.
23. Gillett, N.P., V. Arora, K. Zickfeld , S. Marshall, and B. Merryfield, 2011, Ongoing climate change following a
complete cessation of carbon dioxide emissions, Nature Geoscience, 4:83-87.
22. Zickfeld , K., M.G. Morgan, D.J. Frame, and D.W. Keith, 2010, Expert judgments about transient climate
response to alternative future trajectories of radiative forcing, Proceedings of the National Academy of Science, 107
(28): 12451-12456, doi:10.1073/pnas.0908906107.

470
21. Kuhlbrodt, T., S. Rahmstorf, K. Zickfeld , F. Vikebo, S. Sundby, M. Hofmann, P.M. Link, A. Bondeau, W.
Cramer, and C. Jaeger, 2009, An Integrated Assessment of Changes in the Thermohaline Circulation, Climatic
Change, 96: 489-537, doi:10.1007/s10584-009-9561-y.
20. Zickfeld , K., M. Eby, H.D. Matthews, and A.J. Weaver, 2009, Setting cumulative emissions targets to reduce
the risk of dangerous climate change, Proceedings of the National Academy of Science, 106(38): 16129-16134,
doi:10.1073/pnas.0805800106.
19. Eby, M., K. Zickfeld, A. Montenegro, D. Archer, K.J. Meissner, and A.J. Weaver, 2009, Lifetime of
anthropogenic climate change: Millennial life-times of potential CO2 and temperature perturbations, Journal of
Climate, 22: 2501-2511, doi:10.1175/2008JCLI2554.1.
18. Matthews, H.D., N. Gillett, P. A. Stott, and Zickfeld, K., 2009, The proportionality of global warming to
cumulative carbon emissions, Nature, 459, 829-833.
17. Bruckner, T., and Zickfeld, K., 2009, Low risk emissions corridors for safeguarding the Atlantic thermohaline
circulation, Mitigation and Adaptation Strategies for Global Change, 14, 61-83, doi:10.1007/s11027-008-9150-9.
16. Zickfeld, K., and T. Bruckner, 2008, Reducing the risk of Atlantic thermohaline circulation collapse: sensitivity
analysis of emissions corridors. Climatic Change, 91, 291-315, doi:10.1007/s10584-008-9467-0.
15. Zickfeld, K., M. Eby, and A.J. Weaver, 2008, Carbon-cycle feedbacks of changes in the Atlantic meridional
overturning circulation under future atmospheric CO2, Global Biogeochemical Cycles, 22, GB3024,
doi:10.1029/2007GB003118.
14. Zickfeld, K., J.C. Fyfe, M. Eby, and A.J. Weaver, 2008, Comment on Saturation of the Southern Ocean CO2
sink due to recent climate change. Science, 319, 570b.
13. Knopf, B., Zickfeld, K., M. Flechsig, and V. Petoukhov, 2008, Sensitivity of the Indian monsoon to human
activities. Advances of Atmospheric Sciences, 25(6), 932-945, doi: 10.1007/s00376-008-0932-5.
12. Weaver, A.J., Zickfeld, K., A. Montenegro, and M. Eby, 2007, Long term climate implications of 2050
emission reduction targets. Geophysical Research Letters, 34, L19703, doi:10.1029/2007GL031018.
11. Zickfeld, K., O.A. Saenko, M. Eby, J.C. Fyfe, and A.J. Weaver, 2007, Response of the global carbon cycle to
human-induced changes in Southern Hemisphere winds. Geophysical Research Letters, 34, L12712,
doi:10.1029/2006GL028797.
10. Fyfe, J.C., O.A. Saenko, Zickfeld, K., M. Eby, and A.J. Weaver, 2007, The role of poleward intensifying winds
on Southern Ocean warming. Journal of Climate, 20, 5391-5400.
9. Zickfeld, K., A. Levermann, M.G. Morgan, T. Kuhlbrodt, S. Rahmstorf, and D.W. Keith, 2007, Expert
judgments on the response of the Atlantic meridional overturning circulation to climate change. Climatic Change,
82(3-4): 235-265, doi:10.1007/s10584-007-9246-3.
8. Knopf, B., M. Flechsig, and Zickfeld, K., 2006, Multi parameter uncertainty analysis of a bifurcation point.
Nonlinear Processes in Geophysics, 13, 531-540.
7. Kropp, J.P., A. Block, F. Reusswig, Zickfeld, K., and H.-J. Schellnhuber, 2006, Semiquantitative Assessment of
Regional Climate Vulnerability: The North Rhine - Westphalia Study. Climatic Change, 76(3-4): 265-290,
doi:10.1007/s10584-005-9037-7.
6. Zickfeld, K., B. Knopf, V. Petoukhov, and H.-J. Schellnhuber, 2005, Is the Indian summer monsoon stable
against global change?, Geophysical Research Letters, 32, L15707, doi:10.1029/2005GL022771.
5. Rahmstorf, S., and Zickfeld, K., 2005, Thermohaline circulation changes: a question of risk assessment, Climatic
Change, 68 (1-2), 241-247.
4. Zickfeld, K., T. Slawig and S. Rahmstorf, 2004, A low-order model for the response of the Atlantic thermohaline
circulation to climate change, Ocean Dynamics, 54(1), 8-26.
3. Slawig, T., and Zickfeld, K., 2004, Parameter optimization using algorithmic differentiation in a reduced-form
model of the Atlantic thermohaline circulation, Nonlinear Analysis: Real World Applications, 5(3): 501-518.

471
2. Zickfeld, K., and T. Bruckner, 2003, Reducing the risk of abrupt climate change: emissions corridors preserving
the Atlantic thermohaline circulation, Integrated Assessment, 4(2), 106-115.
1. Zickfeld, K., M.E. Garcia, and K.H. Bennemann, 1999, Theoretical study of the laser-induced femtosecond
dynamics of small Sin clusters, Phys. Rev. B, 59(20), 13422-13430.
Reviewed Book Chapters

Contributing author to Church, J.A., P.U. Clark et al., 2013, Chapter 13: Sea Level Change, in: T.F. Stocker and D.
Qin (Eds.): Climate Change 2013: The Physical Science Basis, Contribution of Working Group I to the Fifth
Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge,
1137-1216.
Contributing author to Collins, M., R. Knutti et al., 2013, Chapter 12: Long-term Climate Change: Projections,
Commitments and Irreversibility, in: T.F. Stocker and D. Qin (Eds.): Climate Change 2013: The Physical Science
Basis, Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate
Change, Cambridge University Press, Cambridge, 1029-1136.
Contributing author to Smith, J.B., H.-J. Schellnhuber, M.Q. Mirza, S. Fankhauser, R. Leemans, L. Erda, L. Ogallo,
B. Pittock, R. Richels, C. Rosenzweig, U. Safriel, R.S.J. Tol, J. Weyant, G. Yohe, 2001, Vulnerability to Climate
Change and Reasons for Concern: A Synthesis, in: J.J. McCarthy, O.F. Canziani, N.A. Leary, D.J. Dokken, K.S.
White (Eds.): Climate Change 2001: Impacts, Adaptation, and Vulnerability, Contribution of Working Group II to
the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press,
Cambridge, 913-967.
Book Chapters
Kropp, J., K. Zickfeld, and K. Eisenack, 2002, Assessment and management of critical events: The breakdown of
marine fisheries and the North Atlantic thermohaline circulation. In: A. Bunde, J. Kropp, H.J. Schellnhuber (Eds.),
The science of disaster: climate disruptions, heart attacks, and market crashes, Springer, Berlin Heidelberg, 193-216.
Conference Proceedings
Bruckner, T., and K. Zickfeld, 2008, Inverse integrated assessment of climate change: the guardrail approach,
International conference on policy modelling, July 2-4 2008, Berlin, Germany.
Bruckner, T., and K. Zickfeld, 2004, Low risk emissions corridors for safeguarding the Atlantic thermohaline
circulation, Expert Workshop on Greenhouse gas emissions and abrupt climate change: positive options and robust
policy, September 30 - October 1, 2004, Paris, France.
Zickfeld, K., and T. Bruckner, 2002, Emissions corridors preserving the Atlantic Ocean thermohaline circulation,
In: A. E. Rizzoli, A.J. Jakeman (Eds.), Integrated assessment and decision support - Proceedings of the 1st biennial
meeting of the International Environmental Modelling and Software Society, 24-27 June 2002, Lugano, Switzerland,
145-150.
Technical Reports
Donner, S., and K. Zickfeld, 2016, Canadas contribution to meeting the temperature limits in the Paris climate
agreement. http://blogs.ubc.ca/sdonner/files/2016/02/Donner-and-Zickfeld-Canada-and-the-Paris-Climate-
Agreement.pdf
K. Zickfeld, 2011, Greenhouse gas emission and climate impacts of the Northern Gateway pipeline, Submission to
the Energy Board Joint Review Panel on the Northern Gateway Pipeline Project.
Theses
Zickfeld, K., 2004, Modeling large-scale singular climate events for integrated assessment, PhD thesis, Potsdam
University, 108 p.
Zickfeld, K.,1998, Femtosecond relaxation dynamics of small silicon clusters: structural change, fragmentation, and
metallization, Diploma thesis, Free University Berlin.

472
Textbooks
Zickfeld, K., 2015, Feedbacks between the carbon cycle and climate, in Environmental Change and Challenge,
P. Dearden and B. Mitchell, Oxford University Press, Fifth Edition.
Other
Zickfeld, K., Pressure is on Canada to present credible climate action plan, straight.com, Op-Ed, June 4 2014
http://www.straight.com/news/659931/kirsten-zickfeld-pressure-canada-present-credible-climate-action-plan
____________________________________________________________
Works Accepted for Publication

____________________________________________________________
Works Submitted for Publication


*Ehlert, D., K. Zickfeld, M. Eby, N. Gillett, 2016, The effect of variations in ocean mixing on the proportionality
between temperature change and cumulative CO2 emissions, Journal of Climate, in revision.
Zickfeld, K., S. Solomon, D.M. Gilford, 2016, Centuries of Thermal Sea Level Rise Due to Anthropogenic
Emissions of Short-Lived Greenhouse Gases, Proceedings National Academy of Sciences USA, in revision.
*Ehlert, D., K. Zickfeld, 2016, What determines the warming commitment after cessation of CO2 emissions?
Environmental Research Letters, submitted.
*Nzotungicimpaye, C.-M., K. Zickfeld, 2016, The contribution from methane to the permafrost carbon feedback,
Current Climate Change Reports, submitted.
____________________________________________________________
Works In Progress

Eby, M., H.D. Matthews, R. Pierrehumbert, and K. Zickfeld, 2016, What is the most policy-relevant carbon
budget?, to be submitted to Nature Climate Change, writing in final stages.
*Ehlert, D., K. Zickfeld, What determines the warming commitment after cessation of carbon dioxide emissions?,
to be submitted t earth System Dynamics, writing in progress.

473
Zickfeld, K., and S. Solomon, Reversibility of sea level rise after elimination of anthropogenic greenhouse gas
emissions, to be submitted to the Proceedings of the National Academy of Science, writing in progress.
Zickfeld, K., and *D. Azevedo, Exploring the symmetry of the climate-carbon cycle response to positive and
negative CO2 emission pulses, writing in progress.
Zickfeld, K., and *M. Wong, Sensitivity of the carbon cycle response to different rates of artificial carbon dioxide
removal, writing in progress.
*Rong, W.T., K. Zickfeld, V. Arora, Nonlinearity of land carbon cycle feedbacks: A CMIP5 model comparison, to
be submitted to Journal of Climate, based on completed honors thesis, writing in progress.
____________________________________________________________
Conferences, Workshops and Presentations

Invited Plenary Talks

January 2011 The role of oceans in anthropogenic CO2 uptake, National Council for Science and the
Environment Conference, January 19-21 2011, Washington DC, USA.
Invited Conference Presentations and Seminars

December 2015 Climate and carbon cycle commitments from early human agriculture, Fall Meeting of the
American Geophysical Union, 14-18 December 2015, San Francisco, CA, USA.
February 2015 The robustness of the proportional relationship of global warming to cumulative carbon
emissions, Sack Lunch Seminar Series, Department of Earth and Planetary Sciences,
Massachusetts Institute of Technology, Cambridge, MA, USA.
September 2014 Respondent to Keynote Lecture by Thomas Homer-Dixon: The Coming Global Energy
Transition, The Politics of Energy: Oils and Beyond?, University of Victoria, 26 September
2014, Victoria, BC, Canada.
January 2014 The challenge of stabilizing Earth's climate: A science perspective, Geography Brown-Bag
Seminar Series, January 23 2014, Simon Fraser University, Burnaby, BC, Canada.
December 2013 What climate changes do 400 ppm commit us to? Fall Meeting of the American Geophysical
Union, 9-13 December 2013, San Francisco, CA, USA.
October 2013 Meeting the 2 degree climate target: prospects and uncertainties, Centre for Coastal Science and
Management Open House, 3 October 2013, Simon Fraser University, Burnaby, BC.
May 2013 Sea level rise projections for the 21st century: Uncertainties and implications for coastal zone
management, Symposium on Coastal Management and Science: Perspectives from Canada and
China, 7 May 2013, Simon Fraser University, Burnaby, BC.
March 2013 Stabilizing the climate system at safe levels: prospects and uncertainties, Geography
Colloquium, 11 March 2013, Humboldt University of Berlin, Berlin, Germany.
March 2012 Cumulative Carbon as a Framework for Climate Stabilization, UBC Geography Colloquium,
March 13 2012, University of British Columbia, Vancouver, BC.
November 2011 Irriversibility of anthropogenic climate change on human timescales, Earth Science Seminar
Series, November 8 2011, Simon Fraser University, Vancouver, BC, Canada.
May 2011 Modelling Marine Climate-Carbon Cycle Feedbacks, Greencycles II Summer School, May 22
2011, Peyresq, France. Keynote Lecture.

474
January 2011 Climate response to carbon emissions: Characteristics and policy implications, Faculty of
Environment Seminar Series, January 25 2011, Simon Fraser University, Vancouver, BC,
Canada.
November 2010 Setting cumulative emissions targets to reduce the risk of dangerous climate change, Resource
and Environmental Management Seminar Series, November 19 2010, Simon Fraser University,
Vancouver, BC, Canada.
May 2010 Irreversibility of anthropogenic climate change on centennial timescales, Canadian Institute for
Advanced Research Workshop, May 5-6, 2010, Vancouver, BC, Canada.
February 2010 Climate response to carbon emissions: characteristics and policy implications, ESRL/CSD
Seminar Series, February 10, 2010, National Oceanographic and Atmospheric
Administration, Boulder, CO, USA.
January 2010 Setting cumulative emissions targets to reduce the risk of dangerous climate change, January 22,
2010, Department of Geography, Simon Fraser University, Vancouver, BC, Canada.
2009 Setting cumulative emissions targets to reduce the risk of dangerous climate change, School of
Earth and Ocean Seminar Series, September 29, 2009, University of Victoria, Canada.
2009 Quantifying uncertainty in allowable emissions consistent with temperature targets, MOCA-09:
IAMAS-IAPSO-IACS 2009 Joint Assembly, July 19-29, 2009, Montreal, Canada.
2009 Climate response to carbon emissions, PIK seminar series, May 18, 2009, Potsdam Institute
for Climate Impact Research, Potsdam, Germany.
2008 Greenhouse gas emissions targets to reduce the risk of dangerous climate change, Department of
Environmental Earth System Science Seminar Series, October 22, 2008, Stanford University,
Palo Alto, USA.
2008 Modelling the effects of future ocean circulation changes on the global carbon cycle, School of
Earth and Ocean Sciences Seminar Series, September 30, 2008, University of Victoria, Canada.
2008 Emissions pathways reducing the risk of dangerous climate change, Pacific Institute of
Mathematical Sciences Workshop, July 21-23, 2008, University of Victoria, Canada.
2008 The University of Victoria Earth System Climate Model - Description and Applications. School
of Resource and Environmental Management, Simon Fraser University, Vancouver, Canada,
June 10, 2008.
2008 Modelling the effects of future ocean circulation changes on the global carbon cycle, Earth,
Atmospheric and Planetary Sciences Department Lecture Series, May 21, 2008, Massachusetts
Institute of Technology, Massachusetts, MA, USA.
2008 Impacts of future ocean circulation changes on the global carbon cycle, Department of
Atmospheric and Oceanic Sciences Seminar, McGill University, Montreal, Canada. February
18, 2008.
2006 Carbon cycle feedbacks of changes in the Atlantic meridional overturning circulation. Topics in
the Atmosphere and Ocean Seminar Series, University of Victoria and Canadian Centre for
Climate Modelling and Analysis, November 21, 2006, Victoria, Canada.
Invited Public Lectures

March 2016 What does 1.5 degrees mean for the global and Canadian carbon budgets? Panel discussion,
March 31, 2016, SFU Harbour Centre, Vancouver, BC.
November 2013 Climate Change in Canada: Why we must care, Keynote Lecture, Saskatchewan Peoples
Climate Hearing, 2 November 2013, Saskatoon, Canada.
January 2012 Limit Climate Change and Stay Healthy, PICS Public Lecture Series, January 25 2012, UBC
Robson Square, Vancouver, BC.

475
Contributed Conference and Workshop Presentations - Lead Author

July 2015 Carbon Cycle Response to Artificial Atmospheric Carbon Dioxide Removal. Our Common
Future under Climate Change Conference, July 7-10, 2015, Paris, France. Poster presentation.
May 2015 Exploring the proportional relationship between global warming and cumulative CO2 emissions
for negative emission scenarios, Joint CMOS and AMS Congress on Polar Meteorology and
Oceanography, May 31-June 4, 2015, Whistler, BC, Canada. Oral presentation.
May 2015 The effectiveness of artificial atmospheric CO2 removal, University of Victoria Earth System
Climate Model Development Workshop, May 4-6, 2015, University of Victoria, BC, Canada.
Oral presentation.
April 2015 The Irreversibility of Sea Level Rise, 7th International Conference on Climate Change: Impacts
and Responses, April 10-11, 2015, Vancouer, BC, Canada. Oral presentation.
December 2014 Carbon Cycle Response to Artificial Atmospheric Carbon Dioxide Removal, AGU Fall
Meeting, December 15-19, 2014, San Francisco, USA. Poster presentation.
December 2014 Historical ocean heat and carbon fluxes: An Earth System model of intermediate complexity
intercomparison, CLIVAR workshop on Ocean's Carbon and Heat Uptake, December 12-14,
2014, San Francisco, USA. Poster presentation.
May 2013 Reversibility of CO2-induced climate change, Joint Scientific Congress of the CMOS, CGU and
CWRA, 26-30 May 2013, Saskatoon, Canada. Oral presentation.
December 2012 Climate change commitment and reversibility in EMIC AR5 simulations, AGU Fall Meeting,
December 3-7, 2012, San Francisco, USA. Poster presentation.
May 2012 Climate change commitment in EMIC AR5 simulations, Congress of the Canadian
Meteorological and Oceanographic Society, May 29-June 1 2012, Montral, Canada. Oral
presentation.
October 2011 Is the climate response to CO2 emissions path dependent? World Climate Research Program
Open Science Conference, 24-28 October 2011, Denver, CO, USA. Poster presentation.
June 2011 Is the climate response to CO2 emissions path dependent? Congress of the Canadian
Meteorological and Oceanographic Society, June 5-9 2011, Victoria, Canada. Oral presentation.
May 2010 Nonlinearity of carbon cycle feedbacks, Congress of the Canadian Meteorological and
Oceanographic Society, May 31-June 4 2010, Ottawa, Canada. Oral presentation.
2009 Expert judgments about transient climate response to radiative forcing, AGU Fall Meeting,
December 14-18 2009, San Francisco, USA. Poster presentation.
2009 Setting cumulative emissions targets to reduce the risk of dangerous climate change, IARU
Climate Change Congress, 10-12 March 2009, Copenhagen, Denmark. Oral presentation.
2008 Setting cumulative emissions targets to reduce the risk of dangerous climate change, AGU Fall
Meeting, December 15-19, 2008, San Francisco, USA. Oral presentation.
2008 Setting cumulative emissions targets to reduce the risk of 'dangerous' climate change, ESF-
FMSH Entre-Sciences Conference New Methodologies and Interdisciplinary Approaches in
Global Change Research, November 5-10, 2008, Porquerolles, France. Oral presentation.
2008 Negative feedback of poleward intensifying Southern Hemisphere winds on atmospheric CO2,
Congress of the Canadian Meteorological and Oceanographic Society, May 25-29 2008,
Kelowna, Canada. Oral presentation.
2007 Expert judgments about transient climate response and climate sensitivity, AGU Fall Meeting,
December 10-14, 2007, San Francisco, USA. Poster presentation.
2007 Impacts of changes in the Atlantic meridional overturning circulation on the global carbon
cycle, CMOS-CGU-AMS Congress, May 28-June 1, 2007, St Johns, Canada. Oral presentation.

476
2007 Emissions corridors reducing the risk of reorganizations of the Atlantic meridional overturning
circulation, European Geosciences Union (EGU) General Assembly, April 15-20, 2007, Vienna,
Austria. Oral presentation.
2006 Sensitivity of the Indian monsoon to human activities. Global Environmental Change Open
Science Conference, November 9-12, 2006, Beijing, China. Poster presentation.
2006 Sensitivity of the Indian monsoon to changes in boundary conditions. International Young
Scientists Global Change Conference, November 5-8, 2006, Beijing, China. Oral presentation.
2006 Emissions corridors reducing the risk of abrupt climate change. Rapid Climate Change
International Science Conference, October 24-27, 2006, Birmingham, UK. Poster presentation.
2006 Effects of potential changes in the Atlantic thermohaline circulation on the global carbon cycle.
Rapid Climate Change International Science Conference, October 24-27, 2006, Birmingham,
UK. Oral presentation.
2004 Safeguarding the Atlantic thermohaline circulation: cost-effective emissions paths and
emissions corridors. 1st General Assembly of the European Geophysical Union (EGU),
Acropolis, April 25-30, 2004, Nice, France. Oral presentation.
2004 Bifurcations in a reduced-form model of the Indian monsoon. 1st General Assembly of the
European Geophysical Union (EGU), Acropolis, April 25-30, 2004, Nice, France. Poster
presentation.
2003 Safeguarding the Atlantic thermohaline circulation: Sensitivity analysis of emission corridors.
International Conference on Earth System Modelling, Sept 15-19, 2003, Hamburg. Germany.
Poster presentation.
2003 Safeguarding the Atlantic thermohaline circulation: a sensitivity analysis of emission corridors.
EGS-AGU-EUG Joint Assembly, Acropolis, April 6-11, 2003, Nice, France. Oral presentation.
2002 Reducing the risk of dangerous climate change. NCCR-Climate Summer School, Sept 7-14,
2002, Grindelwald, Switzerland. Poster presentation.
2002 Reducing the risk of abrupt climate change: emissions corridors preserving the Atlantic
thermohaline circulation. 1st biennial meeting of the International Environmental Modelling and
Software Society (IEMSS), June 24-27, 2002, Lugano, Switzerland. Oral presentation.
2002 Reducing the risk of abrupt climate change: emission corridors preserving the thermohaline
circulation. Annual Meeting of the Center for the Integrated Study of Human Dimensions of
Global Change, May 28-31, 2002, Pittsburgh, USA. Oral presentation.
2002 Expert elicitation on the Atlantic thermohaline circulation. Annual Meeting of the Center for the
Integrated Study of Human Dimensions of Global Change, Carnegie Mellon University, May
28-31, 2002, Pittsburgh, USA. Oral presentation.
2002 Reducing the risk of abrupt climate change: emission corridors preserving the thermohaline
circulation. 27th General Assembly of the European Geophysical Society (EGS), Acropolis,
April 21-26, 2002, Nice, France. Oral presentation.
2001 Emissions corridors preserving the Atlantic ocean thermohaline circulation. Fall Meeting of the
American Geophysical Union (AGU), Dec 10-14, 2001, San Francisco, USA. Poster
presentation.
2000 A reduced-form model for the transient response of the thermohaline circulation to global
warming. Fnfte Deutsche Klimatagung, Oct 2-6, 2000, Hamburg, Germany. Poster
presentation.
2000 A reduced-form model for the transient response of the thermohaline circulation to global
warming. 23rd International conference on mathematical geophysics, La Citadelle, June 18-23,
2000, Villefranche sur Mer, France. Poster presentation.

477
Contributed Conference and Workshop Presentations - Co-author

April 2016 MacDougall, A.H., K. Zickfeld, R. Knutti and H.D. Matthews, 2015, Sensitivity of carbon
budgets to permafrost carbon feedbacks and non-CO2 forcings, EGU General Assembly, April
17-22, 2016, Vienna, Austria. Oral presentation.
July 2015 *Tokarska, K., and K. Zickfeld, The Irreversibility of Sea Level Rise, Our Common Future
under Climate Change Conference, July 7-10, 2015, Paris, France. Poster presentation.
May 2015 *Ehlert, D., K. Zickfeld, The effect of ocean mixing on heat and carbon fluxes and the linearity
between global warming and cumulative CO2 emissions, Joint CMOS and AMS Congress on
Polar Meteorology and Oceanography, May 31-June 4, 2015, Whistler, BC, Canada. Oral
presentation.
April 2015 *Ehlert, D., K. Zickfeld, The linearity between global warming and cumulative CO2 emissions:
examining the effect of ocean mixing, heat, and carbon fluxes, EGU General Assembly 2015,
April 12-17, 2015, Vienna, Austria. Oral presentation.
December 2014 *Tokarska, K., and K. Zickfeld, The role of artificial atmospheric CO2 removal in stabilizing
Earths climate, Fall Meeting of the American Geophysical Union, 15-19 December 2014, San
Francisco, CA, USA. Poster presentation.
December 2014 *Ehlert, D., K. Zickfeld, Exploring the role of ocean heat and carbon uptake in determining the
linear relationship between global warming and cumulative CO2 emissions, Fall Meeting of the
American Geophysical Union, 15-19 December 2014, San Francisco, CA, USA. Poster
presentation.
December 2014 *Ehlert, D., K. Zickfeld, The effect of mixing on ocean heat and carbon uptake and the linear
relationship between warming and cumulative emissions, CLIVAR workshop on Ocean's
Carbon and Heat Uptake, 12-14 December 2014, San Francisco, CA, USA. Poster presentation.
November 2014 *Ehlert, D., K. Zickfeld, Exploring the role of ocean heat and carbon uptake in determining the
proportional relationship between global warming and cumulative emissions, Graduate Climate
Conference, 31 October 2 November 2014, University of Washington, Pack Forest
Conference Centre, WA, USA. Poster presentation.
April 2014 *Tokarska, K., and K. Zickfeld, The role of net-negative CO2 emission scenarios in stabilizing
Earths climate, European Geoscience Union General Assembly, 27 April - 2 May 2014,
Vienna, Austria. Poster presentation.
December 2013 *Ehlert, D., K. Zickfeld, What determines the magnitude of climate change commitment after
cessation of emissions? Fall Meeting of the American Geophysical Union, 9-13 December
2013, San Francisco, CA, USA. Poster presentation.
October 2013 *Herrington, T., and K. Zickfeld, Path Dependence of Regional Climate Change, Centre for
Coastal Science and Management Open House, 3 October 2013, Simon Fraser University,
Burnaby, BC. Poster presentation.
October 2013 *Tokarska, K., and K. Zickfeld, The role of net-negative CO2 emission scenarios in stabilizing
Earths climate, Centre for Coastal Science and Management Open House, 3 October 2013,
Simon Fraser University, Burnaby, BC. Poster presentation.
May 2013 *Tokarska, K., and K. Zickfeld, The role of net-negative CO2 emission scenarios in stabilizing
Earths climate, Joint Scientific Congress of the CMOS, CGU and CWRA, 26-30 May 2013,
Saskatoon, Saskatchewan, Canada. Poster presentation.
Awarded the Campbell Scientific Canada Student Award.
May 2013 *Ehlert, D., A. Levermann, K. Zickfeld, Mechanism for potential strengthening of Atlantic
overturning prior to collapse, Joint Scientific Congress of the CMOS, CGU and CWRA, 26-30
May 2013, Saskatoon, Saskatchewan, Canada. Poster presentation.
April 2013 *Herrington, T., and K. Zickfeld, Path Dependence of Regional Climate Change, European
Geoscience Union General Assembly, 7-12 April 2013, Vienna, Austria. Poster presentation.

478
January 2013 *Herrington, T., and K. Zickfeld, Path Dependence of Regional Climate Change, PICS Climate
Change Research Poster Competition, 30 January 2013, Simon Fraser University, Burnaby, BC.
Poster presentation.
October 2012 *Herrington, T., K. Zickfeld, Cumulative CO2 Emissions Compatible with Avoiding Regional
Climate Tipping Points. Graduate Climate Conference, University of Washington, Pack Forest
Conference Centre, WA, October 26-28, 2012. Poster presentation.
____________________________________________________________
Research/Project Funding - Received

Contract/Grant: Operating Grant Awarded: 2011 Period: 2011 - 2016
Project Title: Modelling the Climate Response to Greenhouse Gas Emissions: Characteristics and
Policy Implications
Funding: NSERC Discovery Grant Type: External Annual: 33,000 Total: 165,000
Involvement: Principal Investigator
Contract/Grant: Research Grant Awarded: 2009 Period: 2009 - 2015

Project Title: Training Program in Interdisciplinary Climate Science
Funding: NSERC CREATE Program Type: External Annual: 289,700 Total: 1,602,300
Involvement: Co-Investigator Collaboration: Principal Investigator: A.J. Weaver, University of
Victoria. Co-Investigators: Arora V.K. (Environment Canada), Christian J.R. (Environment Canada),
Denman K.L. (Environment Canada), Flato G. (Environment Canada), Fyfe J.C. (Environment
Canada), Gillett N.P. (Environment Canada), Ianson D.C. (Department of Fisheries and Oceans),
Juniper S.K. (University of Victoria), Meissner K.J. (University of Victoria), Monahan A.H.
(University of Victoria), Saenko O.A. (Environment Canada)

Project Title: Modelling the climate response to anthropogenic greenhouse gas emissions
Funding: SFU Presidents Research Start-up Grant Type: Internal Total: 65,000

Project Title: Ocean Acidification and Cumulative Carbon Emissions
Funding: Endowed Research Fellowship, Faculty of Environment Type: Internal Total: 5,000
____________________________________________________________
International Project Activity

2010 - 2013 Research, Canada, USA, Japan, Belgium, Denmark, Germany, Switzerland, Russia
Project Title: Earth System Model Intercomparison Project in Support of the Fifth Assessment Report of the
IPCC (EMIC AR5)
Description: The project is a model intercomparison project conducted in support of the Fifth Assessment
Report of the IPCC by groups running Earth System models of intermediate complexity (EMICs). A set of
standardized model experiments was designed which were performed by all participating modeling groups.

479
Affiliated Institution(s): SFU, University of Victoria, University of Bern, Universite Catholique de

Louvain, University of Copenhagen, A.M. Obukhov Institute of Atmospheric Physics, Potsdam Institute for
Climate Impact Research, MIT, University of Minnesota, University of Maryland, University of Tokio, Japan
Agency for Marine-Earth Science and Technology
____________________________________________________________
Supervision of Research Personnel

May 2016 Full Time, Jessie Girling, Undergraduate Research Assistant
Funded by: NSERC USRA
Climate model data analysis
May 2016 Full Time, Kate Bujnowicz, Undergraduate Research Assistant

September 2015 Part Time, Deven Azevedo, Undergraduate Research Assistant
Funded by: NSERC Discovery
Climate modelling
September 2015 - Full Time, Claude-Michel Nzotungicimpaye, M.Sc, Research Assistant

Climate modelling and data analysis
September 2015 - Full Time, Margaret Valerio, BA, Research Assistant

May 2015 August 2015 Full Time, William Morgenstern, Undergraduate Research Assistant
Funded by: VPR USRA
Data collection, climate modelling
May 2015 August 2015 Full Time, Deven Azevedo, Undergraduate Research Assistant
January 2015 April 2015 Part Time, Deven Azevedo, Undergraduate Research Assistant
Funded by: NA
November 2014 January 2015 Part Time, Charles Wang, Undergraduate Research Assistant
Funded by: NA
Climate modelling
September 2014 - July 2015 Full Time, Edward Pollock, M.Sc, Research Assistant
May 2014 August 2014 Part Time, Michael Wong, Undergraduate Research Assistant
Funded by: President's Research Startup Grant
Climate model simulations and data analysis
September 2012 - Full Time, Dana Ehlert, M.Sc, Research Assistant

Funded by: NSERC CREATE/NSERC Discovery

480
September 2012 - August 2014 Full Time, Katarzyna Tokarska, B.Sc, Research Assistant
September 2013 - Part Time, Rong, Wei Tao, Undergraduate Research Assistant
December 2013 Funded by: President's Research Startup Grant
May 2013 - Part Time, Rong, Wei Tao, Undergraduate Research Assistant
August 2013 Funded by: SFU Work Study
September 2011 - December Full Time, Tyler Herrington, B.Sc, Research Assistant
2013 Funded by: NSERC CREATE
January 2010 - June 2010 Full Time, Sylvain Lassonde, M.Sc, Intern
Funded by: NA
September 2006 - Full Time, Karin Kvale, B.Sc, Research Assistant; co-supervised with A.
January 2009 Weaver
Funded by: NSERC Strategic Grant (awarded to A.J. Weaver, University of
Victoria)
Integrated assessment modelling
July 2005 - Full Time, Brigitte Knopf, M.Sc, Research Associate; co-supervised with H.
December 2005 Held
Funded by: German Science Foundation Grant (awarded to H. Held,
Potsdam Institute for Climate Impact Research)
Conceptual climate models, bifurcation analysis
June 2003 - Full Time, Brigitte Knopf, M.Sc, Research Associate; co-supervised with H.
September 2004 Held
Funded by: German Science Foundation Grant (awarded to H. Held,
Potsdam Institute for Climate Impact Research)
Conceptual climate models, bifurcation analysis
____________________________________________________________
Semesterly Activity at Simon Fraser University

Semester Type Course Number Session Type Hours Enrollment
2016-1 Teaching Climate Change GEOG414 D01.00 Lecture 2.00 11
2016-1 Teaching Climate Change GEOG414 D01.02 Laboratory 2.00 11

2016-1 Teaching Climatology GEOG614 G01.00 Seminar 2.00 3
2016-1 Teaching Climatology GEOG614 G01.01 Laboratory 2.00 3
2015-3 Teaching The Climate GEOG314 D01.00 Lecture 2.00 14

System
2015-3 Teaching Directed Readings GEOG691 Directed 1.00 1
Studies

481

Studies
System
2014-3 Teaching Weather and GEOG214 D01.00 Lecture 2.00 36
Climate
2014-3 Teaching Directed Studies ENV391 D02.00 Directed 1.00 1
Studies
2014-1 Teaching Climate Change GEOG414 D01.00 Lecture 2.00 14
2014-1 Teaching Climate Change GEOG414 D01.02 Laboratory 2.00 14

Climate
System
2013-2 Teaching Honours Thesis GEOG491 Undergrad. 1.00 1
Research
2013-1 Teaching Advanced GEOG414 D01.00 Lecture 2.00 9
Climatology
2013-1 Teaching Advanced GEOG414 D01.02 Laboratory 2.00 9
Climatology

System
Climate
Studies
2012-1 Teaching Climate and GEOG214 D01.00 Lecture 2.00 46
Environment
2012-1 Teaching Advanced GEOG414 D01.00 Lecture 2.00 10
Climatology
2012-1 Teaching Advanced GEOG414 D01.01 Laboratory 2.00 10
Climatology

Climate

482

Climate
2010-3 Teaching Climate and GEOG214 D01.00 Lecture 2.00 51
Environment
____________________________________________________________
Teaching at Another Canadian Institution
Semester Institution Course Number Type Hours Enrollment
2010-1 University of The Atmosphere-Ocean EOS550 Lecture 3.00 6

Victoria System
2009-1 University of The Atmosphere-Ocean EOS433/55 Lecture 3.00 3
Victoria System 0
2008-1 University of The Atmosphere-Ocean EOS433/55 Lecture 3.00 6
Victoria System 0
____________________________________________________________
Senior Supervisory Duties of a Thesis/Dissertation/or Major Project
Name Degree Project/Thesis Title Status Began Completed
Valerio, Margaret M.Sc. The effect of mixing on Southern Active 2015-3

Ocean water mass volume and
production
Li, Xinru M.Sc. Reversibility of ocean changes under Active 2015-3
net negative emissions scenarios
Nzotungicimpaye, Ph.D. Permafrost-methane feedback Active 2015-3
Claude-Michel
Pollock, Edward Ph.D. Sensitivity of ocean heat and carbon Withdrawn 2014-3 2015-2
fluxes to boundary layer
parameterizations
Aminipouri, Ph.D. Energy-balance modeling of cities Active 2013-3
Mehdi
Appointed senior supervisor in August 2015
Ehlert, Dana Ph.D. Climate response to anthropogenic Active 2012-3

greenhouse gas emissions
Thesis proposal defended in June 2014

483
Katarzyna, M.Sc. The role of artificial atmospheric Completed 2012-3 2014-2

Tokarska CO2 removal for climate
stabilization
Herrington, Tyler M.Sc. Dependence of Regional Climate Completed 2011-3 2013-3
Change on Greenhouse Gas
Emission Pathway
Rong, Nick Honors Nonlinearity of carbon cycle Completed 2013-2 2013-2
feedbacks - A CMIP5 model
intercomparison
____________________________________________________________
Serving on a Committee of a Thesis/Dissertation/or Major Project
Pulvicki, Alexandra M.Sc. Snow distribution on glaciers Active 2015-3

Rathay, Sarah M.Sc. Response of a fractured bedrock Active 2014-3
aquifer to recharge from heavy
rainfall events
Cunada, Christopher M.Sc. Seasonal Methane Dynamics in Active 2013-3
Mackenzie River Delta Lakes,
Northwest Territories
Committee joined in May 2014
Ho, Hung Chak Ph.D. Spatial influences of heat Completed 2013-1 2016-1
exposure and social vulnerability
on the temperature-mortality
relationship: A case study in the
greater Vancouver area
Hatch, Kristofer M.Sc. Integrating Soft Computing, Completed 2012-3 2014-1
Complex System methods and
GIS for modelling urban land-use
change
Bailey, Harry Ph.D. Addressing intermittency issues Completed 2011-3 2015-3
for renewable energy resources in
British Columbia from the supply
and demand perspectives
I served as internal external examiner for Harrys PhD thesis defense.
Jjumba, Anthony Ph.D. Towards 4D Spatiotemporal Completed 2011-3 2015-2

Modeling of Complex
Geographical Systems
Committee joined in November 2011

484
Holding, Shannon Ph.D. A spatio-temporal risk framework Completed 2011-3 2014-3

for water security assessment on
small islands
I served as external examiner for Shannon's PhD comprehensive exam
(incl. proposal defense) and as internal external examiner for her PhD
thesis defense.
____________________________________________________________
Graduate Supervision Outside SFU (incl. External Examinations)
Name Degree Status Dates Role
Matthews, Robin Ph.D. Completed September 2009 - December 2013 Committee Member
Area: Climate Science
Title: Seasonal variability of sea surface carbonate chemistry and temperature
Institution: University of Victoria
Brennan, Catherine Ph.D. Completed September 2006 - September 2012 Committee Member
Area: Climate Science
Title: Simulating oxygen isotopes in the UVic Earth System Climate Model
Notes: Committee joined in September 2009
Kumar, Ashwin Ph.D. Completed March 2007 - February 2012 External Examiner
Area: Climate Science and Decision Analysis
Title: Studies in climate prediction and decision: aerosol-monsoon effects, climate sensitivity, and
decision markets
Institution: Carngie Mellon University, Pittsburgh, PA
Kvale, Karin M.Sc. Completed September 2006 - January 2009 Committee Member
Area: Atmospheric Science
Title: Carbon dioxide emission pathways avoiding dangerous ocean impacts
____________________________________________________________
Supervision of Teaching Assistants
Level Name Title Course Tutorials Enrollment Began Completed
Doctorate Nzotungicimpaye The Climate GEOG314 3 14 2015-3 2015-3

Claude-Michel System
Doctorate Ehlert, Dana The Climate GEOG314 3 39 2015-1 2015-1
System

485
Level Name Title Course Tutorials Enrollment Began Completed
Doctorate Pollock, Edward Weather and GEOG214 2 36 2014-3 2014-3

Climate
Master Herrington, Tyler Climate GEOG414 1 14 2014-1 2014-1
Change
Master Herrington, Tyler Weather and GEOG214 3 49 2013-3 2013-3
Climate
Master Tokarska, Kasia The Climate GEOG314 2 18 2013-3 2013-3
System
Master Herrington, Tyler Weather and GEOG214 3 35 2012-3 2012-3
Climate
Doctorate Ehlert, Dana The Climate GEOG314 2 18 2012-3 2012-3
System
Master Herrington, Tyler Climate and GEOG214 2 28 2012-1 2012-1
Environment
Master Cross, Ben Climate and GEOG214 1 18 2012-1 2012-1
Environment
Master Cross, Ben Advanced GEOG414 1 10 2012-1 2012-1
Climatology
Master Cross, Ben Weather and GEOG314 1 14 2011-3 2011-3
Climate
Master Baird, Elizabeth Weather and GEOG314 2 16 2011-1 2011-1
Climate
Master Attard, Maureen Climate and GEOG214 3 51 2010-3 2010-3
Environment
____________________________________________________________
New Course Preparation and Course Enhancement

2014 Spring 2014, GEOG 414/614 (Climate Change/Climatology): I substantially revised course materials
to make them consistent with the newly released IPCC report: I replaced a large part of the course
readings, updated all my slides and revised my lecture plans. The total time commitment (in addition to
contact hours and administrative duties) was ~80 hours.
2013 Fall 2013, GEOG 214 (Weather and Climate): I revised the laboratory assignments, and revised my
lecture plans, slides and notes. Several adjustments were necessary so the course material could be
covered in 10 instead of 12 lectures (lectures were scheduled on Mondays, two of which were a
holiday). The total time commitment (in addition to contact hours and administrative duties) was ~70
hours.
2013 Fall 2013, GEOG 314 (The Climate System): I revised laboratory assignments, lecture plans, slides
and notes. The total time commitment (in addition to contact hours and administrative duties) was ~50
hours.
2013 Spring 2013, GEOG 414/614 (Advanced Climatology/Climatology): I substantially revised this course:
I replaced course readings, prepared new course materials, revised lecture plans and slides. I also
revised the laboratory assignments. The total time commitment (in addition to contact hours and
administrative duties) was ~80 hours.

486
2012 Fall 2012, GEOG 214 (Weather and Climate): I completely redesigned two of the laboratory
assignments, revised my lecture plans to increase student participation and revised lecture slides and
notes. The total time commitment (in addition to contact hours and administrative duties) was ~70
hours.
2012 Fall 2012, GEOG 314 (The Climate System): I revised two laboratory assignments, lecture plans,
slides and notes. The total time commitment (in addition to contact hours and administrative duties)
was ~60 hours.
2012 Fall 2012, GEOG 691 (Directed Readings): I compiled a reading list and designed three homework
assignments for this course. The total time commitment (in addition to one contact hour per week) was
~15 hours.
2012 Spring 2012, GEOG 214 (Climate and Environment): I revised the plans, notes and slides for several
lectures and substantially modified two of the laboratory assignments. The total time commitment (in
addition to contact hours and administrative duties) was ~60 hours.
2012 Spring 2012, GEOG 414 (Advanced Climatology): This course was an entirely new development. I
chose articles from the research literature for discussion in class, designed in-class assignments
(including a climate conference simulation), wrote plans and prepared slides and other materials for 12
2-hour lectures. For the laboratory component of the course I designed 4 assignments and a set of term
projects aimed at introducing students to climate modelling and climate data analysis. The total time
commitment (in addition to contact hours and administrative duties) was ~200 hours.
2012 Spring 2012, GEOG 614 (Climatology): I offered this course in combination with GEOG 414
(Advanced Climatology). I designed extra laboratory exercises for graduate students and designed
modelling projects in accordance with students research interests. The total time commitment was ~30
hours.
2011 Fall 2011, GEOG 314 (Weather and Climate): I undertook a major revision of the plans, notes and
slides for 12 lectures. I also designed several new laboratory exercises and two new exams. The total
time commitment (in addition to contact hours and administrative duties) was ~150 hours.
2011 Spring 2011, GEOG 314 (Weather and Climate): This course was an entirely new development and its
preparation proved particularly challenging because no text book covers the course topics at a level
adequate for students in the class. I wrote lecture plans and notes and prepared slides for 12 2-hour
lectures. I also developed 6 laboratory assignments and designed two exams. All materials were
completely original. Lecture notes and slides were made available to students for download through
the course container on WebCT. The total time commitment (in addition to contact hours and
administrative duties) was ~300 hours.
2010 Fall 2010, GEOG 214 (Climate and Environment): This course was the first atmospheric science
course at the 200 level I taught and required very extensive preparation: I wrote lecture plans and notes
and prepared lecture slides for 12 2-hour lectures. All lecture notes and slides were completely original
material and were made available to the students for download through WebCT. I also prepared 6
laboratory assignments, which were modified from earlier versions provided by O. Hertzman, and
designed two exams. The time commitment (in addition to contact hours and administrative duties)
was ~300 hours.
____________________________________________________________
Teaching and Other Professional Development

2014 March 5-7 2014: Advanced Science Communication Workshop, facilitated by Nancy Baron
2013 Fall 2013: Graduate Supervision Brownbag Series, organized by the Associate Dean of Graduate
Studies
2013 January 17, 24 2013: SFU Media Training Workshop (2x3 hours)

487
2012 August 26-31 2012: Leadership Workshop, facilitated by Robert Gass

2011 Spring 2011: "Clear Speach" workshop organized by the SFU Teaching and Learning Center (4 parts,
6 hours total)
May 2-4 2011: Instructional Skills Workshop (ISW) offered by the SFU Teaching and Learning Center
Fall 2011: Best Practices in Graduate Supervision Workshop organized by the Associate Dean of
Graduate Studies
2010 Fall 2010: Discussion series organized by the SFU Teaching and Learning Center for new faculty.
Discussion topics included: Creating the classroom culture, Academic skills and support for
students, Who are todays SFU students?, Academic support for instructors (4 parts, 6 hours).
____________________________________________________________
Active Service to Simon Fraser University

Departmental Committees
November 2015 - May 2016 Member, Search Committee, Surface Hydrology
November 2015 - April 2016 Member, Search Committee, Physical Geography Lecturer
September 2015 - Current Member, Appointments Committee
May 2015 - May 2016 Member, Tenure and Promotion Committee
September 2013-February 2014 Organizer, Geography Distinguished Lecturer Visit
September 2014 - August 2015 Member, Graduate Studies Committee
September 2013 - August 2015 Member, Speaker Series Committee
May 2013 - May 2014 Member, Tenure and Promotion Committee
September 2012 - August 2013 Member, Graduate Studies Committee
September 2011 - August 2012 Member, Undergraduate Studies Committee
September 2011 - August 2012 Member, Resources Committee
May 2011 - April 2012 Member, Tenure and Promotion Committee
Faculty Committees
May 2015 - Current Co-organizer, Solutions and You: Combating Climate Change Lecture Series;
hosted speakers Michael Mann and Jana Sillmann
August 2014 April 2015 Co-organizer, Climate & Energy Research Day, April 14 2015
May 2013 - October 2013 Member, Global Environmental Systems Major Design Committee
University Committees
November 2015 Current Program Committee Member, Pacific Institute for Climate Solutions
November 2012 - November 2012 Department of Geography representative, NSERC postgraduate fellowship
adjudication committee

488
____________________________________________________________
Active Service to the Academic Community

Membership of International Panels
January 2010 - June 2013 Contributing author to the Fifth Assessment Report of the Intergovernmental
Panel on Climate Change (IPCC)
January 1999 - May 2001 Contributing author to the Third Assessment Report of the Intergovernmental
Panel on Climate Change (IPCC)
Committee Membership
July 2015 - Current Member, International Scientific Steering Committee, Carbon Dioxide Removal
Model Intercomparison Project (CDR-MIP)
July 2013 - Current Member, Scientific Committee of the Canadian Meteorological and
Oceanographic Society (CMOS).
2010 - Current Expert Council Member, Climate Collaboratorium (Climate CoLab),
Massachusetts Institute of Technology.
Conference Organizer
January 2010 - June 2011 Member (Communication Lead), Local Arrangements Committee for the 45th
Congress of the Canadian Meteorological and Oceanographic Society (CMOS),
Victoria, BC, June 5-9 2011.
Session Organizer
November 2015 Co-convener, Climate-Carbon Cycle Interactions, Joint Congress of the
Canadian Meteorological and Oceanographic Society and the Canadian
Geophysical Union, May 29 June 2 2016, Fredericton, NB.
November 2014 June 2015 Co-convener, From Carbon Emissions to Climate Change, Congress of the
Canadian Meteorological and Oceanographic Society, May 31 June 4 2015,
Whistler, BC.
April 2014 December 2014 Co-convener, Exploring the nature, limits and policy implications of the climate
response to cumulative CO2 emissions, AGU Fall Meeting, December 15-19
2014, San Francisco, CA.
November 2012 - May 2013 Co-convener, Climate change and the carbon cycle, Joint Scientific Congress of
the CMOS, CGU and CWRA, 26-30 May 2013, Saskatoon, Saskatchewan,
Canada.
November 2011 - June 2012 Co-convener, Climate change and the carbon cycle, Congress of the Canadian
Meteorological and Oceanographic Society, May 29-June 1 2012, Montreal,
QC.
Meteorological and Oceanographic Society, June 5-9 2011, Victoria, BC
Meteorological and Oceanographic Society, May 31-June 4 2010, Ottawa,
Canada.

489
June 2008 - December 2008 Co-convener, Climate Stabilization: Prospects, Trajectories, Uncertainties,
American Geophysical Union (AGU) Fall Meeting, December 15-19 2008, San
Francisco, CA.
Journal Editor
May 2016 Current Section Editor, Carbon and Climate, Current Climate Change Reports
September 2014 Current Guest Co-Editor, Focus Issue Cumulative Emissions, Carbon Budgets and
Climate mitigation, Environmental Research Letters.
Journal Reviewer
2016 Nature Climate Change (1)
2015 Nature Climate Change (3), Environmental Research Letters (6), Review of
Geophysics (1), Climatic Change (1), Current Climate Change Reports (1)
2014 Nature (1), Nature Climate Change (1), Geophysical Research Letters (1),
Climatic Change (1)
2013 Nature Climate Change (2), Proceedings of the National Academy of Science
USA (1), Environmental Research Letters (1), Journal of Climate (1), Earth
System Dynamics (2), Journal of Geophysical Research - Atmosphere (4),
Climate of the Past (1)
2012 Nature Climate Change (4), Environmental Research Letters (1), Journal of
Climate (2), Earth System Dynamics (2)
2011 Nature Climate Change (2), Nature Education (1), Proceedings of the National
Academy of Science (1), Plos One (1)
2010 Proceedings of the National Academy of Science (1), Geophysical Research
Letters (1), Atmosphere-Ocean (1), Climate Dynamics (1), Journal of Climate
(1)
2009 Climatic Change (2), Energy Policy (1), Proceedings of the National Academy
of Science (1), Tellus (1)
2008 Journal of Climate (2)
2007 Climatic Change (2), Journal of Climate (1)
2006 Journal of Climate (4), Climate Dynamics (2), Climatic Change (1),
Meteorology and Atmospheric Science (1)
Grant Proposal Reviewer

2015 Natural Sciences and Engineering Research Council of Canada (1 proposal)
2013 National Environmental Research Council (UK) (1 proposal)
2012 Natural Sciences and Engineering Research Council of Canada (1 proposal),
Swiss National Foundation (1 proposal), Leverhulme Trust (1 proposal)
2011 Natural Sciences and Engineering Research Council of Canada (1 proposal)
2010 National Science Foundation (USA) (1 proposal), National Environmental
Research Council (UK) (1 proposal)
2009 National Oceanographic and Atmospheric Administration (USA) (1 proposal)
2008 National Oceanographic and Atmospheric Administration (USA) (1 proposal)
2007 National Science Foundation (USA) (1 proposal)

490
____________________________________________________________
Membership in the Academic Community

Canadian Meteorological and Oceanographic Society (2011-Current)
American Geophysical Union (2001-Current)
____________________________________________________________
Service to the Community at Large

Media interviews
Sept. 1, 2015 CBC Vancouver, 5:34 pm PDT, host: Andrew Chang, topic: wind storm preparedness
Nov. 13, 2014 CKNW, Mike Eckford Show, 4-5 pm, 1-hour long panel on climate change with Kirsten Zickfeld
(SFU), Simon Donner (UBC), Kathryn Harrison (UBC)
June 2, 2014 CTV News Channel, 3:30 pm PDT, host: van der Heyden, topic: Obamas new policy on coal-fired
power plants
March 26 2014 Breakfast TV, topic: SFU faculty letter asking Board of governors to divest SFU endowment from
fossil fuels
May 9, 2013 CTV News Channel, topic: Scientist Letter asking Hon. Minister Joe Oliver to consider climate
change in energy infrastructure decisions
May 16, 2013 CTV News National, topic: Harper visits Washington to lobby for Keystone pipeline
Dec. 12, 2012 CTV, topic: Coastal protection plan for Vancouver
March 4, 2012 Publication of Nature Climate Change paper Climate response to zeroed emissions of greenhouse
gases and aerosols. Interview with CTV, CFAX, CKMW, 1130 News, Vancouver 24 hours, Canada
Chemical News
Movies
2015 Featuring in Running on Climate, a documentary about the election campaign of Green MLA
Andrew Weaver. Director: Robert Alstead, icycle.ca productions ltd

491
This is Exhibit "Y.." ref'elred to in the

ll affidavit of Kirsten Zickfeld,
before me on the
of April, 2017.
A commissioner for tal<

Blitish Columbia
492
VOLUME 22 JOURNAL OF CLIMATE 15 MAY 2009
Lifetime of Anthropogenic Climate Change: Millennial Time Scales of Potential

CO2 and Surface Temperature Perturbations
M. EBY, K. ZICKFELD, AND A. MONTENEGRO
School of Earth and Ocean Sciences, University of Victoria, Victoria, British Columbia, Canada
D. ARCHER
Department of the Geophysical Sciences, University of Chicago, Chicago, Illinois
K. J. MEISSNER AND A. J. WEAVER

School of Earth and Ocean Sciences, University of Victoria, Victoria, British Columbia, Canada
(Manuscript received 2 April 2008, in final form 15 September 2008)
ABSTRACT
Multimillennial simulations with a fully coupled climatecarbon cycle model are examined to assess the
persistence of the climatic impacts of anthropogenic CO2 emissions. It is found that the time required to absorb
anthropogenic CO2 strongly depends on the total amount of emissions; for emissions similar to known fossil
fuel reserves, the time to absorb 50% of the CO2 is more than 2000 yr. The long-term climate response appears
to be independent of the rate at which CO2 is emitted over the next few centuries. Results further suggest that
the lifetime of the surface air temperature anomaly might be as much as 60% longer than the lifetime
of anthropogenic CO2 and that two-thirds of the maximum temperature anomaly will persist for longer than
10 000 yr. This suggests that the consequences of anthropogenic CO2 emissions will persist for many millennia.
1. Introduction Atmospheric CO2 is currently the dominant anthro-

pogenic greenhouse gas implicated in global warming
The projection of the climatic consequences of anthro-
(Forster et al. 2007); therefore, estimating the lifetime
pogenic CO2 emissions for the twenty-first century has
of anthropogenic climate change will largely depend on
been a major topic of climate research. Nevertheless, the
the perturbation lifetime of CO2. The perturbation
long-term consequences of anthropogenic CO2 remain
lifetime is a measure of the time over which anomalous
highly uncertain. The Intergovernmental Panel on Cli-
levels of CO2 or temperature remain in the atmosphere
mate Change (IPCC) Fourth Assessment Report (AR4)
(defined here to be the time required for a fractional
reported that about 50% of a CO2 increase will be re-
reduction to 1/e). Carbon emissions can be taken up
moved from the atmosphere within 30 years and a further
rapidly by the land, through changes in soil and vege-
30% will be removed within a few centuries (Denman
tation carbon, and by dissolution in the surface ocean.
et al. 2007, p. 501). Although the IPCC estimate of the
Ocean uptake slows as the surface waters equilibrate
time to absorb 50% of CO2 is accurate for relatively small
with the atmosphere and continued uptake depends on
amounts of emissions at the present time, this may be a
the rate of carbon transport to the deep ocean. Ocean
considerable underestimation for large quantities of
uptake is enhanced through dissolution of existing
emissions. Carbon sinks may become saturated in the fu-
CaCO3, often referred to as carbonate compensation.
ture, reducing the systems ability to absorb CO2.
As CO2 is taken up, the ocean becomes more acidic,
eventually releasing CaCO3 from deep sediments. This
increases the ocean alkalinity, allowing the ocean to
Corresponding author address: M. Eby, School of Earth and
Ocean Sciences, University of Victoria, P.O. Box 3055, Victoria, take up additional CO2. Carbonate compensation be-
BC V8W 3P6, Canada. comes important on millennial time scales, whereas
E-mail: eby@uvic.ca changes in the weathering of continental carbonate and
DOI: 10.1175/2008JCLI2554.1
2009 American Meteorological Society 2501

493
2502 JOURNAL OF CLIMATE VOLUME 22
TABLE 1. Simulated global carbon inventories in 1800 and 1994

and their differences (in PgC). The estimated values are taken
from the IPCC Fourth Assessment Report (Denman et al. 2007,
Fig 7.3).
Year Atmosphere Vegetation 1 Soil Ocean Sediment

1800 591 604 1 1366 5 1970 37 237 1166
1994 761 661 1 1437 5 2098 37 340 1166
Difference 170 57 1 71 5 128 103 0
Estimate 165 101 118 0
of the perturbation lasts for more than 5000 yr (Archer

et al. 1998; Archer 2005; Archer and Brovkin 2008;
Lenton and Britton 2006; Lenton et al. 2006; Ridgwell
and Hargreaves 2007; Ridgwell et al. 2007; Mikolajewicz
et al. 2007; Tyrell et al. 2007; Montenegro et al. 2007).
None of these studies attempted to estimate the millen-
FIG. 1. Historical changes in CO2 and SAT. (top) Model simu-
nial time scales of the temperature response or investi-
lated CO2 and (bottom) SAT are compared to historical data gated the multimillennial response as a function of the
(Ethridge et al. 1998; Keeling and Whorf 2005; Jones et al. 2008). magnitude of the perturbation in a systematic way.
The model simulation includes all historical forcings (CO2 emis- Models that have looked at the long-term carbon
sions, insolation, orbital forcing, tropospheric and stratospheric cycle response are usually low resolution, highly pa-
sulfates and non-CO2 greenhouse gases such as CH4, N2O, and
CFCs).
rameterized, or incomplete. For example, Archer
(2005) used highly parameterized climate feedbacks,
whereas Montenegro et al. (2007) used two incomplete
silicate are thought to become important on the 10 000
models: one model lacked a terrestrial carbon cycle and
100 000-yr time scale (Archer 2005; Sarmiento and
the other lacked ocean sediments. The model used here
Gruber 2006; Lenton and Britton 2006).
is currently one of the more complex coupled climate
Earth system models can be used to simulate the ev-
carbon cycle models capable of looking at multimil-
olution of the climate system under different anthro-
lennial time scales. Even given the large range in ex-
pogenic emissions scenarios. There is still a great deal of
isting model predictions, we will show that the lifetime
uncertainty in the climatecarbon cycle response and
of both the anthropogenic CO2 perturbation and the
considerable variation in model predictions. The short
resulting surface air temperature (SAT) change may be
term (century time scale) may be dominated by the
longer than previously thought.
terrestrial carbon cycle response, which is poorly un-
derstood. Over the longer term (millennial time scale)
the ocean biology, sediment, and weathering responses
2. Model description and evaluation
are also highly uncertain. Comprehensive model simu-
lations of the next few centuries suggest that CO2 We use version 2.8 of the University of Victoria (UVic)
anomalies may be relatively long lived (Friedlingstein Earth System Climate Model (ESCM). It consists of a
et al. 2006; Plattner et al. 2008). These studies also primitive equation 3D ocean general circulation model
illustrate the large uncertainties in the modeled short- with isopycnal mixing and a Gent and McWilliams (1990)
term carbon cycle response but they were not designed parameterization of the effect of eddy-induced tracer
to estimate the multimillennial response or the de- transport. For diapycnal mixing, a horizontally constant
pendency of the recovery time scales on the level of profile of diffusivity is applied, with values of about
emissions. 0.3 1024 m2 s21 in the pycnocline. The ocean model is
There are few modeling studies that have consid- coupled to a dynamicthermodynamic sea ice model and
ered the coupled climatecarbon cycle response to an energymoisture balance model of the atmosphere
large anthropogenic emissions on the 10 000-yr time with dynamical feedbacks (Weaver et al. 2001). The land
scale. Differing levels of complexity and experimental surface and terrestrial vegetation components are rep-
design make a detailed comparison of other studies resented by a simplified version of the Hadley Centre
difficult, but most studies suggest that the average Met Office surface exchange scheme (MOSES) coupled
perturbation lifetime of most of the CO2 is on the to the Top-down Representation of Interactive Foliage
order of a few centuries and that as much as a quarter and Flora Including Dynamic vegetation model; Meissner
494
15 MAY 2009 EBY ET AL. 2503
TABLE 2. Modeled and estimated global carbon budgets are for the 1980s, 1990s, and 200005 in PgC yr21. The estimated values are taken
from the IPCC Fourth Assessment Report (Denman et al. 2007, Table 7.1).
1980s 1990s 200005

Model Estimate Model Estimate Model Estimate
Atmospheric increase 3.3 3.3 6 0.1 3.7 3.2 60.1 4.2 4.1 6 0.1
Ocean uptake 21.8 21.8 6 0.8 22.2 22.2 6 0.4 22.4 22.2 6 0.5
Land uptake 22.2 21.7 (23.4 to 0.2) 22.6 22.6 (24.3 to 20.9) 22.8 n/a
et al. 2003). Land carbon fluxes are calculated within 1994) compare relatively well with IPCC AR4 estimates
MOSES and are allocated to vegetation and soil carbon (17501994). The observation-based changes in carbon
pools (Matthews et al. 2004). Ocean carbon is simulated reservoirs during the 1980s, 1990s, and 200005 are well
by means of an Ocean Carbon-Cycle Model Intercom- reproduced by the model (Table 2). The atmospheric
parison Project type inorganic carbon cycle model and a CO2 increase is in close agreement with observations for
nutrientphytoplanktonzooplanktondetritus marine the 1980s and 200005 but is overestimated in the 1990s.
ecosystem model (Schmittner et al. 2008). Sediment Ocean CO2 uptake agrees very well with the observation-
processes are represented using an oxic-only model of based values, but for a slight overestimation in 200005.
sediment respiration (Archer 1996a). Land CO2 uptake falls well within the estimated un-
An earlier version of the UVic ESCM (version 2.7) certainty range for all time periods and is close to the
has undergone extensive evaluation as part of inter- IPCC best estimate.
national model intercomparison projects including the The model reproduces qualitatively and quantita-
Coupled Carbon Cycle Climate Model Intercomparison tively most features of the observation-based patterns of
Project (Friedlingstein et al. 2006), the Paleoclimate
Modeling Intercomparison Project (Weber et al. 2007),
and the coordinated thermohaline circulation experi-
ments (Gregory et al. 2005; Stouffer et al. 2006). The
model has also been used for multicentury climate
projections in support of the IPCC Fourth Assessment
Report (Denman et al. 2007; Meehl et al. 2007). Here,
we evaluate the UVic ESCM version 2.8 primarily with
respect to its ability to simulate characteristics of the
coupled climatecarbon cycle system, including the air
sea flux of CO2, the distribution of ocean dissolved in-
organic carbon (DIC) and alkalinity, the percent of
CaCO3 in sediments, the global carbon budgets of the
last decades and the observation-based evolution of sur-
face air temperature and CO2 over the historical period.
From a preindustrial climate, this version of the model
has a transient climate response of 2.08C and an equilib-
rium climate sensitivity of 3.58C (Weaver et al. 2007).
The simulated evolution of atmospheric CO2 and
surface air temperature over the historical period is in
good agreement with observations (Fig. 1). For the year
2000, the simulated CO2 is about 5 ppmv higher than the
observation-based value. The model does not produce
as much interannual variability as seen in the data but
the long-term trends are well reproduced. Warming
over the twentieth century is 0.78C, in agreement with
the IPCC estimate of 0.68 6 0.28C (Forster et al. 2007).
The simulated inventories of carbon in the atmos-
phere, ocean, and on land in the years 1800 and 1994 FIG. 2. Airsea flux of carbon. (top) Model simulated fluxes at
and their difference are given Table 1. The changes in the year 2000 compared with (bottom) observational estimates
carbon inventories over the historical period (1800 (Takahashi et al. 2009). Negative values denote ocean uptake.
495
FIG. 3. (top) Model simulated zonally averaged DIC at the year 1994 compared with (bottom) GLODAP data (Key
et al. 2004) for (left) ArcticAtlantic and (right) Indo-Pacific oceans.
airsea exchange of CO2 (Fig. 2). These features include 3. Experimental design
outgassing in low latitudes with a maximum in the
eastern tropical Pacific and uptake at mid- and high The model was spun up for 10 000 yr with atmospheric
latitudes with maxima around 408NS in the areas of the carbon dioxide levels and Earths orbital configuration
North Atlantic Current, the Kuroshio Current, and the specified for the year 1800 and the continental CaCO3
Southern Ocean. Model biases include underestimated weathering flux diagnosed from the ocean sediment
uptake in the GreenlandIcelandNorwegian Seas burial flux. The weathering flux was then held fixed
and overestimated uptake in the eastern subtropical while the burial flux of CaCO3 was allowed to evolve
Pacific. with time for all subsequent experiments. Historical
The simulated patterns of DIC and alkalinity show emissions were applied until the end of the year 2000.
good agreement with observations (Figs. 3, 4). The These historical CO2 emissions include contributions
model captures well the surface to deep gradient of both from both fossil fuel burning and land use changes. All
tracers. At depth the model slightly underestimates other transient forcings (insolation, orbital forcing,
carbon while slightly overestimating alkalinity. See tropospheric and stratospheric sulfates, and non-CO2
Table 3 for a summary of the average values and ab- greenhouse gases such as CH4, N2O, and CFCs) were
solute errors of simulated DIC and alkalinity for the held fixed.
global, ArcticAtlantic, and Indo-Pacific oceans. The At the beginning of 2001, pulses of CO2 were ap-
simulated patterns of CaCO3 are also in reasonable plied over 1 yr. The emissions varied from 160 PgC
agreement with observations (Fig. 5). Nevertheless, the (1015 g of carbon) to 5120 PgC (Table 4). The upper
model underestimates deep CaCO3 at tropical latitudes bound approximates all known conventional fossil fuel
and overestimates CaCO3 at high latitudes. Comparing reserves (Rogner 1997). In addition to the pulse ex-
only locations with observations, the global average periments, we also performed simulations with more
percent of CaCO3 in sediments is 34.5% for the data realistic emissions scenarios. As a baseline, we as-
and 31.1% for the model. sumed that emissions follow the A2 scenario up to the
496
15 MAY 2009 EBY ET AL. 2505
FIG. 4. (top) Model simulated zonally averaged alkalinity at the year 1994 compared with (bottom) GLODAP data
(Key et al. 2004) for (left) ArcticAtlantic and (right) Indo-Pacific oceans.
year 2100 and then decline linearly to zero by 2300. This 4. Discussion and conclusions
scenario is designated as A21 (Montenegro et al. 2007).
We then generated a set of scenarios in which the A21 Resulting maximum changes in atmospheric CO2
emissions were scaled such that the cumulative emis- range from 26 to 2352 ppmv (Fig. 6; Table 4). In the
sions reached those of the equivalent pulse simulation pulse experiments, the maximum CO2 anomaly occurs
by the year 2300. A21 and pulse simulations were in- at the beginning, initially decaying very rapidly but
tegrated for 5000 and 10 000 model years, respectively. slowing after several decades. In the A21 experiments,
To explore the consequences of future emissions only, a atmospheric CO2 peaks a few decades before the year
10 000-yr control simulation was also carried out with emissions are set to zero (260286 yr; Table 4). After the
zero emissions after the year 2000. At the end of this peak, CO2 closely approaches the level of the corre-
integration the SAT was again at its year 2000 value sponding pulse experiment after about 500 yr. This
(having dropped 0.18C from its temporary maximum) demonstrates that the long-term atmospheric CO2 re-
whereas CO2 had dropped by 55 ppmv to 321 ppmv. sponse is nearly independent of the rate of CO2
These control results are subtracted from the results of emissions (assuming all emissions occur over the next
the future emissions experiments. 300 yr).
TABLE 3. Model (M), data estimate (D; Key et al. 2004), and absolute error (E) for DIC and Alkalinity averaged over the Global,
ArcticAtlantic, and Indo-Pacific oceans for the year 1994.
Global ArcticAtlantic Indo-Pacific

M D E M D E M D E
23
DIC (mol m ) 2.291 2.309 0.022 2.233 2.246 0.019 2.311 2.331 0.023
Alkalinity (mol m23) 2.424 2.421 0.014 2.396 2.392 0.012 2.434 2.431 0.014
497
A considerable amount (15%30%) of the atmo-

spheric CO2 anomaly persists at the end of the 10 000-yr
simulations (Fig. 6). The time to absorb a given percent
of emissions is strongly dependent on the total amount
of emissions (Fig. 7; Table 4). For emissions up to about
1000 PgC, 50% of the CO2 anomaly is taken up within
100 yr and another 30% is absorbed within 1000 yr,
which is similar to IPCC estimates (Denman et al. 2007).
Above 1000 PgC, the time to absorb 50% of the emissions
increases dramatically, and more than 2000 yr are needed
to absorb half of a 5000-PgC perturbation.
Ocean surface pH is strongly coupled to atmospheric
CO2 (Caldeira and Wicket 2003). Emissions above 1280
PgC result in a decrease in average ocean surface pH
that is larger than the 0.2 guard rail proposed by the
German Advisory Council on Global Change (WGBU;
Schubert et al. 2006; Fig. 8). Given the slow decay of
atmospheric CO2, experiments with emissions of 2560
PgC and larger still have lower pH than the 0.2 guard
rail after 10 000 yr. For high emissions, the change in
surface pH would probably have a significant impact on
oceanic biota. Emissions of 1920 PgC and above result
in minimum pH levels below 7.9, a value that could
bring the aragonite saturation depth to the surface in
the Southern Ocean generating serious adverse effects
on calcifying organisms (Orr et al. 2005).
There is a lag in the response of surface air temper-
FIG. 5. (top) Model simulated percent dry weight CaCO3 at year
ature to the CO2 forcing (Fig. 9). For all but the lowest
2000 compared with (bottom) coretop data (Archer 1996b). Note
that only locations with data are shown in both panels to facilitate emissions, temperature reaches its maximum at least
the comparison. 550 yr after the peak in atmospheric CO2 (Table 4). The
lag is particularly pronounced in the experiments with
TABLE 4. Level and year of maximum CO2 (Max CO2), first year at which 50% of total emissions have been absorbed from the
atmosphere (50% emissions), level and year of maximum SAT (Max SAT), and the first year at which SAT is less than 80% of the
maximum (80% max SAT).
Max CO2 Max SAT

Expt (Pg) (ppmv) (yr) 50% emissions (yr) (8C) (yr) 80% max SAT (yr)
160 69 1 18 0.32 247 527
160_A21 26 270 187 0.32 342 1787
320 139 1 23 0.61 110 2230
640 280 1 36 1.38 3519 4363
640_A21 118 260 201 1.40 3357 4153
960 423 1 63 2.00 2047 3126
1280 568 1 105 2.55 1965 3521
1280_A21 274 269 232 2.53 2110 3583
1920 859 1 218 3.70 1147 3441
2560 1155 1 428 4.75 715 3929
2560_A21 699 278 520 4.72 832 3943
3200 1453 1 781 5.66 809 4441
3840 1752 1 1309 6.48 1076 5066
3840_A21 1223 284 1388 6.43 1147 4986
4480 2051 1 1732 7.24 1085 5248
5120 2352 1 2151 7.86 971 6190
5120_A21 1781 286 2210 7.82 1287 .5000
498
15 MAY 2009 EBY ET AL. 2507
FIG. 6. Temporal changes in CO2. Differences (top) relative to FIG. 8. Temporal changes in sea surface pH. (top) Differences
the control and (bottom) in terms of the percentage of CO2 relative to the control simulation and (bottom) differences in
emissions remaining in the atmosphere. Note the different scales terms of the percentage of the maximum pH anomaly remaining.
along the time axis. Colors indicate total emissions, with solid lines Note the different scales along the time axis. Colors indicate total
for pulse scenarios and dotted lines for equivalent A21 scenarios. emissions, with solid lines for pulse scenarios and dotted lines for
equivalent A21 scenarios. Results for equivalent A21 scenarios
are not shown in the bottom panel for clarity.
FIG. 7. Percentages of anomalies remaining: (top) CO2 and

(bottom) SAT. Stars indicate experimental points and lines are just FIG. 9. Temporal changes in SAT. Differences (top) relative to
visual aids. Note the different scales along the time axis and that the control and as a percentage of the maximum SAT anomaly for
colors indicate different percentages remaining, not total emissions, (middle) high and (bottom) low emissions. High and low emissions
as in Figs. 6, 8, 9. For clarity, results for equivalent A21 scenarios are plotted separately for clarity. Note the different scales along
are not shown. The SAT anomaly is noisy for low emissions due the time axis. Colors indicate total emissions, with solid lines for
to long time-scale climate variability (see Fig. 9 and text). pulse scenarios and dotted lines for equivalent A21 scenarios.
499
total emissions in the range 6401280 PgC, where after

20003500 yr, the planetary cooling is suddenly reversed
and SAT again increases by as much as 0.58C. This
abrupt warming and accompanying increase in CO2 is
caused by flushing events in the Southern Ocean, which
in this model have been shown to be dependent on the
level of atmospheric CO2 (Meissner et al. 2008). Under
the A21 emissions scenarios, the peak in SAT is almost
identical to the corresponding pulse experiments, indi-
cating that the long-term temperature response is in-
dependent of the rate of CO2 emissions (Fig. 9; Table 4).
The SAT anomaly is even longer lived than the CO2
anomaly. For all experiments, at least 50% of the
maximum temperature anomaly persists at the end of FIG. 10. Curve fitting to a double exponential model. Dotted
lines are an exponential fit to simulated CO2 after 1000 yr and are
the simulation. For both the smallest and largest emis-
used to estimate the slow time scale for reducing CO2 (slow).
sion scenarios, the temperature anomaly remaining af- These curves were extrapolated back 1000 yr and the extrapolated
ter 10 000 yr is about 75% of the maximum anomaly. CO2 was subtracted from the simulated CO2. A second exponen-
Similar to CO2, the time to reduce temperature by a tial fit was performed on the remaining CO2 to estimate a fast time
specific percent of the maximum anomaly depends scale for reducing CO2 (fast). The dashed curves are the sum of
two exponential curves (fast 1 slow). Note the different scales
on the total amount of emissions. The time within which
along the time axis.
SAT declines by 20% relative to the peak warming
ranges from about 500 yr for the lowest emission sce-
nario to more than 5000 yr for the highest emissions average lifetime, and A2 is the amount of any very long-
scenarios (Fig. 7; Table 4). lived residual. We restrict our analysis to experiments with
Given that the change in temperature from prein- total emissions greater than 1500 PgC. In simulations with
dustrial to the year 2000 is about 0.88C (Fig. 1), total lower emissions, the response curve is often contaminated
emissions of 640 PgC or more result in average air by noise, making curve fitting imprecise (Figs. 6, 9).
temperatures above the 28C temperature guard rail A gradient-expansion algorithm was used to compute
suggested by the WBGU (Schubert et al. 2006) and the least squares fit of an exponential model to the data.
endorsed by the European Union. The threshold to stay To tease out a fast and slow time scale for uptake of CO2,
below this guard rail would appear to be near 640 PgC an exponential fit was first applied to the CO2 curves
of total emissions from the year 2000. Experiments with after 1000 yr. The data fit an exponential very well (see
emissions of 1280 PgC and larger still exceed the 28C Fig. 10). This curve was then extrapolated back 1000 yr
guard rail after 10 000 yr. and the extrapolated CO2 was subtracted from the sim-
To estimate the perturbation lifetime of anthropo- ulated CO2. A second exponential fit was performed on
genic climate change the response curves of either CO2 the remaining CO2. This fit is clearly not as good as the
or temperature were fit to an exponential formula of the previous fit (Fig. 10). The early response is not a pure
form A0exp(2t/A1) 1 A2. The parameter A0 gives an exponential but a combination of processes with differ-
estimate of the amount a quantity is reduced, A1 is the ent time scales (Joos et al. 1996). Still, this analysis
TABLE 5. Average perturbation lifetimes in years and percentages reduced. The average perturbation lifetimes are calculated from
exponential fits to model results. Percentages are of either total CO2 emissions or maximum SAT. All are calculated from differences
with the control (control has zero emissions from year 2001 onward).
CO2 SAT
Fast Slow Slow
Expt (Pg) (years) (%) (yr) (%) 10 000 yr (%) (yr) (%) 10 000 yr (%)
1920 146 56 3000 24 20 3400 39 61
2560 149 50 3000 28 22 3900 37 63
3200 136 43 2700 33 24 3900 34 66
3840 129 36 2900 38 26 4300 34 66
4480 102 29 2600 43 28 4100 31 69
5120 107 27 2900 44 29 4600 31 69
500
15 MAY 2009 EBY ET AL. 2509
FIG. 11. Temporal changes in carbon pools. Differences in car-

bon relative to the control simulation for the 2560-PgC pulse ex- FIG. 12. Portion of anomalous CO2, radiative forcing, and sur-
periment. The sediment pool includes changes due to continental face temperature relative to 1500 yr after the start of the simula-
weathering. Note the different scales along the time axis. tion. Although sometimes indistinct in the figure, the radiative
forcing and temperature curves are similar: both show longer time
scales (decline less steeply) than CO2 and time scales become
longer as emissions increase.
provides a reasonable, if somewhat uncertain, estimate
of the overall fast absorption time scale. Although the
estimated short-term-response time scale may be de- fits the temperature response very well. The average
pendent on the number of exponentials used in the fit perturbation lifetime is about 4000 yr, or 40% longer
(Maier-Reimer and Hasselmann 1987), the longer re- than the average for CO2. The temperature perturba-
sponse time scale (after 1000 yr) is quite robust and tion lifetime also appears to be more dependent on the
reasonably independent of the section of the curve used level of total emissions than the CO2 perturbation life-
in the fit. The perturbation lifetime of CO2 is thus broken time (Table 4).
up into a period of rapid absorption, a period of slow Radiative forcing from atmospheric CO2 depends on
absorption, and a residual that represents CO2, which the logarithm of CO2, but for the first 1000 yr, the
stays in the atmosphere for longer than this method can thermal inertia of the ocean and climate feedbacks are
resolve (10 000 yr). To derive a perturbation lifetime important in keeping SAT below what would be ex-
for temperature, we also fit an exponential model to pected from the radiative forcing alone (Meehl et al.
the temperature response curves after the year 1000. 2007). After 1000 yr, the time scale for reducing SAT
We find that the response curves for CO2 can be well becomes very similar to the time scale of the CO2 ra-
approximated by the superposition of exponentials with diative forcing and this time scale is considerably longer
two different time scales. The average lifetime for the than for CO2. The logarithmic dependence of the radi-
short time scale is about 130 yr whereas the long time ative forcing on CO2 is also why the SAT perturbation
scale has an average lifetime closer to 2900 yr (Table 5). lifetime depends on the total amount of emissions, even
The amount of CO2 absorbed by processes associated though the time scale of CO2 absorption itself appears
with the short time-scale sink are nearly constant (1075 to be relatively constant.
1382 PgC; calculated from Table 5). About 400 PgC of Figure 12 shows the portion of CO2, radiative forcing
the short time-scale sink is associated with increased from CO2, and surface temperature normalized to their
land uptake (mostly through CO2 fertilization), whereas values at 1500 yr. The spread in the time scales for
the rest (;900 PgC) are due to relatively rapid disso- CO2 (illustrated by the spread in the curves) is relatively
lution in the surface ocean (Fig. 11). The longer time small and larger emissions seem to show slightly shorter
scale of the deep-ocean sink is associated with slow rates time scales (steeper slopes) than smaller emissions (also
of deep-ocean transport and carbonate dissolution. The see Table 5). Radiative-forcing time scales are longer
amount taken up by the deep-ocean sink is not constant than for CO2 alone and, as with temperature, the time
but increases at higher levels of emissions, implying that scale for the decay of the radiative forcing increases as
the sink is not saturated. The absorption time scale for emissions increase. The temperature time-scale depen-
CO2 does not seem to be very sensitive to the amount of dency on emissions can mostly be explained by the
emissions (Table 5). changes in radiative-forcing time scales, although other
For high-emission experiments, after year 1000 (roughly feedbacks make the spread in temperature time scales
the year of maximum temperature), a single exponential even larger.
501
In summary, this study suggests that for emissions less Etheridge, D. M., L. P. Steele, R. L. Langenfelds, R. J. Francey,
than about 1500 PgC, most of the CO2 will be absorbed J.-M. Barnola, and V. I. Morgan, 1998: Historical CO2 rec-
ords from the Law Dome DE08, DE08-2, and DSS ice cores.
within a few centuries, which is in agreement with ear-
Trends: A Compendium of Data on Global Change, Carbon
lier work. Temperature anomalies may last much lon- Dioxide Information Analysis Center. [Available online at
ger. With larger emissions, the time to absorb most of http://cdiac.esd.ornl.gov/trends/co2/lawdome.html.]
the CO2 increases rapidly (Table 4; Fig. 7). This de- Forster, P., and Coauthors, 2007: Changes in atmospheric con-
pendency of the CO2 response on the level of emissions stituents and in radiative forcing. Climate Change 2007: The
Physical Science Basis, S. Solomon et al., Eds., Cambridge
has important policy implications and needs to be in-
University Press, 129234.
vestigated with other models. A long-term model in- Friedlingstein, P., and Coauthors, 2006: Climatecarbon cycle
tercomparison project (LTMIP) with standardized ex- feedback analysis: Results from the C4MIP model intercom-
periments has recently been initiated and this will parison. J. Climate, 19, 33373353.
hopefully further increase our understanding and re- Gent, P. R., and J. C. McWilliams, 1990: Isopycnal mixing in ocean
duce the uncertainty in the long-term carbon cycle re- circulation models. J. Phys. Oceanogr., 20, 150155.
Gregory, J. M., and Coauthors, 2005: A model intercomparison of
sponse. Preliminary results from nine models (including changes in the Atlantic thermohaline circulation in response
the one used here) can be found in Archer et al. (2009). to increasing atmospheric CO2 concentration. Geophys. Res.
Although the long-term climatecarbon cycle re- Lett., 32, L12703, doi:10.1029/2005GL023209.
sponse still remains highly uncertain, the model used in Jones, P. D., D. E. Parker, T. J. Osborn, and K. R. Briffa, 2008:
this study suggests that for large emissions, the pertur- Global and hemispheric temperature anomaliesLand and
marine instrumental records. Trends: A Compendium of Data
bation lifetime of both CO2 and surface temperature
on Global Change, Carbon Dioxide Information Analysis
might be longer than previously thought. The long-term Center. [Available online at http://cdiac.ornl.gov/trends/temp/
climate response appears to be independent of the rate jonescru/jones.html.]
at which CO2 is emitted over the next few centuries. Joos, F., M. Bruno, R. Fink, U. Siegenthaler, T. F. Stocker, and
Regardless of the future emissions trajectory, changes C. LeQuere, 1996: An efficient and accurate representation
of complex oceanic and biospheric models of anthropogenic
to the earths climate will likely persist for several
carbon uptake. Tellus, 48B, 397417.
thousands of years. The logarithmic relationship be- Keeling, C. D., and T. P. Whorf, 2005: Atmospheric CO2 records
tween CO2 and its radiative forcing implies that the time from sites in the SIO air sampling network. Trends: A Com-
scale at which atmospheric temperature declines will be pendium of Data on Global Change, Carbon Dioxide Informa-
longer than the time scale of CO2. For ecosystems tion Analysis Center. [Available online at http://cdiac.ornl.gov/
trends/co2/sio-keel.html.]
having already adapted to a warmer world, slow cooling
Key, R. M., and Coauthors, 2004: A global ocean carbon climatol-
may be beneficial. Nevertheless, it is sobering to ponder ogy: Results from Global Data Analysis Project (GLODAP).
the notion that the carbon we emit over a handful of Global Biogeochem. Cycles, 18, GB4031, doi:10.1029/
human lifetimes may significantly affect the earths cli- 2004GB002247.
mate over tens of thousands of years. Lenton, T. M., and C. Britton, 2006: Enhanced carbonate and
silicate weathering accelerates recovery from fossil fuel CO2
REFERENCES perturbations. Global Biogeochem. Cycles, 20, GB3009,
doi:10.1029/2005GB002678.
Archer, D., 1996a: A data-driven model of the global calcite ly- , and Coauthors, 2006: Millennial timescale carbon cycle and
socline. Global Biogeochem. Cycles, 10, 511526. climate change in an efficient Earth system model. Climate
, 1996b: An atlas of the distribution of calcium carbonate in Dyn., 26, 687711.
sediments of the deep sea. Global Biogeochem. Cycles, 10, Maier-Reimer, E., and K. Hasselmann, 1987: Transport and stor-
159174. age of CO2 in the oceanAn inorganic ocean-circulation
, 2005: Fate of fossil fuel CO2 in geologic time. J. Geophys. carbon cycle model. Climate Dyn., 2, 6390.
Res., 110, C09S05, doi:10.1029/2004JC002625. Matthews, H. D., A. J. Weaver, K. J. Meissner, N. P. Gillett, and
, and V. Brovkin, 2008: The millennial atmospheric life- M. Eby, 2004: Natural and anthropogenic climate change: In-
time of anthropogenic CO2. Climatic Change, 90, 283297, corporating historical land cover change, vegetation dynamics
doi:10.1007/s10584-008-9413-1. and the global carbon cycle. Climate Dyn., 22, 461479.
, H. Kheshgi, and E. Maier-Reimer, 1998: Dynamics of fossil Meehl, G. A., and Coauthors, 2007: Global climate projections.
fuel CO2 neutralization by marine CaCO3. Global Biogeochem. Climate Change 2007: The Physical Science Basis, S. Solomon
Cycles, 12, 259276. et al., Eds., Cambridge University Press, 747845.
, and Coauthors, 2009: Atmospheric lifetime of fossil-fuel Meissner, K. J., A. J. Weaver, H. D. Matthews, and P. M. Cox, 2003:
carbon dioxide. Annu. Rev. Earth Planet. Sci., 37, 117134. The role of land surface dynamics in glacial inception: A
Caldeira, K., and M. E. Wickett, 2003: Anthropogenic carbon and study with the UVic Earth System model. Climate Dyn., 21,
ocean pH. Nature, 425, 365. 515537.
Denman, K. L., and Coauthors, 2007: Couplings between changes , M. Eby, A. J. Weaver, and O. A. Saenko, 2008: CO2
in the climate system and biogeochemistry. Climate Change threshold for millennial-scale oscillations in the climate
2007: The Physical Science Basis, S. Solomon et al., Eds., system: Implications for global warming scenarios. Climate
Cambridge University Press, 589662. Dyn., 30, 161174.
502
15 MAY 2009 EBY ET AL. 2511
Mikolajewicz, U., M. Groger, E. Maier-Reimer, G. Schurgers, and biogeochemical cycling simulated for a business-as-usual
M. Vizcano, and A. M. E. Winguth, 2007: Long-term effects CO2 emission scenario until year 4000 AD. Global Bio-
of anthropogenic CO2 emissions simulated with a complex geochem. Cycles, 22, GB1013, doi:10.1029/2007GB002953.
earth system model. Climate Dyn., 28, 599633. Schubert, R., and Coauthors, 2006: The future oceansWarming
Montenegro, A., V. Brovkin, M. Eby, D. Archer, and A. J. Weaver, up, rising high, turning sour. Wissenschaftlicher Beirat der
2007: Long term fate of anthropogenic carbon. Geophys. Res. Bundesregierung Globale Umweltveranderungen Special
Lett., 34, L19707, doi:10.1029/2007GL030905. Rep., 110 pp.
Orr, J. C., and Coauthors, 2005: Anthropogenic ocean acidification Stouffer, R. J., and Coauthors, 2006: Investigating the causes of the
over the twenty-first century and its impacts on calcifying response of the thermohaline circulation to past and future
organisms. Nature, 437, 681686. climate changes. J. Climate, 19, 13651387.
Plattner, G.-K., and Coauthors, 2008: Long-term climate commit- Takahashi, T., and Coauthors, 2009: Climatological mean and
ments projected with climatecarbon cycle models. J. Climate,
decadal changes in surface ocean pCO2, and net sea-air CO2
21, 27212751.
flux over the global oceans. Deep-Sea Res. II, in press.
Ridgwell, A., and J. C. Hargreaves, 2007: Regulation of atmo-
Tyrrell, T., J. G. Shepherd, and S. Castle, 2007: The long-term
spheric CO2 by deep-sea sediments in an Earth system
legacy of fossil fuels. Tellus, 59B, 664672.
model. Global Biogeochem. Cycles, 21, GB2008, doi:10.1029/
Weaver, A. J., and Coauthors, 2001: The UVic Earth System Cli-
2006GB002764.
, I. Zondervan, J. C. Hargreaves, J. Bijma, and T. M. Lenton, mate Model: Model description, climatology, and applications
2007: Assessing the potential long-term increase of oceanic to past, present and future climates. Atmos.Ocean, 39,
fossil fuel CO2 uptake due to CO2-calcification feedback. 361428.
Biogeosciences, 4, 481492. , M. Eby, M. Kienast, and O. A. Saenko, 2007: Response of
Rogner, H. H., 1997: An assessment of world hydrocarbon re- the Atlantic meridional overturning circulation to increasing
sources. Annu. Rev. Energy Environ., 22, 217262. atmospheric CO2: Sensitivity to mean climate state. Geophys.
Sarmiento, J. L., and N. Gruber, 2006: Ocean Biogeochemical Res. Lett., 34, L05708, doi:10.1029/2006GL028756.
Dynamics. Princeton University Press, 526 pp. Weber, S. L., and Coauthors, 2007: The modern and glacial over-
Schmittner, A., A. Oschlies, H. D. Matthews, and E. D. Galbraith, turning circulation in the Atlantic Ocean in PMIP coupled
2008: Future changes in climate, ocean circulation, ecosystems model simulations. Climate Past, 3, 5164.
503
This is Exhibit ". Z" referled to in the

I't affidavit of Kirsten Zickfeld,
before me on the
of Aoril, 2017.
A commissioner for
British Columbia
Vol 459 | 11 June 2009 | doi:10.1038/nature08047 504
LETTERS
The proportionality of global warming to cumulative
carbon emissions
H. Damon Matthews1, Nathan P. Gillett2, Peter A. Stott3 & Kirsten Zickfeld2
The global temperature response to increasing atmospheric CO2 is (CCR). The CCR is illustrated schematically in Fig. 1, which shows
often quantified by metrics such as equilibrium climate sensitivity the progression from carbon emissions to climate change. The CCR
and transient climate response1. These approaches, however, do not incorporates the standard concept of climate sensitivity (the temper-
account for carbon cycle feedbacks and therefore do not fully ature response to increased atmospheric CO2), in addition to a
represent the net response of the Earth system to anthropogenic carbon sensitivity (the amount by which atmospheric CO2 concen-
CO2 emissions. Climatecarbon modelling experiments have trations increase in response to CO2 emissions, as mediated by
shown that: (1) the warming per unit CO2 emitted does not depend natural carbon sinks, and including also the effect of feedbacks
on the background CO2 concentration2; (2) the total allowable between climate change and carbon uptake).
emissions for climate stabilization do not depend on the timing The CCR thus represents the net climate response to CO2 emis-
of those emissions35; and (3) the temperature response to a pulse sions, and can be defined as DT/ET, where DT is the global mean
of CO2 is approximately constant on timescales of decades to temperature change over some period of time, and ET is the total
centuries3,68. Here we generalize these results and show that the cumulative carbon dioxide emitted over that period. We assign units
carbonclimate response (CCR), defined as the ratio of temper- of trillion tonnes of carbon to ET (1 Tt 5 1 teratonne, or 1018 grams,
ature change to cumulative carbon emissions, is approximately of carbon, which is equivalent to 3.7 trillion tonnes of CO2), so the
independent of both the atmospheric CO2 concentration and its CCR as defined here carries units of uC per Tt C emitted. CCR can be
rate of change on these timescales. From observational constraints, written as:
we estimate CCR to be in the range 1.02.1 6C per trillion tonnes of
carbon (Tt C) emitted (5th to 95th percentiles), consistent with CCR 5 DT/ET
twenty-first-century CCR values simulated by climatecarbon 5 (DT/DCA) 3 (DCA/ET)
models. Uncertainty in land-use CO2 emissions and aerosol
forcing, however, means that higher observationally constrained where DCA is the change in atmospheric carbon (in Tt C). Written in
values cannot be excluded. The CCR, when evaluated from climate this way, CCR represents the product of the temperature change per
carbon models under idealized conditions, represents a simple yet unit atmospheric carbon increase (DT/DCA) and the airborne frac-
robust metric for comparing models, which aggregates both tion of cumulative carbon emissions (DCA/DET). If defined under
climate feedbacks and carbon cycle feedbacks. CCR is also likely conditions of constant doubled pre-industrial atmospheric CO2, DT
to be a useful concept for climate change mitigation and policy; by is equal to the equilibrium climate sensitivity, and if defined under
combining the uncertainties associated with climate sensitivity, doubled CO2 conditions in a simulation in which CO2 increases at
carbon sinks and climatecarbon feedbacks into a single quantity, 1% per year, DT is equal to the transient climate response1.
the CCR allows CO2-induced global mean temperature change to Both the airborne fraction of cumulative emissions and the tem-
be inferred directly from cumulative carbon emissions. perature change per unit atmospheric carbon increase are dependent
We propose a new measure of the climate response to anthro- on the atmospheric CO2 concentration and its rate of increase;
pogenic carbon dioxide emissions: the carbonclimate response however, the CCR (as the product of the two) shows a remarkable
constancy with time. This can be seen in Fig. 2, which shows three
Climatecarbon feedbacks model simulations using the University of Victoria Earth System
Carbon sensitivity Climate sensitivity Climate Model9 (UVic ESCM, see Methods), an intermediate-
complexity coupled climatecarbon model. In all simulations, we
CO2 emission CO2 concentration Climate change prescribed atmospheric CO2 concentrations and used the models
interactive carbon sinks to diagnose the implied anthropogenic
Carbonclimate response (CCR)
CO2 emissions consistent with the prescribed concentration
Figure 1 | Schematic representation of the progression from CO2 emissions changes10. In the first simulation (Fig. 2a) we increased atmospheric
to climate change. We define carbon sensitivity as the increase in CO2 by 1% per year for 70 years; in the second and third simulations
atmospheric CO2 concentrations that results from CO2 emissions, as (Fig. 2b), atmospheric CO2 was doubled (solid lines) or quadrupled
determined by the strength of natural carbon sinks. Climate sensitivity is (dashed lines) instantaneously and held constant for 1,000 years. In
shown here as a general characterization of the temperature response to
all simulations, the airborne fraction of cumulative emissions
atmospheric CO2 changes. Feedbacks between climate change and the
strength of carbon sinks are shown as the upper dotted arrow decreased over time, whereas the temperature change per unit change
(climatecarbon feedbacks). The CCR aggregates the climate and carbon in atmospheric carbon increased with time. After an initial adjust-
sensitivities (including climatecarbon feedbacks) into a single metric ment period of about a decade, the CCR remained almost constant at
representing the net temperature change per unit carbon emitted. ,1.7 uC per Tt C emitted.
1
Department of Geography, Planning and Environment, Concordia University, 1455 de Maisonneuve Blvd W., Montreal, Quebec, H3G 1M8, Canada. 2Canadian Centre for Climate
Modelling and Analysis, Environment Canada, 3800 Finnerty Road, Victoria, British Columbia, V8P 5C2, Canada. 3Met Office Hadley Centre, FitzRoy Road, Exeter, Devon, EX1 3PB, UK.
829
2009 Macmillan Publishers Limited. All rights reserved
LETTERS 505 NATURE | Vol 459 | 11 June 2009
a CCR to be more closely constrained in simulations in which cumu-

1.0 4
lative emissions vary smoothly. Nonetheless, if used as a metric for
T/CA model intercomparison, we recommend that CCR be defined under
0.8 standard conditions, such as at the time of CO2 doubling in a tran-
3 sient simulation with 1% CO2 increase per year. Defined in this way,
CCR generalizes previously proposed metrics (such as the temper-
0.6 CA/ET ature response to a small pulse or constant sustained emission6see
2 Supplementary Information for additional discussion) into a single
Temperature change per unit carbon (C per Tt C)

0.4 robust and versatile quantity which can be easily estimated from
CCR
current standard model experiments, and yet represents the climate
response to a wide range of CO2 emissions scenarios.
Airborne fraction (Tt C per Tt C)
1
0.2 In a given model, CCR is approximately constant with respect to
time and emissions scenario; however, we would expect the value of
0.0 0
CCR to vary among models owing to differences in both climate and
0 10 20 30 40 50 60 70 carbon sensitivities. Its time and scenario independence mean that
b the CCR can be estimated from any model simulation with either
1.0 6 prescribed CO2 emissions, or prescribed CO2 concentrations and
prognostic model carbon sinks. Consequently, the simulations
0.8
5 performed as part of the Coupled Climate Carbon Cycle Model
T/CA Intercomparison Project (C4MIP11) provide a means of estimating
4 the range of CCR values among the current generation of coupled
0.6 climatecarbon models.
3 Figure 3 shows results from the 11 C4MIP models and the
0.4 CA/ET ensemble mean, with global temperature change plotted as a function
2 of cumulative carbon emissions (Fig. 3a) and temperature change per
CCR
unit carbon emitted plotted as a function of time (Fig. 3b). Most
0.2
1
models (and the ensemble mean) show a nearly linear relationship
between temperature change and cumulative emissions (Fig. 3a),
suggesting that this may be a robust property of the coupled
0.0 0
0 200 400 600 800 1,000 climatecarbon system. Some models do deviate from linearity,
Model year particularly early in the simulations, which is at least partly due to
Figure 2 | Idealized model simulations of the CCR. a, Simulation with a 1% the influence of decadal temperature variability. However, by the
per year atmospheric CO2 increase for 70 years, showing temperature change middle of the twenty-first century, all models converge to an intrinsic
per unit atmospheric carbon increase (DT/DCA: thin red line, right axis), value of temperature change per unit carbon emitted, which remains
airborne fraction of cumulative carbon emissions (DCA/ET: thin blue line, approximately stable for the remainder of the simulation (Fig. 3b).
left axis) and CCR (thick red line, right axis). In this simulation, cumulative CCR values calculated at the time of CO2 doubling in each model
airborne fraction decreased with time owing to a delayed carbon cycle simulation are given in Supplementary Table 1. Model values of CCR
response to a rapid prescribed rate of atmospheric CO2 increase. This is range from 1.0 to 2.1 uC per Tt C, with an ensemble mean value of
consistent with saturating carbon sinks at higher atmospheric CO2, which 1.6 uC per Tt C (see Supplementary Information for additional dis-
leads to an increased airborne fraction of annual emissions with increasing
cussion of model CCR values).
atmospheric CO2. b, Simulations with an instantaneous doubling (solid
lines) and quadrupling (dashed lines) of atmospheric CO2 for 1,000 years The CCR can also be estimated from historical carbon emissions
(colours as in a). In all cases, the cumulative airborne fraction decreased with data and observed temperature changes. To calculate CCR from obser-
time, whereas the temperature change per unit atmospheric carbon vations, we first estimated decadal-mean CO2-attributable warming
increased with time; consequently, the CCR (defined as the product of these relative to 190009 by scaling an estimate of greenhouse-gas-
two quantities) remained constant in time. attributable warming12 by the ratio of CO2 to greenhouse-gas forcing.
We then calculated CCR by dividing CO2-attributable warming by
In these simulations, the CCR is independent of both time and cumulative anthropogenic CO2 emissions between 190009 and each
CO2 emission (or concentration) scenario. At a given CO2 concen- subsequent decade, including emissions from land-use change, fossil
tration (see, for example, Fig. 2b), the time-independence of CCR fuels and cement production (see Methods).
arises from a cancellation of a decreasing airborne fraction of cumu- Figure 4 shows an estimate of CCR for 199099 of 1.02.1 uC per
lative emissions, and an increasing temperature change per unit Tt C (5 to 95% confidence interval), with a best estimate of 1.5 uC per
atmospheric CO2 over time. This may relate in part to the uptake Tt C. Similar estimates of CCR, albeit with larger uncertainties, are
of heat and carbon by the ocean being driven by the same deep-ocean obtained for previous decades. We note that these estimates are less
mixing processes on long timescales3,7. However, as can be seen in contaminated with internal climate variability than those derived from
Fig. 2a and b, CCR is also independent of CO2 concentration and, by single simulations in Fig. 3 because the greenhouse-gas-attributable
extension, of the CO2 emission scenario. This scenario independence warming is based on a scaled ensemble mean of 11 simulations.
emerges owing to the approximate cancellation of the saturation of Nonetheless, assuming the simulated temporal evolution of the green-
carbon sinks and the saturation of CO2 radiative forcing with increas- house gas response is realistic, these results provide further evidence
ing atmospheric CO2. As a result, at higher atmospheric concentra- for the constancy of CCR as a function of time.
tions, a given CO2 emission will lead to a larger increase in Recent climatecarbon model experiments have shown that elimi-
atmospheric CO2, but the temperature change per unit change in nating CO2 emissions leads to approximately stable, or slowly
atmospheric CO2 will be smaller2. decreasing global temperatures over time3,7,13; this implies that close
Even in the extreme case of instantaneous pulse emissions8, the to zero net anthropogenic carbon emissions are required to stabilize
temperature change per unit carbon emitted in the UVic ESCM is global mean temperature3, and conversely that there may be neg-
found to be constant to within 10% on timescales of between 20 and ligible future warming commitment as a result of past CO2 emis-
1,000 years, and for cumulative emissions of up to 2 Tt C (see sions3,7,13. Consequently, the CCR, defined here as the ratio of
Supplementary Fig. 1). As is seen, however, in Fig. 2a, we expect instantaneous temperature change to past CO2 emissions, can also
830
NATURE | Vol 459 | 11 June 2009 506 LETTERS
a 3
BERN-CC IPSL-CM4-LOOP
Carbonclimate response (C per Tt C)

4 CSM-1 LLNL 95th
perc
CLIMBER2-LPJ MPI entile
FRCGC UMD
2
HADCM3LC UVIC-2.7
Temperature change (C)
3 IPSL-CM2C Mean
Best estimate
2 1
5th percentile
1
0
1900 1920 1940 1960 1980 2000
Year
0
0 0.5 1.0 1.5 2.0 Figure 4 | Observational estimates of CCR. CCR was estimated for each
Cumulative carbon emissions (Tt C) decade of the twentieth century after 1910 by scaling an observationally
constrained estimate of greenhouse-gas-attributable warming relative to
b
4 190009 by the ratio of CO2 forcing to total greenhouse gas forcing, and
dividing by cumulative anthropogenic carbon emissions over the same
period. This observationally constrained estimate of CCR is both stable in
time and consistent with the estimates derived from model simulations.
Carbonclimate response (C per Tt C)
quantified uncertainties in historical land-use change emissions and

structural uncertainties in the simulated sulphate aerosol response.
2 For example, the allowable emissions for a particular warming
target calculated by ref. 5 were lower, because they used a higher
observational estimate of CO2-attributable warming as well as a
1 climatecarbon model which simulated non-negligible zero emis-
sions commitment under conditions of high climate sensitivity.
We note also that our analysis of allowable emissions applies specif-
ically to CO2-induced warming, and does not account for the effects
0
of other greenhouse gases or aerosols.
The CCR is a simple, yet robust representation of the global tem-
perature response to anthropogenic CO2 emissions, and as such is
1 directly relevant to current policy negotiations surrounding inter-
1900 1950 2000 2050 2100
Model year
national climate mitigation efforts. The European Union has proposed
restricting global warming to less than 2 uC above pre-industrial tem-
Figure 3 | CCR estimated from the C4MIP simulations11. a, Decadal-average peratures16; however, large uncertainty in equilibrium climate sensi-
temperature change plotted as a function of cumulative carbon emissions, tivity17 prevents confident estimates of the CO2 stabilization level
showing a near-linear relationship for both individual models (coloured
required to avoid 2 uC warming, and climate sensitivity alone provides
lines) and the ensemble mean (black line). b, Temperature change per
cumulative carbon emitted for each decade from 1900 to 2100 relative to the no policy-useful information about the allowable CO2 emissions for a
first decade of each model simulation. Over most of the twenty-first century given stabilization level. The CCR represents a synthesis of previous
portion of the simulations, CCR values in each model are remarkably efforts to quantify the temperature response to anthropogenic CO2
constant in time. emissions by aggregating the uncertainties associated with climate
sensitivity, carbon sinks and climatecarbon feedbacks into a single
well-constrained metric of climate change that is related directly to
be used as an estimate of the centennial-scale temperature legacy of
cumulative carbon emissions.
these emissions. As a result, our estimates of CCR can be inverted to
estimate the total allowable anthropogenic carbon emissions per
METHODS SUMMARY
degree of long-term temperature change.
For the idealized model experiments (1% per year CO2 increase; doubled/quad-
From our model-based estimate of CCR, we estimate allowable rupled CO2) we used the UVic ESCM version 2.8 (refs 9, 1820). The UVic
emissions of 1.25 Tt C (range, 0.952 Tt C) for 2 uC warming relative ESCM is a computationally efficient coupled climatecarbon model, with inter-
to pre-industrial temperature; our observationally based best estimate active representations of three-dimensional ocean circulation, atmospheric
of allowable emissions for 2 uC of warming is 1.4 Tt C (595% con- energy and moisture balances, sea ice dynamics and thermodynamics, dynamic
fidence interval, 1.0 to 1.9 Tt C). Given total CO2 emissions until now vegetation and the global carbon cycle (including land and both inorganic and
of approximately 0.5 Tt C from fossil fuels and land-use change14,15, organic ocean carbon). Version 2.7 of the UVic ESCM was one of the 11 par-
this implies that total future carbon emissions consistent with 2 uC of ticipating models in C4MIP11, in which models were driven by a common CO2
warming must be restricted to a best estimate of about 0.8 Tt C emissions scenario and carbon sinks and atmospheric CO2 concentrations were
calculated interactively until the year 2100. From the C4MIP simulations, we
(0.7 Tt C based on the model ensemble mean; 0.9 Tt C based on obser-
estimated CCR using globally averaged temperature change and accumulated
vational constraints). carbon emissions at the year of CO2 doubling in each simulation.
We emphasize, however, that the calculated uncertainty on this Our observational estimate of CCR was derived using estimates of CO2-attri-
number is quite large (0.4 to 1.5 Tt C). Furthermore, we are unable to butable warming and cumulative CO2 emissions for each decade of the twentieth
exclude the possibility of higher values of CCR (and consequently century relative to 190009. We estimated CO2-attributable warming using an
lower values of allowable emissions), owing particularly to poorly estimate of greenhouse-gas-attributable warming12, scaled by the ratio of CO2 to
831
LETTERS 507 NATURE | Vol 459 | 11 June 2009
total greenhouse-gas forcing21, where greenhouse-gas forcing was first scaled by 15. Houghton, R. A. Carbon flux to the atmosphere from land-use changes:
an estimate of the mean efficacy of long-lived greenhouse gases22. We calculated 18502005. In TRENDS, A Compendium of Data on Global Change (Carbon Dioxide
uncertainties in greenhouse-gas-attributable warming, accounting for internal Information Analysis Center, Oak Ridge National Laboratory, US DOE, 2008).
variability and inter-model uncertainty12, and assumed normally and Student-t 16. European Commission. Limiting global climate change to 2 degrees Celsius: the
way ahead for 2020 and beyond. (Commission of the European Communities,
distributed uncertainties for radiative forcings and greenhouse-gas efficacy,
2007).
respectively22. We calculated cumulative carbon emissions from fossil fuels
17. Meehl, G. A. et al. in Climate Change 2007: The Physical Science Basis (eds Solomon,
and land-use change13,14,23, and assumed a one-sigma systematic uncertainty S. et al.) 747845 (Cambridge University Press, 2007).
on land-use emissions of 60.5 Pg C per year24. Our central estimates for CO2- 18. Meissner, K. J., Weaver, A. J., Matthews, H. D. & Cox, P. M. The role of land-
attributable warming and cumulative emissions at 199099 relative to 190009 surface dynamics in glacial inception: a study with the UVic Earth System Climate
were 0.492 uC and 0.338 Tt C, respectively. We calculated a probability density Model. Clim. Dyn. 21, 515537 (2003).
function for CCR based on the probability distributions of the constituent terms, 19. Matthews, H. D., Weaver, A. J. & Meissner, K. J. Terrestrial carbon cycle dynamics
which we used to estimate the mean and the 5th and 95th percentiles. under recent and future climate change. J. Clim. 18, 16091628 (2005).
20. Schmittner, A., Oschlies, A., Matthews, H. D. & Galbraith, E. D. Future changes in
Full Methods and any associated references are available in the online version of climate, ocean circulation, ecosystems and biogeochemical cycling simulated for
the paper at www.nature.com/nature. a businessas-usual CO2 emissions scenario until year 4000 AD. Glob.
Biogeochem. Cycles 22, GB1013 (2008).
Received 4 December 2008; accepted 14 April 2009. 21. Gregory, J. M. & Forster, P. M. Transient climate response estimated from
radiative forcing and observed temperature change. J. Geophys. Res. 113, D23105
1. Randall, D. A. et al. in Climate Change 2007: The Physical Science Basis (eds
(2008).
Solomon, S. et al.) 589845 (Cambridge University Press, 2007).
2. Caldeira, K. & Kasting, J. F. Insensitivity of global warming potentials to carbon 22. Forster, P. et al. in Climate Change 2007: The Physical Science Basis (eds Solomon, S.
dioxide emissions scenarios. Nature 366, 251253 (1993). et al.) 129234 (Cambridge University Press, 2007).
3. Matthews, H. D. & Caldeira, K. Stabilizing climate requires near-zero emissions. 23. Denman, K. et al. in Climate Change 2007: The Physical Science Basis (eds Solomon,
Geophys. Res. Lett. 35, L04705 (2008). S. et al.) 129234 (Cambridge University Press, 2007).
4. Zickfeld, K., Eby, M., Matthews, H. D. & Weaver, A. J. Setting cumulative 24. Canadell, J. G. et al. Contributions to accelerating atmospheric CO2 growth from
emissions targets to reduce the risk of dangerous climate change. Proc. Natl Acad. economic activity, carbon intensity, and efficiency of natural sinks. Proc. Natl Acad.
Sci. USA. (submitted). Sci. USA 104, 1886618870 (2007).
5. Allen, M. R. et al. Warming caused by cumulative carbon emissions towards the
Supplementary Information is linked to the online version of the paper at
trillionth tonne. Nature 458, 11631166 (2009).
www.nature.com/nature.
6. Shine, K. P., Fuglestvedt, J. S., Hailemariam, K. & Struber, N. Alternatives to the
global warming potential for comparing climate impacts of emissions of Acknowledgements We thank A. Weaver, M. Eby, V. Arora, N. Ramankutty,
greenhouse gases. Clim. Change 68, 281302 (2005). M. Allen, S. Solomon, K. Keller, K. Caldeira and S. Turner for commentary and
7. Solomon, S., Kasper Plattner, G., Knutti, R. & Friedlingstein, P. Irreversible climate discussions on this work. We also thank P. Forster for providing radiative forcing
change due to carbon dioxide emissions. Proc. Natl Acad. Sci. USA 106, 17041709 time series, and P. Friedlingstein and the C4MIP modelling community for the
(2009). availability of their model output. H.D.M. acknowledges support from the National
8. Eby, M. et al. Lifetime of anthropogenic climate change: millennial time-scales of Science and Engineering Research Council of Canada, and the Canadian
potential CO2 and surface temperature perturbations. J. Clim. 22, 25012511 (2009). Foundation for Climate and Atmospheric Sciences Project Grants. P.A.S. was
9. Weaver, A. J. et al. The UVic Earth System Climate Model: model description, supported by the Joint DECC, Defra and MoD Integrated Climate Programme.
climatology and applications to past, present and future climates. Atmos. Ocean N.P.G. received support from the Leverhulme Trust. N.P.G. and P.A.S. acknowledge
39, 361428 (2001). support from the Climate Change Detection and Attribution Project, jointly funded
10. Matthews, H. D. Emissions targets for CO2 stabilization as modified by carbon
by NOAAs Office of Global Programs and the US Department of Energy.
cycle feedbacks. Tellus 55B, 591602 (2006).
11. Friedlingstein, P. et al. Climate-carbon cycle feedback analysis, results from the Author Contributions H.D.M. proposed the study, carried out model simulations
C4MIP model intercomparison. J. Clim. 19, 33373353 (2006). and analysis, and wrote most of the paper. N.P.G. proposed the inclusion of
12. Huntingford, C., Stott, P. A., Allen, M. R. & Lambert, F. H. Incorporating model observational constraints, N.P.G. and P.A.S. carried out this analysis, and N.P.G.
uncertainty into attribution of observed temperature change. Geophys. Res. Lett. wrote the sections of the paper and methods describing these results. K.Z.
33, L05710 (2006). provided additional model simulations and analysis as described in the
13. Plattner, G.-K. et al. Long-term climate commitments projected with climate- Supplementary Information. All authors participated in discussions pertaining to
carbon cycle models. J. Clim. 21, 27212751 (2008). interpretation and presentation of results.
14. Marland, G., Boden, T. A. & Andres, R. J. Global, regional, and national fossil fuel
CO2 emissions. In TRENDS, A Compendium of Data on Global Change (Carbon Author Information Reprints and permissions information is available at
Dioxide Information Analysis Center, Oak Ridge National Laboratory, US DOE, www.nature.com/reprints. Correspondence and requests for materials should be
2008). addressed to H.D.M. (dmatthew@alcor.concordia.ca).
832
doi:10.1038/nature08047 508
METHODS efficacy for long-lived greenhouse gases (shown in figure 2.19 of ref. 22) to
UVic ESCM. The UVic ESCM is an intermediate-complexity coupled climate account for the larger temperature response per unit radiative forcing for other
carbon model. The climate component consists of a reduced-complexity energy greenhouse gases compared to CO2. Tropospheric ozone changes were not
moisture balance atmosphere coupled to a general circulation ocean and specified in the simulations used by ref. 12, so we did not include them in our
dynamic/thermodynamic sea-ice model9. The carbon cycle component of estimate of total greenhouse gas forcing, under the assumption that the response
version 2.8 consists of a biochemical dynamic vegetation model18,19 and an to tropospheric ozone is spatially and temporally dissimilar to that due to the well-
organic/inorganic ocean carbon cycle model20. Version 2.7 of the UVic ESCM mixed greenhouse gases and is therefore unlikely to be aliased in the multiple
was one of the 11 participating models in the C4MIP11, as well as a contributing regression (the inclusion of tropospheric ozone forcing in the total greenhouse-
model to the long-term climate and carbon cycle projections highlighted in ref. 17. gas forcing estimate reduces our observational estimate of CCR to 0.91.8 uC per
C4MIP. The C4MIP compared the simulated climate and carbon cycle changes Tt C). Our calculation also assumes that climate forcings other than CO2 emis-
from 11 coupled climatecarbon models (including seven atmosphereocean sions have had little influence on atmospheric CO2 concentration. This is a
general circulation models, and four intermediate-complexity models)11. reasonable assumption given a near-cancellation over the past century of positive
Models were driven by a common CO2 emissions scenario (including specified non-CO2 greenhouse-gas forcing and negative aerosol forcing.
emissions from both fossil fuels and land-use change), with carbon sinks and Uncertainties in greenhouse-gas-attributable warming were calculated follow-
atmospheric CO2 calculated interactively until the year 2100. To calculate the ing ref. 12; uncertainties in radiative forcings were estimated from ref. 22 (FAQ
CCR for each model, we used globally averaged temperature changes from the 2.1, Fig. 2) and were assumed to be normally distributed; uncertainties in effi-
coupled simulations, along with a running total of specified CO2 emissions. The cacies were estimated from figure 2.19 of ref. 22, and were assumed to be
values of CCR presented here and in the Supplementary Information were cal- Student-t distributed. Land use, fossil fuel and cement emissions were taken
culated using a ten-year average of temperature increases and cumulative emis- from CDIAC14,15. A one-sigma uncertainty on fossil fuel and cement emissions
sions, centred at the time of CO2 doubling in each simulation. of 65% was assumed following ref. 23 and a one-sigma systematic uncertainty
Observationally constrained CCR estimate. We calculated observational esti- on land-use emissions of 60.5 Pg C per year was assumed following ref. 24; both
mates of CCR by taking the ratio of CO2-attributable warming and cumulative were assumed to be normally distributed. A probability density function was
emissions in the decade 190009 and each subsequent decade of the twentieth calculated for CCR based on the probability density functions of the constituent
century. We began with a multi-model estimate of greenhouse-gas-attributable terms, and this was used to derive the mean and the 5th and 95th percentiles. The
warming for each decade of the twentieth century. This was derived by scaling uncertainty in land-use emissions was the largest single contributor to the overall
the mean simulated temperature response to prescribed historical well-mixed uncertainty in CCR. Given this, we tested the sensitivity of our results to setting
greenhouse-gas concentrations from HadCM3, GFDL and PCM to best-fit land-use emissions to zero; this gave an estimate of CCR for the decade 199099
HadCRUT2v temperature observations, based on a multiple regression together of 1.62.7 uC per Tt C, though we emphasize that this should not be taken as a
with the response to sulphate aerosol and natural forcing12. The calculated un- realistic upper bound for CCR, because zero land-use emissions are not consis-
certainty in this greenhouse-gas-attributable warming includes an estimate of tent with observed atmospheric CO2 increases. Uncertainties in the overall
internal variability based on control simulations and an estimate of model un- magnitude of aerosol forcing are fully accounted for in our estimate of green-
certainty based on inter-model differences in forcings and simulated responses12. house-gas-attributable warming; however, uncertainties in the temporal or spa-
We scaled the greenhouse-gas-attributable warming by the ratio of CO2 forcing tial pattern of the response to aerosol forcing are only accounted for to the extent
to total well-mixed greenhouse gas forcing, with all forcings expressed as differ- that they are sampled in the three global climate models we used, and errors in
ences between 190009 and subsequent decades of the twentieth century21. Before these patterns could lead to values of CCR outside our estimated uncertainty
this scaling, we multiplied the well-mixed greenhouse-gas forcing by the mean range.

509
is is Exhibit ".h." refelred to in the

't affidavit of Kirsten Zickfeld,
affirmed before mc on the
day of April, 2017.
A commissioner for'
Blitish Columbia
510
Setting cumulative emissions targets to reduce the

risk of dangerous climate change
Kirsten Zickfelda,1,2, Michael Ebya, H. Damon Matthewsb, and Andrew J. Weavera
aSchoolof Earth and Ocean Sciences, University of Victoria, Victoria, BC, Canada V8W 3V6; and bDepartment of Geography, Planning, and Environment,
Concordia University, Montreal, QC, Canada H3G 1M8
Edited by Hans Joachim Schellnhuber, Potsdam Institute for Climate Impact Research, Potsdam, Germany, and approved July 20, 2009 (received for review
June 16, 2008)
Avoiding dangerous anthropogenic interference with the climate approach, whereby we work backwards from a specified tem-
system requires stabilization of atmospheric greenhouse gas perature target to CO2 emissions (12, 13); (ii) extend the
concentrations and substantial reductions in anthropogenic emis- cause-effect chain to include the linkage between CO2 emissions
sions. Here, we present an inverse approach to coupled climate- and concentrations (through an explicit representation of the
carbon cycle modeling, which allows us to estimate the probability carbon cycle); and (iii) use a state-of-the-art coupled climate-
that any given level of carbon dioxide (CO2) emissions will exceed carbon cycle model. This approach allows us to gain new insights
specified long-term global mean temperature targets for danger- into the transient evolution of the coupled climate-carbon cycle
ous anthropogenic interference, taking into consideration uncer- system toward temperature stabilization and to consistently
tainties in climate sensitivity and the carbon cycle response to derive cumulative CO2 emissions levels to reduce the risk of
climate change. We show that to stabilize global mean tempera- DAI.
ture increase at 2 C above preindustrial levels with a probability We adopt a risk management approach, whereby key uncertain-
of at least 0.66, cumulative CO2 emissions from 2000 to 2500 must ties in the coupled system are described in terms of probability
not exceed a median estimate of 590 petagrams of carbon (PgC) density functions (PDFs). Major uncertainties in the long-term
(range, 200 to 950 PgC). If the 2 C temperature stabilization target climate system response to specified emissions trajectories include
is to be met with a probability of at least 0.9, median total uncertainties in the climate sensitivity and the response of the
allowable CO2 emissions are 170 PgC (range, 220 to 700 PgC). marine and terrestrial carbon sinks to climate change (14). The
Furthermore, these estimates of cumulative CO2 emissions, com- climate sensitivity summarizes the feedbacks in the response of the
patible with a specified temperature stabilization target, are in- physical climate system to radiative forcing and is here defined as
dependent of the path taken to stabilization. Our analysis there- the expected equilibrium global mean surface temperature re-
fore supports an international policy framework aimed at avoiding sponse to a doubling of the preindustrial atmospheric CO2 con-
dangerous anthropogenic interference formulated on the basis of centration. In recent years, a growing number of studies have
total allowable greenhouse gas emissions. attempted to constrain climate sensitivity from present-day clima-
tology (15, 16, 17, 18), the historical temperature evolution (19, 20,
climate carbon cycle feedbacks cumulative emissions budget 21, 22, 23, 24), and paleo-climatic records (25). However, climate
dangerous anthropogenic interference uncertainty analysis 2 C target sensitivity is only weakly constrained by any of these observations,
so that large uncertainty remains as to its value. Here, we use several
of the published PDFs for climate sensitivity to derive the likelihood
T he ultimate goal of climate policies is to reduce the amount
of anthropogenic greenhouse gas (GHG) emissions to
achieve stabilization of greenhouse gas concentrations in the
that a given temperature target is exceeded for a specific CO2
emissions level.
atmosphere at a level that would prevent dangerous anthropo- Unlike climate sensitivity, no probabilistic measure has been
genic interference with the climate system (United Nations derived for the uncertainty in the carbon cycle response. The
Framework Convention on Climate Change, Article 2). Com- modification of the marine and terrestrial carbon sinks because
monly, dangerous anthropogenic interference (DAI) is charac- of climate change is usually quantified in terms of the climate-
terized in terms of the impacts of climate change. For instance, carbon cycle feedback (26). Although varying widely in magni-
the Intergovernmental Panel on Climate Change (IPCC) devel- tude, this feedback is positive in most models (14), implying
oped a framework that relates different categories of impacts, faster accumulation of CO2 in the atmosphere. Accordingly,
ranging from impacts on unique and vulnerable systems to climate-carbon cycle feedbacks reduce the amount of emissions
large-scale geophysical discontinuities, to the level of tempera- compatible with a given CO2 concentration or global mean
ture change at which they are likely to occur (1, 2). Despite temperature target (27, 28). In our analysis, we incorporate
increasing evidence about the consequences of climate change, carbon cycle uncertainty by exploring the sensitivity of the
there can be no objective scientific definition of what constitutes resulting allowable CO2 emissions to the strength of the climate-
DAI, because such a definition is ultimately a normative deci- carbon cycle feedback.*
SUSTAINABILITY
sion, influenced by value judgments. In recent years, interna-

SCIENCE
tional climate policy discussions have been framed around

Author contributions: K.Z., M.E., H.D.M., and A.J.W. designed research; K.Z. performed
limiting the global mean temperature increase to 2 C relative to research; K.Z., M.E., and H.D.M. contributed new reagents/analytic tools; K.Z. and M.E.
preindustrial times, a number that some have argued represents analyzed data; and K.Z. wrote the paper.
a threshold beyond which climate impacts become dangerous. The authors declare no conflict of interest.
For example, in 1996 the European Council adopted the target This article is a PNAS Direct Submission.
to limit global mean warming to 2 C. This target has since been 1To whom correspondence should be addressed. E-mail: kirsten.zickfeld@ec.gc.ca.
reaffirmed by the European Union on a number of occasions, 2Present address: Canadian Centre for Climate Modelling and Analysis, Environment
such as March 2005 (3) and January 2007 (4). Canada, Victoria, BC, Canada V8W 3V6.
In this study we derive allowable CO2 emissions levels that are *Note that we neglect the part of the overall carbon sink uncertainty associated with
compatible with a set of long-term temperature targets to avoid uncertainty in the drivers of sinks (e.g., CO2 fertilization).
DAI. Our study differs from earlier attempts to derive safe This article contains supporting information online at www.pnas.org/cgi/content/full/
emissions levels (5, 6, 7, 8, 9, 10, 11) in that we (i) use an inverse 0805800106/DCSupplemental.
www.pnas.orgcgidoi10.1073pnas.0805800106 PNAS September 22, 2009 vol. 106 no. 38 16129 16134
511
We use the University of Victoria Earth System Climate

Model (UVic ESCM) version 2.8 (29) (see Methods). The model
A
1000
was integrated over the historical period (18002000) by using
Cumulative emissions (PgC)

known natural and anthropogenic forcings (including forcing-
from-greenhouse gases such as CO2, CH4, NO2, and halocar- 800
bons, sulfate aerosols, land-use change, solar irradiance, volca-
noes and orbital changes). The resulting change in global mean 600
temperature in the year 2000 is 0.68 C relative to preindustrial
times. Following an inverse approach, we then performed an
ensemble of model simulations spanning the 20002500 period, 400
whereby we diagnosed CO2 emissions compatible with pre-
scribed temperature trajectories. These simulations differed in 200
the prescribed temperature stabilization profile and the value of
the climate sensitivity (see Methods).
0
Results and Discussion 2000 2100 2200 2300 2400 2500
Time (calendar year)
Our results indicate that the cumulative CO2 emissions compatible
with a given global mean temperature target are independent of the
path taken to stabilization. Fig. 1 illustrates this path independency B 600
for trajectories stabilizing at 2 C (relative to preindustrial): No
matter whether temperature is stabilized in 2150, 2200, or 2300, or 550
the target is temporarily exceeded (overshoot scenario), admis-
sible cumulative CO2 emissions converge to the same value at about
CO 2 (ppmv)
2350. In 2500, cumulative emissions (computed from the beginning 500
of 2001) are 716 PgC for all temperature trajectories. For all
profiles, stabilization at 2 C requires atmospheric CO2 to peak and
decline afterward to compensate for the oceans thermal inertia. In 450
the long-term, atmospheric CO2 must be 450 parts per million by
volume (ppmv), given the standard value for climate sensitivity of
400
the UVic ESCM (3.6 C). Note that in the case of the overshoot
scenario (light blue curve in Fig. 1), the 2 C stabilization target can
only be met if CO2 is artificially removed from the atmosphere, 350
resulting in negative emissions. 2000 2100 2200 2300 2400 2500
The above finding has important implications as it allows one Time (calendar year)
to relate CO2 cumulative emissions to the temperature target
independently of the specific trajectory taken to stabilization. If 3.0
C
Surface air temperature anomaly (oC)
global mean temperature thresholds for DAI of 3 C and 4 C are

assumed, allowable cumulative emissions from 2001 to 2500 are
found to be 1,260 PgC and 1,790 PgC, respectively. In terms of 2.5
long-term atmospheric CO2 concentrations, the above targets
imply levels of 550 ppmv and 700 ppmv (red-dashed lines in
2.0
Fig. 2).
These figures are highly sensitive to the value of climate
sensitivity. Using an uncertainty range for climate sensitivity 1.5
(CS) of 1 C to 9 C, which encompasses the 595% range of
most published PDFs for climate sensitivity (Box 10.2 in ref. 34),
we find that the admissible cumulative emissions from 2001 to 1.0
2500 vary widely (Fig. 2). The range is 110 to 4,230 PgC for the
2 C temperature stabilization target, 30 to 8,080 PgC for the
0.5
3 C target, and 140 to 15,940 PgC for the 4 C target. Fig. 2 2000 2100 2200 2300 2400 2500
indicates that the sensitivity is largest at low values of CS and Time (calendar year)
decreases with increasing values of CS. Interestingly, for all
three targets long-term temperature stabilization results in Fig. 1. Path independency of cumulative CO2 emissions. (A) Cumulative CO2
near-constant cumulative emissions, implying that annual emis- emissions and (B) CO2 concentrations compatible with a global mean temper-
ature increase of 2 C relative to preindustrial times. The different curves refer
to experiments with different prescribed temperature change trajectories (C).
Note that this estimate of allowable emissions, compatible with the 2 C target, differs The red-dashed trajectory is the standard trajectory used throughout the
from the value of 500 PgC that can be inferred from an earlier study by the authors (30). analysis. Cumulative emissions are computed from the year 2001 onwards.
The reason for this difference is the use of a different solar forcing over the historical
period. In ref. 30, we employed the solar forcing of Lean (31). Here, we use the forcing of
Krivova et al. (32), which better reflects the most recent IPCC estimate of a change in
radiative forcing of 0.12 Wm2 from 1750 to present day (33).
sions must decrease to zero (35). However, this result is less
true for very low CS, which allows small positive CO2 emissions
The reason is twofold: First, this nonlinearity is due to the logarithmic dependency of
radiative forcing on atmospheric CO2, which implies that at higher CS a smaller increment
far into the future, or for very high CS, which requires negative
in CO2 is required to attain the same global mean temperature change than at lower CS. future emissions.
Second, the sink capacity of the biosphere is greatly enhanced at low values of CS. In fact, As discussed above, the sensitivity of allowable cumulative
given the same temperature target, admissible CO2 concentrations are significantly higher
emissions to the assumed CS value is large, with lower limits of
if a low CS is assumed than under assumption of a high CS. Because our model simulates
a substantial terrestrial carbon sink in response to higher atmospheric CO2 levels (CO2 the computed ranges being rather ambitious and upper limits
fertilization), relatively higher emissions are allowed at lower CS values. exceeding the estimate of carbon bound in known fossil fuel
16130 www.pnas.orgcgidoi10.1073pnas.0805800106 Zickfeld et al.

512
S W S W S W
4000 A 4000 B 4000 C
Cumulative emissions (PgC)
3000 T = 2 oC 3000 T = 3 oC 3000 T = 4 oC
2000 2000 2000
1000 1000 1000
0 0 0
2000 2100 2200 2300 2400 2500 2000 2100 2200 2300 2400 2500 2000 2100 2200 2300 2400 2500
1400 D cs=2
cs=3
cs=6
cs=7
1400 E 1400 F
cs=3.6 cs=8
1200 1200 1200
T = 2 oC cs=4 cs=9 T = 3 oC T = 4 oC
CO2 (ppmv)
cs=5
1000 1000 1000
800 800 800
600 600 600
400 400 400
2000 2100 2200 2300 2400 2500 2000 2100 2200 2300 2400 2500 2000 2100 2200 2300 2400 2500
Time (calendar year) Time (calendar year) Time (calendar year)
Fig. 2. Dependence of cumulative CO2 emissions on climate sensitivity. Cumulative CO2 emissions (AC) and CO2 concentrations (DF) compatible with a
long-term global mean temperature increase, (T), of 2 C (A and D), 3 C (B and E) and 4 C (C and F) relative to preindustrial times (the prescribed temperature
trajectories are displayed in Fig. S3). The different curves refer to experiments with climate sensitivities (cs) in the range 29 C (the curves for cs 1 C are not
shown for clarity). The standard climate sensitivity of the UVic ESCM is 3.6 C (red-dashed curves). (AC Right) Cumulative emissions in 2500 under assumption
of weak (W) and strong (S) climate-carbon cycle feedbacks are shown.
reserves (5,000 PgC) (36). In terms of climate policy, such a need to be 1,020 PgC (range, 7001,300 PgC). The associated
large range implies responses ranging from very stringent CO2 median value for the 4 C target is 1,450 PgC (range, 1,000 to
mitigation policies to virtually no mitigation at all (business- 1,900 PgC).
as-usual). Clearly, to be of any value for climate decision- It can be expected that these figures are sensitive to the
making, the above estimates of cumulative emissions need to be strength of the climate-carbon cycle feedback. According to the
weighted by their associated probabilities. To obtain such a Coupled Carbon Cycle Climate Model Intercomparison Project
weighting, we used several published PDFs for climate sensitivity (C4MIP) (14), the UVic ESCM simulates a climate-carbon cycle
(1620, 22, 23). For each value of CS used in this study (19 C), feedback that lies approximately in the middle of the range
we then computed the cumulative probability that the actual spanned by state-of-the-art climate-carbon cycle models. If the
value of CS is larger then the assumed value, for one specific feedback simulated by the UVic ESCM was stronger, allowable
PDF. For each simulation with a specific CS, we can then relate emissions under a specific temperature threshold for DAI would
this probability to the probability that the associated allowable be lower, and vice versa. To estimate the range of permissible
CO2 emissions would lead to exceeding the specified tempera- emissions under consideration of different strengths of the
ture-stabilization target. Assume, for instance, a CS of 3 C. For climate-carbon cycle feedback, we reconstructed the allowable
the PDF of Forest et al. (22), there is a 0.67 cumulative emissions that would be simulated by the C4MIP models with the
probability that the actual CS value is larger than 3 C. It follows strongest (HadCM3LC) and the weakest (IPSL-CM4-LOOP)
that the probability that the allowable cumulative emissions feedback under the same temperature trajectory. This recon-
computed with a CS of 3 C will exceed the intended target in struction was achieved by scaling the reduction in allowable
2500 is also 0.67. emissions because of the feedback as simulated by the UVic
Fig. 3 displays the probabilities of exceeding temperature ESCM based on the respective normalized gain factor (see
stabilization targets of 24 C for a range of cumulative emis- Methods). Fig. 2 AC Right displays allowable CO2 emissions in
sions and seven published PDFs for climate sensitivity. For a 2500 under assumption of weak (IPSL-CM4-LOOP) and strong
specific temperature target, the probability of exceeding that (HadCM3LC) climate-carbon cycle feedbacks. Taking this ad-
SUSTAINABILITY
target increases with higher cumulative emissions. Conversely, ditional uncertainty into account, the total range in year-2500
SCIENCE
lower acceptable probabilities of exceeding require lower emis- cumulative emissions for the 2 C target is 260 to 4,410 PgC
sions levels. We find that for the 2 C target, cumulative CO2 (compared with 110 to 4,230 PgC, taking into account uncer-
emissions after 2000 need to be limited to a median value of 550 tainty in climate sensitivity alone). Note that for a specific
PgC (with a range of 300770 PgC depending on the climate temperature target the range in allowable emissions because of
sensitivity PDF), if the probability of exceeding that target is to uncertainty in the climate-carbon cycle feedback is independent
be kept within the unlikely domain (corresponding to a of the CS value, because the strength of the feedback depends
probability between 0.1 and 0.33 in the IPCC terminology) (37). only on the global mean temperature change and not on the
If the acceptable probability of exceeding the 2 C target is taken atmospheric CO2 level. Accordingly, the relative uncertainty
to be 0.1, allowable CO2 emissions must not exceed a median range because of climate-carbon cycle feedbacks is smaller at
of 120 PgC (with a range of 100 to 500 PgC). Assuming a lower CS.
threshold for DAI of 3 C and an acceptable probability of Fig. 3 displays the effect of different assumptions regarding
exceeding that target of 0.33, allowable cumulative emissions the climate-carbon cycle feedback on the probability of exceed-
Zickfeld et al. PNAS September 22, 2009 vol. 106 no. 38 16131
513
climate-carbon cycle feedback. If the acceptable probability of

A 1.0 exceeding the 2 C target is taken to be 0.33 (keeping it in the
likely
very
unlikely domain), median allowable CO2 emissions, under in-
clusion of the uncertainty in both climate sensitivity and the
Probability of exceeding 2oC
likely
0.8 climate-carbon cycle feedback, are 590 PgC (range, 200 to 950
PgC). If the acceptable probability of exceeding the 2 C target
is taken to be 0.1 (very unlikely), allowable CO2 emissions must
0.6 Knutti 06, Fig. 7 dashed not exceed a median of 170 PgC (range, 220 to 700 PgC).
likelihood
medium
Forest 02, uniform prior
We emphasize that the allowable emissions estimates pre-
Forest 06, uniform prior
0.4 sented in this study refer purely to CO2 and are only valid more
Hegerl 06, all proxies
Gregory 02
generally if the radiative forcing of other greenhouse gases
continues to be approximately compensated by that of sulfate
unlikely
Murphy 04, weighted
0.2 Piani 05 aerosols, as has been approximately true in the past.
Although the probabilities of exceeding given in this article
unlikely
have been derived for the long-term (year-2500) temperature
very
0.0 response, they also apply, with some approximation, to the
0 1000 2000 3000 4000
instantaneous temperature response. Fig. S1 demonstrates that
Cumulative emissions 2001-2500 (PgC)
even if the total allowable emissions compatible with the 2 C
target are emitted very quickly, the resulting temperature anom-
B 1.0 aly exceeds the target transiently by only a very small amount (SI
likely
very
Text).
It is notable that the range of allowable CO2 emissions for 2 C
likely
0.8
warming derived from transient climate-carbon model simula-
tions (38), albeit over a more limited range of climate sensitiv-
0.6 ities, are nevertheless consistent with the results we report here.
Our estimate of allowable emissions compatible with the 2 C
likelihood
medium
target is not directly comparable to that of Meinshausen et al.

0.4 (39), who consider emissions budgets for the 20002050 period.
In view of climate policy, an interesting question regards the
least-cost allocation of allowable CO2 emissions over time. The
unlikely
0.2 cumulative emissions estimates presented in this article, which

have been derived by using a three-dimensional, state-of-the-art
unlikely
climate model, can be used as a constraint for the derivation of

very
0.0
economically optimal emissions paths by using a model of the
0 1000 2000 3000 4000
Cumulative emissions 2001-2500 (PgC) energy-technology sector and the world economy. Because in
such a cost-effectiveness framework no additional specification
of climate parameters is required, the resulting optimal emis-
1.0 sions trajectory would be entirely consistent with the tempera-
C
likely
very
ture targets initially prescribed to the UVic ESCM.

The independence of the amount of total allowable emissions
likely
0.8
on the emissions trajectory found in this study supports an
international policy framework aimed at avoiding DAI, which is
0.6 formulated on the basis of total allowable GHG emissions. Such
a framework could avoid some of the complications arising from
likelihood
medium
getting all major GHG emitting nations to agree to a common

0.4 timetable for emissions reductions. Given the large uncertainties
associated with total emissions compatible with long-term tem-
perature targets, interim revision of the allowable emissions in
unlikely
0.2 light of the latest scientific evidence should be warranted under

such a framework. Also, by focusing on absolute temperature
unlikely
very
0.0
targets, such an approach would neglect the role of path-
0 1000 2000 3000 4000
Cumulative emissions 2001-2500 (PgC) This value was calculated based on the median of the seven PDFs for climate sensitivity
used in this study and the median feedback gain factor from ref. 14, normalized to UVic
Fig. 3. Probability of exceeding specified global mean temperature targets
ESCMs climate sensitivity.
for different CO2 emissions levels. (A) 2 C target. (B) 3 C target. (C) 4 C
If
target. The brighter colors denote the range in emissions spanned by the the focus is on compliance with maximal or instantaneous (as opposed to long-term)
different climate sensitivity PDFs, and the weaker colors denote the range temperature targets, one has to take the cumulative emissions at the time of temperature
spanned by additional uncertainty in the strength of the climate-carbon cycle stabilization instead of 2500. The reason is that holding temperature constant over several
centuries allows for extra emissions (or, in the case of high climate sensitivities, requires
feedback. Using the nomenclature suggested by the IPCC (37), we have
negative emissions; see Fig. 2 and related discussion). To interpret the given probabilities
divided the probability range into very unlikely (0.01 P 0.1), unlikely in terms of transient overshoot probabilities is therefore only an approximation, which
(0.1 P 0.33), medium likelihood (0.33 P 0.66), likely (0.66 P becomes less valid as the cumulative emissions at the time of temperature stabilization
0.9), and very likely (0.9 P 0.99). differ from those at 2500 (i.e., for high and low climate sensitivities).
Unlike the approach suggested here, inconsistencies are likely to arise if previously
determined allowable CO2 concentration levels are used to derive cost-efficient CO2
emissions pathways, for instance by so-called integrated assessment models, as the
ing specified temperature targets. Generally, the probability that
simulated climate response, which in turn determines the strength of the climate-carbon
any given emissions level will exceed the specified temperature cycle feedback that may differ between the model originally used to derive the safe CO2
target decreases (increases) under assumption of a weak (strong) stabilization level and the integrated assessment model.

514
dependencies in causing DAI. In fact, if the intended temper- optimize the tracking, by making sure that the control parameter, E(t), would
ature target is reached earlier, implying faster rates of climate not react too quickly nor too slowly to the diagnosed temperature error. Fig.
change, or temporarily exceeded, dangerous climate impact S4 Top demonstrates the accuracy of this procedure. The future temperature
tracking runs were integrated from a common year-2000 state to the year
thresholds are more likely to be crossed (7, 40).
2500 by holding all non-CO2 forcings (except orbital forcing) fixed at year-
2000 levels.
Materials and Methods
Model Description. The UVic ESCM version 2.8 consists of a 19-layer ocean
Climate Sensitivity Uncertainty. We constructed an ensemble of 13 model
general circulation model with isopycnal mixing and a Gent and McWilliams
versions for which the equilibrium climate sensitivity varies from 1 C to 9 C.
(41) parameterization of the effect of eddy-induced tracer transport. For
To alter the equilibrium temperature response to a doubling of the atmo-
diapycnal mixing, it uses a horizontally constant profile of diffusivity with
spheric CO2 concentration, we included an adjustable temperature-longwave
values of 0.3 104 m2s1 in the pycnocline. The ocean model is coupled to
radiation feedback (50):
a dynamic-thermodynamic sea-ice model and an energy-moisture balance
model of the atmosphere (29). Atmospheric heat and moisture transports are
Lout t Loutt cTt To
parameterized through Fickian diffusion. All experiments presented here are
computed with advection of heat and specific humidity. The atmospheric where Lout is the unmodified outgoing longwave radiation and Lout
is the new
model includes a parameterization of water vapor/planetary long-wave feed- outgoing longwave radiation. The feedback term is proportional to the
backs, and the radiative forcing associated with changes in atmospheric CO2 difference between global mean surface air temperature and the year-1800
is included as a modification of planetary long-wave flux. The land surface and temperature, T(t) T0. The chosen parameterization serves to increase the net
terrestrial vegetation are represented by a simplified version of the Hadley climate feedback by damping the increase in outgoing longwave radiation
Centres MOSES land-surface scheme coupled to the dynamic vegetation with increased surface temperature (mimicking, for example, a strong positive
model TRIFFID (42). Ocean carbon is simulated by means of a OCMIP-type cloud feedback). The proportionality constants, c, corresponding to specific
inorganic carbon-cycle model (J. Orr, R. Najjar, C. Sabine, and F. Joos, Abiotic equilibrium climate sensitivities, were determined from a set of prior model
how-to document, 2000, available at http://www.ipsl.jussieu.fr/OCMIP) and a simulations. For those runs (seven in total), we specified values of the constant
marine ecosystem model solving prognostic equations for nutrients, phyto- that we guessed would give rise to climate sensitivities in the chosen range
plankton, zooplankton, and detritus (43). Sediment processes are represented (19 C). We then forced the model with an instantaneous doubling of the
using an oxic-only model of sediment respiration (44). The coupled model has preindustrial CO2 concentration and diagnosed the equilibrium global mean
a resolution of 3.6 in longitude and 1.8 in latitude and conserves energy, temperature response. For chosen climate sensitivities lying between diag-
water, and carbon to machine precision without the use of flux adjustment. nosed values, we derived the respective constant by interpolation. To repro-
The model has participated in a number of model intercomparison projects duce the observed historical temperature trajectory for model versions with
including the C4MIP (14), the Paleoclimate Modeling Intercomparison Project climate sensitivities different from the standard value of the UVic ESCM
(PMIP) (45), and the coordinated thermohaline circulation experiments (46, (3.6 C), we compensated the feedback associated with different climate
47). In addition, the model was used as an assessment tool in the 4th Assess- sensitivities [i.e., the term c(T(t) T0)] by adjusting poorly known transient
ment Report of the IPCC (34). radiative forcings (mainly from sulfate aerosols) through the historical period.
This procedure ensures that all future simulations are initialized from a
Historical Simulations. The model was spun up for 10,000 years under year- common year-2000 state of the model. After 2000, the forcing anomalies
1800 radiative forcing (prescribed atmospheric CO2 concentration of 284 (relative to the standard case) are taken to vary proportionally to CO2 emis-
ppm). Afterward, the model was integrated in the fully coupled mode for sions, eventually declining to near zero. The motivation for this choice is that
another 5,000 years. Over the last 4,000 years of simulation, the model did not emissions of sulfate aerosol precursors, such as SO2, are tightly linked to CO2
exhibit any discernible drift. The model was then integrated forward to the emissions. Note that the rate of ocean heat uptake is the same for all model
end of year 2000 by prescribing the observed atmospheric CO2 profile (with versions. The reason is that in a complex three-dimensional coupled climate-
2000 having an average concentration of 368 ppm). The climatic effect of carbon cycle model such as the UVic ESCM it is a difficult task to generate
land-use change over the 20th century was accounted for by changing spec- different model versions with different mixing parameterizations and a rea-
ified surface albedo in regions of croplands (48). Other forcings varying over sonable preindustrial state and transient response over the historical period.
the historical period included non-CO2 greenhouse gases (CH4, N2O, halocar- In such a model (as opposed to two-dimensional physical climate models such
bons), sulfate aerosols (direct effect), solar irradiance, volcanoes, and orbital as those used in refs.15 and 20), varying the mixing parameterization would
changes. The resulting temperature change in 2000 was 0.68 C. Fig. S2 not only alter the ocean heat uptake, but would also degrade the spatial
displays the temperature trajectory for 1800 2000 as compared with obser- distribution of physical and biogeochemical tracers, such as salinity, temper-
vational data (49). To ensure a smooth transition for future projections, we ature, dissolved inorganic carbon, alkalinity, etc.
performed a second simulation over the historical period (1800 2000) by
using the temperature trajectory from the experiment described above to Climate-Carbon Cycle Feedback Uncertainty. A common procedure to diagnose
diagnose CO2 emissions compatible with that trajectory (with all non-CO2 the climate-carbon cycle feedback is to compute the difference in atmospheric
forcings specified as above). The procedure used to track a specified tem- CO2 between the fully coupled experiment and an experiment where the
perature profile is described below. The year-2000 state of the coupled additional CO2 is treated as a radiatively inactive gas (referred to as the
climate-carbon cycle system as computed by this historical run was used to uncoupled experiment) (14, 26). When forced by historical CO2 emissions for
initialize the model simulations spanning the 2000 2500 period. 1800 2000 and the Special Report on Emission Scenarios (SRES) A2 scenario
thereafter, the climate-carbon cycle feedback simulated by the UVic ESCM is
Future Simulations. For the future simulations we diagnosed CO2 emissions 110 ppm in 2100. The associated feedback gain factor is 0.18, lying approxi-
compatible with prescribed temperature trajectories. These temperature pro- mately in the middle of the range simulated by the climate-carbon cycle
files were synthetically constructed by specifying an initial rate of temperature models that participated in the C4MIP model intercomparison (14). In the case
SUSTAINABILITY
change (which was taken to be the rate of temperature change averaged over of the inverse modeling framework used here, an alternative method to
SCIENCE
the last 10 years of the historical simulation), a target temperature, and a year compute the feedback has to be applied, because the uncoupled experiment
of temperature stabilization. The standard profiles used in this study for described above is not meaningful (if the modeled atmosphere does not
stabilization at 2 C, 3 C, and 4 C (relative to preindustrial) are shown in Fig. feel the climatic effects of additional CO2, the allowable cumulative emis-
S3. The temperature tracking was implemented through a simple propor- sions compatible with a given temperature profile would be infinitely large).
tional control equation, of the form We proceed as for the standard temperature tracking experiments described
above, but prescribe preindustrial atmospheric CO2 levels to the ocean and the
Et pkTDATAt Tt terrestrial biosphere. From these experiments we can then derive the reduc-
tion in allowable emissions because of climate-carbon cycle feedbacks by
where E(t) are CO2 emissions, and TDATA(t) T(t) is the error between diagnosing the effect of climate change on ocean and land carbon invento-
prescribed and simulated temperature change at a specific time, t. The pro- ries. We find that the reductions in allowable emissions are 247 gigatons of
portionality constant pk includes factors converting temperature to CO2 carbon (GtC), 392 GtC, and 522 GtC in 2500 for climate targets of 2 C, 3 C, and
concentrations (CO2 concentration divided by climate sensitivity) and CO2 4 C, respectively. These numbers are independent of the value of climate
concentrations to emissions. Also, pk is taken to be inversely proportional to sensitivity, because the feedback depends only on the change in global mean
the response time scale of the system. This time scale was tuned so as to temperature (which is the same given one stabilization target) and not on the
Zickfeld et al. PNAS September 22, 2009 vol. 106 no. 38 16133
515
atmospheric CO2 level. Fig. S5 displays allowable emissions with and without emissions that would be simulated by the C4MIP models by scaling the reduc-
climate-carbon cycle feedbacks for the three chosen temperature targets. tion in allowable emissions computed with the UVic ESCM based on the ratio
Given the uncertainty in the strength of the climate-carbon cycle feedback, of the respective models normalized gain factor to that of the UVic model
the curves without feedback can be interpreted as upper bounds on the (28). Note that here we use a g value for the UVic ESCM of 0.18 and an value
allowable emissions. To estimate the range of permissible emissions consid- of 0.0055 C ppm1 that are lower than the values indicated in the C4MIP study
ering different strengths of the climate-carbon cycle feedback, we drew on (14), which were calculated based on an earlier version of the model. Note also
results of the C4MIP model intercomparison (14). The feedback simulated by that in our procedure we assume that the relation between the models gain
the participating models varies widely, with the feedback gain factor (g) factors remains constant through time and across emissions scenarios, which
is not entirely accurate (14).
ranging from 0.04 to 0.31. Because these gain factors are dependent on the
respective models climate sensitivity, we normalized them to the climate
ACKNOWLEDGMENTS. This work was supported by the Canadian Foundation
sensitivity of the UVic ESCM, using the year-2100 surface-air temperature
for Climate and Atmospheric Sciences Polar Climate Stability Research Net-
change per unit CO2 concentration change () listed in ref. 14. This procedure work grant and the Climate Decision Making Centre, which has been created
gives a range in normalized gain factors of 0.05 (for the IPSL-CM4-LOOP through a cooperative agreement between the National Science Foundation
model) to 0.26 (HadCM3LC). We then reconstructed the range in allowable (SES-0345798) and Carnegie Mellon University.
1. Smith J, Schellnhuber HJ, Mirza M (2001) Lines of evidence for vulnerability to climate 26. Cox PM, Betts RA, Jones CD, Spall SA, Totterdell IJ (2000) Acceleration of global
change: A synthesis. In Climate Change 2001: Impacts, Adaptation and Vulnerability - warming due to carbon-cycle feedbacks in a coupled climate model. Nature 408:184
Contribution of Working Group II to the Third Assessment Report of the IPCC, eds 187.
McCarthy J, Canziani O, Leary N, Dokken D, White K (Cambridge Univ Press, Cambridge, 27. Matthews HD (2006) Emissions targets for CO2 stabilization as modified by carbon cycle
UK), pp 914 967. feedbacks. Tellus B 58:591 602.
2. Schneider S, et al. (2007) Assessing key vulnerabilities and the risk from climate change. 28. Jones CD, Cox PM, Huntingford C (2006) Climate-carbon cycle feedbacks under stabi-
In Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of lization: Uncertainty and observational constraints. Tellus B 58:603 613.
Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on 29. Weaver AJ, et al. (2001) The UVic Earth System Climate Model: Model description,
Climate Change, eds Parry M, Canziani O, Palutikof J, Van der Linden P, Hanson C climatology, and applications to past, present and future climates. Atmos Ocean
(Cambridge Univ Press, Cambridge, UK), pp 779 810. 39:361 428.
3. European Council Presidency Conclusions (2005) (Brussels). Available at http:// 30. Weaver AJ, Zickfeld K, Montenegro A, Eby M (2007) Long term climate implications of
europa.eu.int/rapid/pressReleasesAction.do?referenceDOC/05/1&f. 2050 emission reduction targets. Geophys Res Lett 34:L19703.
4. Commission for European Communities (2007) Limiting Global Climate Change to 2 31. Lean J (2000) Evolution of the suns spectral irradiance since the maunder minimum.
Degrees Celsius: The Way Ahead for 2020 and Beyond (Commission for European Geophys Res Lett 27:24252428.
Communities, Brussels). 32. Krivova NA, Balmaceda L, Solanki SK (2007) Reconstruction of solar total irradiance
5. Wigley TML (2004) Choosing a stabilization target for CO2. Clim Change 67:111. since 1700 from the surface magnetic flux. Astron Astrophys 467:335346.
6. Kriegler E, Bruckner T (2004) Sensitivity analysis of emissions corridors for the 21st 33. Forster P, Ramaswamy V (2007) Changes in Atmospheric Constituents and in Radiative
century. Clim Change 66:345387. Forcing. In Climate Change 2007: The Physical Science Basis, Contribution of WG I to
7. Schneider SH, Mastrandrea MD (2005) Probabilistic assessment of dangerous climate the Fourth Assessment Report of the IPCC eds Solomon S, Qin D, Manning M
change and emissions pathways. Proc Natl Acad Sci USA 102:15728 15735. (Cambridge Univ Press, Cambridge, UK), pp 129 234.
8. Knutti R, Joos F, Muller SA, Plattner GK, Stocker TF (2005) Probabilistic climate change 34. Meehl G, Stocker T (2007) Global climate projections. In Climate Change 2007: The
projections for CO2 stabilization profiles. Geophys Res Lett 32:L20707. Physical Science Basis, Contribution of WG I to the Fourth Assessment Report of the
9. Meinshausen M (2006) What does a 2 C target mean for greenhouse gas concentra- IPCC eds Solomon S, Qin D, Manning M (Cambridge Univ Press, Cambridge, UK), pp
tions? A brief analysis based on multi-gas emissions pathways and several climate 747 845.
sensitivity uncertainty estimates. In Avoiding Dangerous Climate Change, eds Schelln- 35. Matthews HD, Caldeira K (2008) Stabilizing climate requires near-zero emissions.
huber HJ, Cramer W, Nakicenovic N, Wigley T, Yohe G (Cambridge Univ Press, Cam- Geophys Res Lett 35:L04705.
bridge, UK), pp 265280. 36. Rogner HH (1997) An assessment of world hydrocarbon resources. Annu Rev Energ Env
10. Meinshausen M, et al. (2006) Multi-gas emissions pathways to meet climate targets. 22:217262.
Clim Change 75:151194. 37. Salomon S, Qin D, Manning M (2007) Technical summary. In Climate Change 2007: The
11. Harvey LDD (2007) Allowable CO2 concentrations under the United Nations Frame- Physical Science Basis, Contribution of WG I to the Fourth Assessment Report of the
work Convention on Climate Change as a function of the climate sensitivity probability IPCC eds Solomon S, Qin D, Manning M (Cambridge Univ Press, Cambridge, UK).
distribution function. Environ Res Lett 2:014001. 38. Matthews HD, Gillett NP, Stott PA, Zickfeld K (2009) The proportionality of global
12. Petschel-Held G, Schellnhuber HJ, Bruckner T, Toth F, Hasselmann K (1999) The toler- warming to cumulative carbon emissions. Nature 459:829 832.
able windows approach: Theoretical and methodological foundations. Clim Change 39. Meinshausen M, et al. (2009) Greenhouse-gas emission targets for limiting global
41:303331. warming to 2 C. Nature 458:1158 1162.
13. Bruckner T, et al. (1999) Climate change decision support and the tolerable windows 40. ONeill BC, Oppenheimer M (2004) Climate change impacts are sensitive to the con-
approach. Env Mod Ass 4:217234. centration stabilization path. Proc Natl Acad Sci USA 101:1641116416.
14. Friedlingstein P, et al. (2006) Climate-carbon cycle feedback analysis: Results from the 41. Gent P, McWilliams J (1990) Isopycnal mixing in ocean general circulation models. J
C4MIP model intercomparison. J Clim 19:33373353. Phys Oceanogr 20:150 155.
15. Knutti R, Stocker T, Joos F, Plattner GK (2002) Constraints on radiative forcing and 42. Meissner KJ, Weaver AJ, Matthews HD, Cox PM (2003) The role of land surface dynamics
future climate change from observations and climate model ensembles. Nature in glacial inception: A study with the UVic Earth System model. Clim Dynam 21:515
416:719 723. 537.
16. Murphy JM, et al. (2004) Quantification of modelling uncertainties in a large ensemble 43. Schmittner A, Oschlies A, Giraud X, Eby M, Simmons HL (2005) A global model of the
of climate change simulations. Nature 430:768 772. marine ecosystem for long-term simulations: Sensitivity to ocean mixing, buoyancy
17. Piani C, Frame DJ, Stainforth DA, Allen MR (2005) Constraints on climate change from forcing, particle sinking, and dissolved organic matter cycling. Global Biogeochem Cy
a multi-thousand member ensemble of simulations. Geophys Res Lett 32:L23825. 19:GB3004.
18. Knutti R, Meehl GA, Allen MR, Stainforth DA (2006) Constraining climate sensitivity 44. Archer D (1996) A data-driven model of the global calcite lysocline. Global Biogeochem
from the seasonal cycle in surface temperature. J Clim 19:4224 4233. Cy 10:511526.
19. Gregory J, Stouffer R, Raper S, Stott P, Rayner N (2002) An observationally based 45. Weber SL, et al. (2007) The modern and glacial overturning circulation in the Atlantic
estimate of the climate sensitivity. J Climate 15:31173121. Ocean in PMIP coupled model simulations. Clim Past 3:51 64.
20. Forest C, Stone P, Sokolov A, Allen M, Webster M (2002) Quantifying uncertainties in 46. Gregory J, et al. (2005) A model intercomparison of changes in the Atlantic thermo-
climate system properties with the use of recent climate observations. Science 295:113 haline circulation in response to increasing atmospheric CO2 concentration. Geophys
117. Res Lett 32:L12703.
21. Frame DJ, et al. (2005) Constraining climate forecasts: The role of prior assumptions. 47. Stouffer RJ, et al. (2006) Investigating the causes of the response of the thermohaline
Geophys Res Lett 32:L09702. circulation to past and future climate changes. J Climate 19:13651387.
22. Forest CE, Stone PH, Sokolov AP (2006) Estimated PDFs of climate system properties 48. Matthews HD, Weaver AJ, Eby M, Meissner KJ (2003) Radiative forcing of climate by
including natural and anthropogenic forcings. Geophys Res Lett 33:L01705. historical land cover change. Geophys Res Lett 30:1055.
23. Hegerl GC, Crowley TJ, Hyde WT, Frame DJ (2006) Climate sensitivity constrained by 49. Jones PD, Parker DE, Osborn TJ, Briffa KR (2006) Global and hemispheric temperature
temperature reconstructions over the past seven centuries. Nature 440:1029 1032. anomaliesland and marine instrumental records. In Trends: A Compendium of Data
24. Annan JD, Hargreaves JC (2006) Using multiple observationally-based constraints to on Global Change (Carbon Dioxide Information Analysis Center, Oak Ridge National
estimate climate sensitivity. Geophys Res Lett 33:L06704. Laboratory, U.S. Department of Energy, Oak Ridge, TN).
25. von Deimling TS, Held H, Ganopolski A, Rahmstorf S (2006) Climate sensitivity esti- 50. Matthews HD, Caldeira K (2007) Transient climate-carbon simulations of planetary
mated from ensemble simulations of glacial climate. Clim Dynam 27:149 163. geoengineering. Proc Natl Acad Sci USA 104:9949 9954.

516
This is Exhibit "..1..." referred to in the

!l affidavit of Kirsten Zickfeld,
before me on the
of April, 2017.
A commissioner
British Columbia
517
LETTERS
PUBLISHED ONLINE: 6 MAY 2012 | DOI: 10.1038/NCLIMATE1496
Equivalence of greenhouse-gas emissions for peak

temperature limits
Stephen M. Smith1,2 *, Jason A. Lowe2 , Niel H. A. Bowerman3 , Laila K. Gohar2 , Chris Huntingford4
and Myles R. Allen3
Climate policies address emissions of many greenhouse and cumulative CO2 emissions. GTPs measure the change in global
gases including carbon dioxide, methane, nitrous oxide and temperature at a single point in time after emission, taken relative
various halogen-containing compounds. These are aggregated to that for the same mass of emitted CO2 (ref. 11). The chosen
and traded on a CO2 -equivalent basis using the 100-year point can either be at a constant time horizon after emission, or
global warming potential (GWP100 ); however, the GWP100 can be fixed at the time of overall peak temperature (1Tmax ), in
has received scientific and economic criticism as a tool for which case the time horizon progressively decreases and GTP values
policy14 . In particular, given international agreement to limit become dependent on timing of emission. Like the GWP, the GTP
global average warming to 2 C, the GWP100 does not measure is critically dependent on the choice of a timescale (either the
temperature and does not clearly signal the need to limit fixed time horizon or the timing of 1Tmax ) and does not account
cumulative CO2 emissions57 . Here, we show that future peak for warming beyond this horizon. It therefore leaves scope for
temperature is constrained by cumulative emissions of several perverse trade-offs between long-lived CO2 and more potent but
long-lived gases (including CO2 and N2 O) and emission rates shorter-lived emissions such as CH4 (ref. 20).
of a separate basket of shorter-lived species (including CH4 ). Indeed, CO2 perturbations are so long lived that temperature
For each basket we develop an emissions-equivalence metric will stabilize within centuries only if CO2 emissions are brought
allowing peak temperature to be estimated directly for any near to zero21 . Furthermore, recent CO2 -only studies have shown
emissions scenario. Todays emissions of shorter-lived species that cumulative emissions are an effective constraint on 1Tmax
have a lesser impact on ultimate peak temperature than those (refs 57), providing a conceptually simple way of framing CO2
nearer the time of peaking. The 2 C limit could therefore emissions policy. These studies do not address how non-CO2
be met by setting a limit to cumulative long-lived CO2 - emissions might be included in such a framework. Such a multigas
equivalent emissions while setting a maximum future rate for strategy would be economically advantageous22 , but it is likely
shorter-lived emissions. that incorporating other GHGs would have implications for the
Greenhouse gases (GHGs) and aerosols vary widely in their allowable CO2 budget23 .
atmospheric lifetime and radiative efficiency, which together Here, we present a multigas framework that recognizes both the
determine their effect on climate. To give a common currency 2 C limit and the role of cumulative CO2 emissions. To illustrate,
to these effects the Intergovernmental Panel on Climate Change we start with the simplest possible energy-balance equation for
(IPCC) has provided GWP values throughout its assessments. These the climate system (see Methods) with a response timescale ( ) of
are defined as the global mean radiative forcing integrated over a 31 yr. We assume no feedback of climate onto GHG concentrations.
given period following a one-kilogram pulse emission, relative to Figure 1 shows the normalized increase in global temperature
that from a one-kilogram pulse of CO2 (ref. 8). GWPs therefore predicted for sustained emissions of several well-mixed GHGs listed
enable a CO2 -equivalent value to be assigned to any given quantity in the IPCC fourth assessment report (AR4; ref. 8). Temperature
of GHG emission. Helped by its transparency and relative ease of asymptotes towards a maximum of x Ax Ex , where Ex is the
calculation, the GWP100 has developed from an indicative tool to emission rate for gas x (kg yr1 ), Ax is its radiative efficiency
the established metric for weighting emissions. (W m2 kg1 ), x is its atmospheric lifetime (yr) and is the climate
Alternative metrics for emissions equivalence exist in the litera- sensitivity parameter ( C (W m2 )1 ). The time taken to approach
ture based on radiative forcing9,10 , global temperature change1114 , this maximum is critically dependent on x because it determines
or economic costs of mitigation1517 . Each has potential merits, but how soon atmospheric concentrations reach equilibrium. Some
selecting the most useful for policy requires a clearly quantified goal species are near maximum within a few decades (for example CH4 ,
in limiting climate change18 . Dangerous limits have previously been perturbation lifetime 12 yr) whereas others cause temperature to
proposed in terms of quantities such as atmospheric concentrations, rise for each additional unit of emission even after several centuries
global temperature and rates of change19 . However, the Cancn (for example CF4 , lifetime 50,000 yr).
Agreements now provide an internationally recognized limit to We propose that emissions can be divided into two separate
warming of 2 C above pre-industrial levels, or possibly 1.5 C (http: baskets of long-lived and shorter-lived GHGs. Such a multibasket
//unfccc.int/resource/docs/2010/cop16/eng/07a01.pdf), suggesting approach has been suggested before24,25 , however we go further
a key role for metrics based on global temperature. by developing a method for determining which basket a given gas
Two methods for linking emissions to temperature have gained should belong to and by identifying metrics linking each basket to
recent attention: the global temperature-change potential (GTP) peak temperature. Our hypothesis is that gases satisfying x >
1 Committee on Climate Change, 7 Holbein Place, London SW1W 8NR, UK, 2 Met Office Hadley Centre, FitzRoy Road, Exeter, Devon EX1 3PB, UK,
3 Atmospheric,Oceanic and Planetary Physics, University of Oxford, Oxford OX1 3PU, UK, 4 Centre for Ecology and Hydrology, Wallingford, Oxon OX10
8BB, UK. *e-mail: stephen.smith@theccc.gsi.gov.uk.
NATURE CLIMATE CHANGE | ADVANCE ONLINE PUBLICATION | www.nature.com/natureclimatechange 1

2012 Macmillan Publishers Limited. All rights reserved.
518
LETTERS NATURE CLIMATE CHANGE DOI: 10.1038/NCLIMATE1496
1.0 1.00
HFC-32 (4.9 yr) HFC-32 (4.9 yr)
CH4 (12 yr) HFC-125 (29 yr)
0.8 CH4 (12 yr) HFC-143a
HFC-125 (29 yr) (52 yr)
Normalized T
0.6 HFC-227ea (34.2 yr) HFC-227ea

0.95 (34. 2 yr)
HFC-143a (52 yr)
r2 (Tmax,Emax)
0.4 N2O (114 yr)
N2O (114 yr)
HFE-125 (136 yr)
0.2
SF6 (3,200 yr)
HFE-125
CF4 (50,000 yr) 0.90
(136 yr)
0
0 100 200 300 400
Year after start of sustained emission
CO2
Figure 1 | Time evolution of global average temperature rise for several SF6 (3,200 yr)
0.85
GHGs under constant emission rates. Temperature change (1T) is shown 0.85 0.90 0.95 1.00
normalized to the eventual maximum reached for each gas. Gases shown r2 (Tmax, E)
span the range of atmospheric lifetimes (in brackets next to each gas)
found in IPCC AR4 Table 2.14 (ref. 8). CO2 is not included because Figure 2 | Correlation between peak temperature rise and cumulative
temperature does not asymptote under constant emissions using the emissions (r2 (1Tmax , 6E)) versus correlation between peak temperature
analytical expression for atmospheric decay cited by the IPCC. rise and maximum emission rate (r2 (1Tmax , Emax )) for key GHGs. Line of
equal correlations shown in grey.
are long-lived gases that, like CO2 , determine 1Tmax through
cumulative emissions; those with x < are shorter-lived agents rate, calculated from the simple planetary energy-balance equation
that contribute according to their sustained emission rate. (y Ay for gas y). We call this the absolute sustained emission
To test this we created more than 140 emission trajectories for temperature (aSET, C (kg yr1 )1 ). It is equivalent to the long-term
each GHG listed in IPCC AR4 Table 2.14 (ref. 8), excluding those absolute GTP value for sustained emission11 and the constant
controlled by the Montreal Protocol, using a range of growth rates, emission commitment used by the IPCC29 . We also note that
peak emission years and post-peak decline rates. We then estimated the concept has been discussed in earlier work, which showed
global temperature increase using a reduced-complexity coupled that the ratio of aSET values for two gases is equivalent to their
carbon cycleclimate model26 (see Methods). The model also ratio of absolute GWP values using an infinite time horizon
exhibits = 31 yr, as measured using the method in ref. 27. Figure 2 (aGWP )14 . Supplementary Table S3 lists aSET values for relevant
shows the correlation to 1Tmax of both cumulative emissions GHGs, along with uncertainty ranges owing to climate-system
and maximum emission rate for representative GHGs, including parameters, and their CH4 -equivalent (we use CH4 as an illustrative
all those with atmospheric lifetimes in the range 20200 yr. CH4 baseline, but note an inert shorter-lived gas such as HFC-32
and other gases with lifetimes less than 30 yr show a stronger could be used instead).
link between 1Tmax and emission rate, whereas N2 O, CO2 and SET values could also be derived for very short-lived forcing
other gases with lifetimes greater than 100 yr are more closely agents such as black carbon and other aerosols, although there
linked to 1Tmax through cumulative emissions. This result is are issues in applying any global average metric to such spatially
robust to uncertainty in , with only hydrofluorocarbon-143a inhomogeneous forcings3 . Our construction of the metric assumes
(HFC-143a; lifetime 52 yr) and HFC-227ea (lifetime 34.2 yr) fixed equilibrium climate sensitivity and does not include carbon-
showing intermediate behaviour (Supplementary Fig. S2). Together cycle feedbacks, although sustained emissions of shorter-lived
these two HFCs contribute no more than around 1% of total GHG gases will cause some further warming through the carbon cycle
radiative forcing at any time in future representative concentration (Supplementary Fig. S4).
pathways (RCPs)28 . We use the results in Fig. 2 as our basis for The two metrics together indicate that GHG emissions up to
distinguishing between long- and shorter-lived emissions. time t could lead to an eventual 1T M (t ), relative to a pre-industrial
For long-lived GHGs we define the absolute peak commitment t = 0 when all emissions are zero, given by:
temperature metric (aPCT, C kg1 ) as the increase in 1Tmax that XX t X
is committed to per unit emission. The aPCT is a generalization 1T M (t ) = Ex (t 0 )aPCTx + Ey (t )aSETy
of the cumulative warming commitment5 and the climatecarbon x t 0 =0 y
response6 concepts used in previous CO2 -only studies. We estimate where Ex (t ) and Ey (t ) are the emission rates at time t of long-lived
aPCT values for each long-lived GHG using plots of cumulative GHG x and shorter-lived GHG y, respectively. The peak value
emissions versus 1Tmax (Supplementary Fig. S3). Supplementary of 1T M (t ) provides a guide to the ultimate GHG-induced peak
Table S1 lists these values and CO2 -equivalent PCT values (calcu- temperature 1Tmax liable under a given emissions scenario. Figure 3
lated by dividing by the aPCT for CO2 ), together with indicative illustrates this by applying the mean aPCT values from Supple-
value ranges owing to uncertainty in climate-system parameters. mentary Tables S1 and S2 and the aSET values in Supplementary
We find an aPCT for CO2 of 4.3 1016 C kg1 (range 2.6 Table S3 to past and future emissions consistent with RCP2.6.
1016 8.01016 C kg1 ), or 1.6 C (1.02.9 C) per trillion tonnes Shaded regions show, for each year to 2100, predicted warming
of carbon, closely matching previous CO2 -only estimates5,6 . This re- from cumulative emissions of long-lived GHGs up to that year
sult remains the same to within one decimal place when we remove (Fig. 3a) and the additional warming that would arise if shorter-
all non-CO2 forcings throughout the simulation, indicating that the lived GHG emissions continued at the level in that year (Fig. 3b).
previous estimates may hold more generally for multigas scenarios. Results for 2005 indicate that long-lived GHG emissions have
aPCT values show little scenario dependence (Supplementary already contributed nearly 0.9 C, of which more than 0.7 C is due
Table S2). For all the GHGs, uncertainties in CO2 -equivalent PCT to CO2 and 0.1 C due to N2 O. This is committed warming, even if
values are smaller than uncertainties in aPCT values. all future emissions (long- and shorter-lived) were somehow halted.
For the shorter-lived GHGs we compare the peak temperature Emissions of shorter-lived GHGs at 2005 rates would eventually
rise that would occur if emissions were maintained at a unit lead to further warming of nearly 0.6 C, almost all as a result of
2 NATURE CLIMATE CHANGE | ADVANCE ONLINE PUBLICATION | www.nature.com/natureclimatechange

519
NATURE CLIMATE CHANGE DOI: 10.1038/NCLIMATE1496 LETTERS
a 2.0 0.30 Black carbon
Contribution to Tmax (C)

1.8
T M from long-lived GHGs
(C above pre-industrial)
SF6 0.25
1.6 Trop. O3 precursors
HFC-23
1.4
C6F14 0.20 CH4
1.2 C2F6
1.0 CF4 Shorter-lived F-gases
0.15
0.8 N2O
0.6 CO2 Long-lived F-gases
0.10
0.4 MAGICC, all long-
lived GHGs N2O
0.2 0.05
0 CO2
2000 2020 2040 2060 2080 2100
0
Year 2010s 2020s 2030s 2040s 2050s 2060s
b 0.7 Decade of emission
T M from short-lived GHGs
HFC-245fa
0.6
HFC-227ea Figure 4 | MAGICC-modelled contributions to 1Tmax in the RCP2.6
0.5 HFC-143a scenario from emissions of GHGs, black carbon and tropospheric ozone
0.4 HFC-134a (trop. O3 ) precursors in each decade. Trop. O3 precursors denote
HFC-125
0.3
emissions of carbon monoxide, oxides of nitrogen and volatile organic
HFC-43-10mee
HFC-32 compounds and includes the indirect effect on CH4 through OH
0.2
CH4 concentrations. F-gases denote halogen-containing compounds. CH4
0.1 MAGICC, all short- contribution includes indirect effects on trop. O3 and stratospheric water
lived GHGs
0 vapour; black-carbon contribution includes the effect on snow albedo. For
2000 2020 2040 2060 2080 2100 this model configuration and all anthropogenic forcings in the RCP2.6
Year
scenario, 1Tmax occurs in 2062.
c 2.0
1.8 Tests with a higher warming scenario show similarly good
1.6 Total SET

metric predictions for both long-lived GHG warming and the
1.4
Total PCT timing of overall GHG-induced 1Tmax (Supplementary Fig. S5).
Total T M
1.2
1.0 The SET-based overestimate of warming from shorter-lived GHGs
MAGICC, all
0.8 forcings
increases, however, which we attribute mainly to the slower pace of
0.6 MAGICC, all emissions reduction near 1Tmax .
0.4 GHGs Applying this framework to emissions policy would require a
0.2 departure from the single-basket approach of the Kyoto Protocol
0
2000 2020 2040 2060 2080 2100 in which all GHGs are traded together. Instead, long-lived GHGs
Year would be traded with other long-lived GHGs using the PCT metric
to provide emissions-equivalence values, while the SET metric
Figure 3 | PCT- and SET-estimated temperature change for GHGs in the would carry out the same role separately for the shorter-lived
RCP2.6 scenario (1T M ) compared with realized GHG warming simulated species. To satisfy a given limit to 1Tmax such as 2 C, long-term
by MAGICC (black lines). a, Component owing to long-lived GHG targets could be set for cumulative emissions of long-lived GHGs,
emissions. Shaded areas show cumulative emissions up to that year and for emission rates of shorter-lived GHGs by the time peak
multiplied by aPCT value for each gas. b, Component owing to shorter-lived temperature is reached. Although these would be separate targets,
GHG emissions. Shaded areas show emissions in that year multiplied by there is a trade-off to be made between the two: higher emission
aSET value for each gas. c, Sum of a and b. Metrics estimate total rates of shorter-lived GHGs will require lower cumulative emissions
GHG-induced 1Tmax of 1.96 C around 2080, direct modelling predicts of long-lived GHGs, and vice versa.
1.89 C around 2070. Dashed line shows MAGICC output for all forcings In the time leading up to 1Tmax the relevance of shorter-lived
(including aerosols and natural factors). emissions will increase. Figure 4 shows direct model outputs for
contributions to 1Tmax over time from both long- and shorter-lived
CH4 (the realized warming, shown by the black line, is only half as emissions in the RCP2.6 scenario. It was created by removing
large because emissions rates are steadily increasing up to this point, each emission species in each decade (and only that decade) and
hence atmospheric concentrations and global temperatures are a simulating the resulting change in 1Tmax . In this scenario CO2
long way from reaching equilibrium). As emission rates decline, emissions decline rapidly over the coming decades, hence the figure
committed warming from long-lived species continues to increase shows their total influence diminishing, however the effect on peak
whereas the additional warming from shorter-lived species declines. temperature per unit CO2 emission remains similar regardless of
Results for 2100 thus show an increased total of 1.6 C from PCT timing. Earlier CO2 emissions may in fact be more important,
gases and a reduced total of just over 0.3 C from SET gases. because the rate of additional warming declines as high cumulative
Total metric-estimated 1T M (t ) from both long- and shorter- emissions totals are reached (Supplementary Fig. S3). In contrast,
lived GHG emissions is shown in Fig. 3c. Direct comparison with todays emissions of CH4 and other short-lived forcings (such as
realized warming (black solid line) shows close agreement on both ozone and black carbon, included in the figure for completeness)
the magnitude and timing of 1Tmax (1.89 C around 2070 from have a lesser effect on the magnitude of the peak than those
the model, 1.96 C around 2080 from the metrics). By this time emissions nearer the time of peaking. This is even more evident for
emission rates have declined rapidly enough for the SET prediction scenarios that show a later peak (Supplementary Fig. S6).
to fall to within 0.1 C of the realized temperature from shorter- Accepting that sustained shorter-lived forcings have a secondary,
lived GHGs. A scenario of even more rapid emissions reductions long-term influence on 1Tmax through the carbon cycle and that
could cause the SET metric to underestimate realized temperature. details change between scenarios, the qualitative result remains the
The overall modelled temperature profile for RCP2.6 (grey dashed same: present emissions of long-lived GHGs such as CO2 have a
line) shows a lower and earlier 1Tmax of 1.70 C because it includes greater impact on ultimate 1Tmax . Immediate reductions in CH4 ,
other natural and anthropogenic forcings such as aerosols. black carbon and ozone precursors will help to slow near-term rates

520
LETTERS NATURE CLIMATE CHANGE DOI: 10.1038/NCLIMATE1496
of warming and bring other benefits30 . They may also indirectly help 6. Matthews, H. D., Gillett, N. P., Stott, P. A. & Zickfeld, K. The proportionality of
to reduce 1Tmax by stimulating technological and economic devel- global warming to cumulative carbon emissions. Nature 459, 829832 (2009).
7. Zickfeld, K., Eby, M., Matthews, H. D. & Weaver, A. J. Setting cumulative
opments that allow further reductions before peaking is reached.
emissions targets to reduce the risk of dangerous climate change. Proc. Natl
Their direct influence on peak temperature is, however, limited. Acad. Sci. USA 106, 1612916134 (2009).
We have presented two simple metrics (PCT and SET) that 8. Forster, P. et al. in IPCC Climate Change 2007: The Physical Science Basis
can be used for GHG emissions equivalence and allow rapid (eds Solomon, S. et al.) (Cambridge Univ. Press, 2007).
estimation of peak temperature for a given scenario. The time- 9. Lenton, T. 2 C or not 2 C? That is the climate question. Nature 473,
7 (2011).
horizon choice embedded in the GWP100 that is used at present 10. Wigley, T. M. L. The Kyoto Protocol: CO2 , CH4 and climate implications.
may obscure the trade-off available in meeting the 2 C limit Geophys. Res. Lett. 25, 22852288 (1998).
between cumulative long-lived GHG emissions and emission rates 11. Shine, K., Fuglestvedt, J., Hailemariam, K. & Stuber, N. Alternatives to the
of shorter-lived species. Our division of different radiative species global warming potential for comparing climate impacts of emissions of
based on their lifetimes allows this trade-off to be presented in greenhouse gases. Climatic Change 68, 281302 (2005).
12. Tanaka, K., ONeill, B., Rokityanskiy, D., Obersteiner, M. & Tol, R. Evaluating
simple terms. Compared with the time-dependent GTP, these global warming potentials with historical temperature. Climatic Change 96,
metrics provide less detail on the changing contribution to 1Tmax of 443466 (2009).
each emission over time, and the SET metric in particular provides 13. Johansson, D. Economics- and physical-based metrics for comparing
less accurate temperature predictions. But GTP values are sensitive greenhouse gases. Climatic Change 110, 123141 (2012).
14. Fisher, D. A., Hales, C. H., Wang, W-C., Ko, M. K. W. & Sze, N. D. Model
to uncertainty over when 1Tmax is predicted to occur20 , and calculations of the relative effects of CFCs and their replacements on global
although the two-basket framework still requires such a prediction warming. Nature 344, 513516 (1990).
to set a timescale for reductions in shorter-lived species, PCT and 15. Manne, A. S. & Richels, R. G. An alternative approach to establishing trade-offs
SET values themselves are robust to this uncertainty. Furthermore, among greenhouse gases. Nature 410, 675677 (2001).
they are still applicable in the decades following 1Tmax . 16. Kandlikar, M. Indices for comparing greenhouse gas emissions: Integrating
science and economics. Energ. Econ. 18, 265281 (1996).
Interesting work remains in applying this framework to the 17. Eckaus, R. S. Comparing the effects of greenhouse gas emissions on global
economics of emissions reduction and in considering possible warming. Energy J. 13, 2536 (1992).
mechanisms for trade-offs between the two types of emission 18. Plattner, G-K., Stocker, T., Midgley, P. & Tignor, M. IPCC Expert Meeting on
that reflect their changing importance for meeting tempera- the Science of Alternative Metrics (IPCC, 2009).
19. Oppenheimer, M. & Petsonk, A. Article 2 of the UNFCCC: Historical origins,
ture limits over time. recent interpretations. Climatic Change 73, 195226 (2005).
Methods 20. Shine, K. P., Berntsen, T. K., Fuglestvedt, J. S., Skeie, R. B. & Stuber, N.
Comparing the climate effect of emissions of short- and long-lived climate
Figure 1 and aSET values were calculated using the energy-balance
agents. Phil. Tran. R. Soc. A 365, 19031914 (2007).
equation in ref. 11:
21. Matthews, H. D. & Caldeira, K. Stabilizing climate requires near-zero emissions.
d1T (t )
= RF(t ) 1T (t ) Geophys. Res. Lett. 35, L04705 (2008).
dt 22. Van Vuuren, D. P., Weyant, J. & De la Chesnaye, F. Multi-gas scenarios to
where 1T (t ) is the global average surface temperature change at time t , RF (t ) stabilize radiative forcing. Energ. Econ. 28, 102120 (2006).
is the radiative forcing and is the response timescale equal to C (C being the 23. Cox, P. M. & Jeffery, H. A. Methane radiative forcing controls the allowable
effective heat capacity of the climate system). Consistent with the model for the CO2 emissions for climate stabilization. Curr. Opin. Environ. Sustain. 2,
assessment of greenhouse-gas induced climate change (MAGICC) settings below, 404408 (2010).
we used a climate sensitivity parameter = 0.8 C (Wm2 )1 ( here is the inverse 24. Jackson, S. C. Parallel pursuit of near-term and long-term climate mitigation.
of that used in ref. 27). Science 326, 526527 (2009).
All other calculations were carried out using MAGICC v5.3 (http://www. 25. Daniel, J. et al. Limitations of single-basket trading: Lessons from the Montreal
cgd.ucar.edu/cas/wigley/magicc/). Default carbon-cycle parameters were used Protocol for climate policy. Climatic Change 111, 241248 (2012).
with an equilibrium climate sensitivity of 3 C, 60 m ocean mixed-layer depth 26. Wigley, T. M. L. & Raper, S. C. B. Interpretation of high projections for
and 2.3 cm2 s1 ocean vertical diffusivity, yielding a transient climate response of global-mean warming. Science 293, 451454 (2001).
1.82 C. Atmospheric concentrations and fluxes of CO2 , CH4 , N2 O and halogenated 27. Andrews, D. G. & Allen, M. R. Diagnosis of climate models in terms of transient
GHGs up to the year 2000 were adjusted for consistency with the RCP scenarios. climate response and feedback response time . Atmos. Sci. Lett. 9, 712 (2008).
More than 140 emission trajectories were developed for each gas to cover the 28. Meinshausen, M. et al. The RCP greenhouse gas concentrations and their
period 20002400. Each started on the level provided in the RCP2.6-harmonized extensions from 1765 to 2300. Climatic Change 109, 213241 (2011).
emission profiles (http://www.pik-potsdam.de/mmalte/rcps/), from which we 29. Meehl, G. A. et al. in IPCC Climate Change 2007: The Physical Science Basis
projected four different multiples of emissions growth up to a ten-year peaking (eds Solomon, S. et al.) (Cambridge Univ. Press, 2007).
phase centred on 2010, 2030, 2050, 2070 or 2090. Growth rates were chosen 30. Shindell, D. et al. Simultaneously mitigating near-term climate change and
so that trajectories enclosed the range of peak emissions found across all RCPs. improving human health and food security. Science 335, 183189 (2012).
After the peaking phase, emissions decline at a rate of 1%, 2%, 3%, 4%, 5%, 7%
or 10%. We included an additional case of zero emissions after 2000, used as a
Acknowledgements
baseline against which to measure peak temperature and emissions. Each curve We thank K. Shine and D. Shindell for discussions during the progress of this research.
was combined with RCP2.6 harmonized emissions for all other species (including We are grateful to T. Wigley and the other model developers for making MAGICC freely
aerosol, ozone precursors and ozone-depleting GHGs) and then analysed using available. S.M.S. was supported by the UK Committee on Climate Change (the views
MAGICC (Supplementary Fig. S1). Trajectories that had not reached peak warming expressed herein represent those of the authors and do not necessarily represent the views
by the end of the model run, and CO2 runs exceeding 4 C, were filtered out (only of the committee). J.A.L. and L.K.G. were supported by the joint Department of Energy
perfluorocarbons and SF6 had fewer than 130 runs remaining, with 73 for C5 F12 ). and Climate Change/Department for Environment, Food and Rural Affairs Met Office
Hadley Centre Climate Programme (GA01101). N.H.A.B. was supported by a Natural
Received 18 November 2011; accepted 25 March 2012; Environment Research Council CASE studentship with the Met Office. M.R.A. received
published online 6 May 2012 additional support from the US National Oceanic and Atmospheric Administration
and the US Department of Energy through the International Detection and Attribution
References Group. C.H. was supported by the Centre for Ecology & Hydrology science budget fund.
1. ONeill, B. C. The jury is still out on global warming potentials. Climatic Change
44, 427443 (2000). Author contributions
2. Shine, K. The global warming potentialthe need for an interdisciplinary S.M.S. designed the research with input from J.A.L. Decadal emissions plots were devised
retrial. Climatic Change 96, 467472 (2009). by N.A.H.B., C.H. and M.R.A. S.M.S. carried out all simulations with advice on model
3. Fuglestvedt, J. S. et al. Transport impacts on atmosphere and climate: Metrics. setup from L.K.G. All authors contributed to writing the paper.
Atmos. Environ. 44, 46484677 (2010).
4. Gillett, N. P. & Matthews, H. D. Accounting for carbon cycle feedbacks in a Additional information
comparison of the global warming effects of greenhouse gases. Environ. Res. The authors declare no competing financial interests. Supplementary information
Lett. 5, 034011 (2010). accompanies this paper on www.nature.com/natureclimatechange. Reprints and
5. Allen, M. R. et al. Warming caused by cumulative carbon emissions towards permissions information is available online at www.nature.com/reprints. Correspondence
the trillionth tonne. Nature 458, 11631166 (2009). and requests for materials should be addressed to S.M.S.

521
/
This is Exhibit "!4." r'eferred to in the
!l affidavit of Kirsten Zickfeld,
afhrmed before me on the
day of Aoril, 2017.
A commissioner for'
Blitish Columbia
522
PERSPECTIVE
PUBLISHED ONLINE: 24 FEBRUARY 2016|DOI: 10.1038/NCLIMATE2868
Differences between carbon budget

estimates unravelled
Joeri Rogelj1,2*, Michiel Schaeffer3,4, Pierre Friedlingstein5, Nathan P. Gillett6, Detlef P. van Vuuren7,8,
Keywan Riahi1,9, Myles Allen10,11 and Reto Knutti2
Several methods exist to estimate the cumulative carbon emissions that would keep global warming to below a given temperature
limit. Here we review estimates reported by the IPCC and the recent literature, and discuss the reasons underlying their differ-
ences. The most scientifically robust number the carbon budget for CO2-induced warming only is also the least relevant
for real-world policy. Including all greenhouse gases and using methods based on scenarios that avoid instead of exceed a given
temperature limit results in lower carbon budgets. For a >66% chance of limiting warming below the internationally agreed
temperature limit of 2C relative to pre-industrial levels, the most appropriate carbon budget estimate is 5901,240GtCO2
from 2015 onwards. Variations within this range depend on the probability of staying below 2C and on end-of-century non-CO2
warming. Current CO2 emissions are about 40GtCO2yr1, and global CO2 emissions thus have to be reduced urgently to keep
within a 2C-compatible budget.
T
he ultimate objective of international climate negotiations is The near-linearity between peak global-mean temperature rise
to prevent dangerous anthropogenic interference with the cli- and cumulative CO2 emissions is the result of an incidental interplay
mate system1. At the 2015 Paris Conference, this objective was of several compensating feedback processes in both the carbon cycle
further specified as limiting global-mean temperature increase to and the climate: the logarithmic relationship between atmospheric
well below 2 C relative to pre-industrial levels and pursuing further CO2 concentrations and radiative forcing, the decline of ocean heat-
efforts for limiting temperature increase to below 1.5 C (ref. 2). uptake efficiency over time, as well as the changes in the airborne
Over the past decade, a large body of literature has been published fraction of anthropogenic CO2 emissions15. This compensating rela-
that shows that the maximum global-mean temperature increase as tionship is robust over a range of CO2 emissions and over timescales
a result of CO2 emissions is nearly linearly proportional to the total of up to a few centuries, with very few exceptions16. Such a relation-
cumulative carbon (CO2) emissions311. Maximum warming is also ship is not generally shown for other anthropogenic radiatively
influenced by the amount of non-CO2 forcing leading up to the time active species. An approximate proportionality exists for other long-
of the peak1214. This has culminated in the most recent assessment lived greenhouse gases (GHGs) for warming during this century 12,
of the IPCC in the form of several estimates of emission budgets whereas for short-lived climate forcers the rate of emissions leading
compatible with limiting warming to below specific temperature up to the time of peak warming is important 1214.
limits. Here, we first explain the underlying scientific rationale The unique characteristics of the Earth systems response to
for such budgets and then continue with a detailed account of the anthropogenic carbon emissions allow the definition of a quantity
strengths and limitations of the various budgets reported in both the called the transient climate response to cumulative emissions of
IPCC Fifth Assessment Report (AR5) and the recent literature, and carbon (TCRE). TCRE is defined as global average surface tempera-
of the differences between them. ture change per unit of total cumulative anthropogenic CO2 emis-
sions, typically 1,000PgC. In AR5, TCRE was assessed to be likely
The purpose of budgets to lie (that is, with greater than 66% probability 17) between 0.8 to
The IPCC AR5 Working Group I (WGI) report 15 indicated that the 2.5C per 1,000PgC for cumulative CO2 emissions less than about
total net cumulative emission of anthropogenic CO2 is the principal 2,000PgC and until the time at which temperature peaks.
driver of long-term warming since pre-industrial times. Therefore, The constancy of TCRE means that it can also be assessed for the
to limit the warming caused by CO2 emissions to below a given tem- real world by dividing an estimate of CO2-induced warming to date
perature threshold, cumulative CO2 emissions from all anthropo- by an estimate of anthropogenic CO2 emissions5,10. Such an approach
genic sources need to be capped to a specific amount, sometimes relies on a calculation of GHG-attributable warming using a regres-
referred to as the carbon budget or quota (which, in the context sion of observed warming onto the simulated response to GHGs and
of this Perspective, refers to global values and not to the emission other forcings, and an estimate of the ratio of CO2 to total GHG radi-
allowances of single countries). ative forcing or temperature response. Alternatively TCRE may be
1
ENE Program, International Institute for Applied Systems Analysis (IIASA) Schlossplatz 1, A-2361 Laxenburg, Austria. 2Institute for Atmospheric and
Climate Science, ETH Zurich, Universittstrasse 16, CH-8092 Zrich, Switzerland. 3Climate Analytics, Karl-Liebknechtstrasse 5, 10178 Berlin, Germany.
4
Environmental Systems Analysis Group, Wageningen University and Research Centre, PO Box47, 6700AA Wageningen, the Netherlands. 5College
of Engineering, Mathematics and Physical Sciences, University of Exeter, Exeter EX4 4QF, UK. 6Canadian Centre for Climate Modelling and Analysis,
Environment Canada, University of Victoria, PO Box1700, STN CSC, Victoria, British Columbia V8W 2Y2, Canada. 7PBL Netherlands Environmental
Assessment Agency, PO Box303, 3720AH Bilthoven, the Netherlands. 8Copernicus Institute of Sustainable Development, Faculty of Geosciences,
Utrecht University, Budapestlaan 4, 3584CD Utrecht, the Netherlands. 9Graz University of Technology, Inffeldgasse, A-8010 Graz, Austria..10ECI,
School of Geography and the Environment, University of Oxford, Oxford OX1 3QY, UK. 11Department of Physics, University of Oxford, Parks Road,
Oxford OX1 3PU, UK. *e-mail: rogelj@iiasa.ac.at
NATURE CLIMATE CHANGE | VOL 6 | MARCH 2016 | www.nature.com/natureclimatechange 245

523
PERSPECTIVE NATURE CLIMATE CHANGE DOI: 10.1038/NCLIMATE2868
a 12 Budget for CO2-induced warming only. The most direct application
of TCRE is to derive cumulative carbon budgets consistent with lim-
combustion and industry (PgC yr1) 10 iting CO2-induced warming to below a specific temperature thresh-
Carbon emissions from fossil fuel
old. For instance, WGI indicates26 that limiting anthropogenic

8 CO2-induced warming to below 2 C relative to 18611880 with
an assessed probability of greater than 50% will require cumulative
6 CO2 emissions from all anthropogenic sources since that period to
stay approximately below 4,440GtCO2. Alternatively, doing so with
4
a greater than 66% probability would imply a 3,670GtCO2 budget.
2
These values assume a normal distribution of which the standard
deviation (1) range is given by the assessed likely TCRE range of
0 0.8to 2.5C per 1,000PgC (that is, about 3,670GtCO2), and make
use of the near-linearity of the ratio of CO2-induced warming to
cumulative CO2 emissions15.
2000 2020 2040 2060 2080 2100 Although this is the most robust translation of the TCRE concept
Year into a cumulative carbon budget, it is at the same time also the least
b directly useful to policy-making. In the real world, non-CO2 forcing
2.2
also plays a role, and its global-mean temperature effect is superim-
2 posed on the CO2-induced warming. A carbon budget derived from
a TCRE-based estimate should thus not be used in isolation.
increase relative to 18501875 (C)
1.8 The near-linear relationship of TCRE does hence not necessarily

Global mean temperature
1.6
apply to the ratio of total human-induced warming to cumulative
carbon emissions (as might be suggested by Fig. SPM.10in ref.26).
1.4 The latter relationship is scenario dependent, because, for example,
the percentage contribution of non-CO2 climate drivers to total
1.2
anthropogenic warming increases in the future in many scenarios.
1 Therefore, to take into account the influence of non-CO2 forcing on
carbon budgets, the TCRE-based approach can be extended using
0.8 multi-gas emission scenarios. Multi-gas emission scenarios pro-
vide an internally consistent evolution over time of all radiatively
2000 2020 2040 2060 2080 2100 active species of anthropogenic origin. They are often created with
Year integrated assessment models (IAMs), which represent interactions
within the global energyeconomyland system (for examples, see
Figure 1 | Proportionality of global-mean temperature increase to refs2729).
cumulative emissions of CO2. a,b, Four CO2 emission pathways with
identical cumulative carbon emissions over the twenty-first century (a) and Threshold exceedance budgets. Here we define a straightforward
their corresponding temperature projections (b). The grey area in b shows methodology of extending TCRE-based carbon budgets for CO2-
the central 66% uncertainty range of temperature projections around the induced warming to budgets that also takes into account non-CO2
thick purple line. Figure adapted with permission from ref.15, 2013 IPCC. warming as threshold exceedance budgets (TEBs) for multi-gas
warming (Table1).
assessed from observations by applying observational constraints to This approach uses multiple realizations of the simulated response
the parameters of a simple carbon-cycle climate model7,8, and evalu- to a multi-gas emission scenario. These realizations can either be
ating the ratio of warming to emissions for the constrained model. multi-model ensembles or perturbed parameter ensembles. An
For a carbon budget approach to make sense, TCRE must be example of the former would be simulations of the Representative
reasonably independent of the pathway of emissions. Earlier studies Concentration Pathways30,31 (RCPs) by Earth system models (ESMs)
have indeed shown that this is the case7,8,18,19, at least for peak warm- that were contributed to the Fifth Phase of the Coupled Model
ing and monotonously increasing cumulative carbon emissions. If a Intercomparison Project 32 (CMIP5). An example of the latter would
set carbon budget limit is exceeded, CO2 needs to be actively removed be the use of a simple climate model in a probabilistic setup7,23,24, as
from the atmosphere afterwards2022 to bring emissions back to used in the assessments of the IPCC3335 as well as in other recent
within the budget. Figure1 illustrates this path independency (even studies3638. From such multi-model or perturbed parameter ensem-
for moderate amounts of net negative CO2 emissions), and shows bles, the carbon budget is estimated at the time a specified share
with the simple carbon cycle and climate model MAGICC7,23,24 that (for example, 50% or one-third) of realizations exceeds a given tem-
even with large variations in the pathway of CO2 emissions during perature limit (that is, 50% or two-thirds of the ensemble members
the twenty-first century, the transient temperature paths as a func- remain below the limit; see orange scenario in Fig.2).
tion of cumulative CO2 emissions are very similar a characteristic The TEB approach was used by WGI for determining carbon
also found in other models18,25. Once all pathways achieve the same budgets that account for non-CO2 forcing 15. Applying this meth-
end-of-century cumulative CO2 emissions, the temperature projec- odology to the CMIP5 RCP8.5 (ref. 39) simulations of ESMs10,40
tions are virtually identical (Fig.1). and ESMs of intermediate complexity 41 (EMICs), they found that
Given these considerations, carbon budgets are a useful guide for compatible CO2 emissions since 1870 are about 3,010 GtCO2 and
defining and characterizing the emissions pathways that limit warm- 2,900 GtCO2 to limit warming to less than 2 C since the period
ing to certain levels, such as 2C relative to pre-industrial levels. 18611880 in more than 50% and 66% of the available model runs,
respectively. Other recent studies36 have used an extended version
An abundance of carbon budgets of this approach that computes TEBs based on perturbed param-
Despite the simplicity of carbon budgets, many (often very different) eter ensembles of a subset of scenarios from the IPCC AR5 Scenario
estimates have been published. Here we provide an overview of how Database (hosted at the International Institute for Applied Systems
these budgets are defined and calculated. Analysis (IIASA); https://secure.iiasa.ac.at/web-apps/ene/AR5DB).
246 NATURE CLIMATE CHANGE | VOL 6 | MARCH 2016 | www.nature.com/natureclimatechange

524
NATURE CLIMATE CHANGE DOI: 10.1038/NCLIMATE2868 PERSPECTIVE
Table 1 | Three different types of carbon budgets and their definitions.
Carbon budget type Abbreviation Definition and description
Budget for CO2-induced CO2-only budget Amount of cumulative carbon emissions that are compatible with limiting warming to below a specific
warming temperature threshold with a given probability in the hypothetical case that CO2 is the only source of
anthropogenic radiative forcing. This budget can be inferred from the assessed range of TCRE.
Threshold exceedance TEB Amount of cumulative carbon emissions at the time a specific temperature threshold is exceeded with a given
budget probability in a particular multi-gas emission scenarios. This budget thus takes into account the impact of
non-CO2 warming at the time of exceeding the threshold of interest.
Threshold TAB Amount of cumulative carbon emissions over a given time period of a multi-gas emission scenario that limits
avoidance budget global-mean temperature increase to below a specific threshold with a given probability. This budget thus takes
into account the impact of non-CO2 warming at peak global-mean warming, which is approximately the time
when global CO2 emissions become zero and global-mean temperature is stabilized.
The results of a TEB approach are most useful if the warming for limiting warming to below 2 C with at least 66% chance of
due to non-CO2 forcing as a function of cumulative CO2 emissions 2,5503,150GtCO2 from 1870 until peak warming.
is similar across scenarios, meaning that the conclusions are not
strongly dependent on the scenario chosen. However, Fig.3a shows The numbers compared
that there is quite a large variation in non-CO2 forcing for a given To understand what the different approaches mean in terms of the
level of cumulative CO2 emissions when looking at all scenarios actual values of carbon budgets, we compare the available budg-
available in the IPCC AR5 Scenario Database. Caution is therefore ets relating to the 2C limit. Table2 provides an overview for all
advised when deriving carbon budgets on the basis of one single of the numbers discussed in this section, relative to two com-
multi-gas scenario (see below). Finally, the use of TEBs for limiting mon base years (2011 and 2015). Taking into account that about
warming to below a given temperature limit assumes that non-CO2 2,050GtCO2 (approximately 560PgC) had already been emitted by
warming never increases beyond the level it reached at the time the the end of 201436, a CO2-only budget approach would indicate that
TEB was computed (Fig.2). Non-CO2 forcing thus needs to be kept 1,620 GtCO2 (or 440 PgC) remain to have a >66% probability of
within limits over time. limiting warming to below 2C relative to pre-industrial levels (here
defined as the 18611880 period26). Using a TEB approach and
Threshold avoidance budgets. Carbon budgets defined in the pre- assuming non-CO2 forcing as in RCP8.5, this amount is reduced to
vious section are derived at the time a given scenario exceeds a spe- 850GtCO2 (or 230PgC). When assessed with the latter approach, a
cific temperature threshold or limit. A complementary approach is 1,620GtCO2 budget would limit warming to below 2C in less than
to consider multiple emission scenarios and evaluate carbon budg- 33% of the available models42.
ets for the subset of scenarios that avoids crossing such a threshold It is worth noting that the IPCC assessment of the CO2-only
with a given probability. We name these budgets threshold avoid- budget is based on an assessed uncertainty range of TCRE, drawing
ance budgets (TABs, Table1). Because, by definition, such scenarios on many lines of evidence. The WGI numbers including non-CO2
do not exceed the limit of interest at any specific point in time (with forcing are based on CMIP5 simulations of the response to RCPs,
a given probability), a time horizon needs to be defined up to which which although being a valid approach provide a narrower
a budget is computed. This time horizon can either be a predefined scientific basis. At least for the four RCPs used by WGI, a similar
period, for example the 20112050 or the 20112100 period, or warming as a function of cumulative CO2 emissions is found (see
more variable in nature, for example the time period until peak Fig. TFE.8in ref.42), despite having different non-CO2 evolutions
warming (see yellow scenario in Fig.2). Both of these approaches (Fig.3a). This counterintuitive result is explained further below.
were used in AR5, and more sophisticated approaches based on the When extensively varying the non-CO2 assumptions for TEBs
TAB methodology have been used in the literature7. using a subset of baseline and weak mitigation scenarios from the
IPCC Working Group III (WGIII) computed TABs for the IPCC AR5 Scenario Database (which all exceed the 2 C limit), a
periods 20112050 and 20112100 by assessing probabilistic range of 8501,550GtCO2 (5th95th percentile range across all TEB
temperature projections in 210034,35. For this, WGIII categorized scenarios, from 2015 onwards) is associated with limiting warming
a large number of scenarios on the basis of end-of-century CO2- to below 2C with 66% probability 36. The difference between this
equivalent concentrations. The reported TAB values for example, range and the 850GtCO2 number quoted above is, on the one hand,
in Table6.3 in the WGIII Report 35 or Table SPM.1 in the Synthesis caused by the different modelling frameworks and, on the other
Report 33,34 (SYR) are therefore the result of an assessment of hand, by the fact that the non-CO2 forcing evolution of RCP8.5is
the exceedance probability outcomes found in each of the CO2- situated amongst the highest percentiles of the non-CO2 forcing in
equivalent concentration categories. Alternatively, scenarios could other high-emission scenarios that exceed the 2C threshold (Fig.3).
have been categorized on the basis of median temperature, prob- When considering TABs until peak warming, based on the strin-
abilities to limit warming to below a specific temperature limit, gent mitigation scenarios of the IPCC AR5 Scenario Database, a
or even carbon budgets. For scenarios that limit end-of-century range of 5901,240GtCO2 is found for limiting warming to below
warming to below 2C with a likely probability, the WGIII assess- 2 C with >66% probability 33 (10th90th percentile range, as
ment 34 reports that the TABs in terms of cumulative CO2 emissions reported by WGIII, from 2015 onwards). Finally, for TABs calcu-
in the periods 20112050and 20112100are 1501,300GtCO2 and lated over the 20152100 period, an assessment of the stringent mit-
6301,180GtCO2, respectively. igation scenarios available in the IPCC AR5 Scenario Database and
In the IPCC SYR33, TABs are also computed on the basis of their temperature outcomes results in a range of 4701,020GtCO2
the scenarios available in the IPCC AR5 Scenario Database see (10th90th percentile range) for limiting warming to below 2 C
Table2.2in ref.33. However, the SYR categorizes scenarios directly with a likely chance35.
based on their probability of keeping peak warming to below a In conclusion, moving from a CO2-only budget 42 to a multi-gas
specific temperature threshold (1.5 C, 2 C or 3 C) during the multi-scenario TEB budget 36 removes around 420 GtCO2 (that is,
twenty-first century. For example, the IPCC SYR reports TABs the average of the 70770GtCO2 range) from the CO2 budget from

525
CO2 emissions
30
Global carbon emissions 3
Different carbon budgets
20
(PgC yr1)
3 Threshold exceedance budget (TEB)

10 2
1
0
Temperature increase relative

to pre-industrial levels (C)
2
Threshold
1950 2000 2050 2100 2 Threshold avoidance budget (TAB)
(until peak warming)
Varying non-CO2 emissions
1
CH4 emissions
1,000
Global methane emissions
800 1
0
(TgCH4 yr1)
600 1950 2000 2050 2100
400 Timing of 2
peak warming
200 Timing of
exceeding threshold
0
1950 2000 2050 2100
Figure 2 | An illustration of the methods for computing TEBs versus TABs. In the first step (arrows labelled 1), temperature outcomes are computed
from multi-gas emission scenarios that either exceed (orange) or avoid (yellow) a given temperature threshold. Based on either the timing of exceeding
the chosen threshold or the timing of peak warming, carbon budgets compatible with the chosen temperature threshold are computed in the second step
(arrow labelled 2) by summing the carbon emissions of the underlying scenarios until the timing of exceeding the threshold or peak warming for the TEB
or TAB (arrow labelled 3), respectively.
2015 onward for limiting warming to below 2C with 66% chance. the 20152100 period would be of the order of 120220 GtCO2.
Subsequently moving to a TAB budget until peak warming 33 or Furthermore, the budget type definition also influences other fac-
over the 20152100 period35 and a >66% chance would also remove tors, such as scenario selection, whose impact on the carbon budget
about 260310GtCO2 and 380530GtCO2, respectively. (Note that is explained in more detail below.
these values are illustrative as they are obtained by comparing ranges
that are defined in different ways.) Underlying data and modelling. Some of the differences between
The TAB range for limiting warming to below 2C with greater the quantitative budgets estimates are simply driven by differences in
than 66% probability of 4701,020GtCO2 for the 20152100 period the underlying data and models. In general, these differences apply
is thus 35 to 70% below what would have been inferred from a CO2- to TEBs and TABs alike. For example, although the WGI CO2-only
only budget with a TEB approach. budget is based on the interpretation of an assessed uncertainty
range, the other TEB and TAB budgets were computed either from
Strengths and limitations CMIP5 RCP results (in the WGI report and the SYR) or from a sim-
The various approaches to computing carbon budgets each come ple climate model (MAGICC) in a probabilistic setup7,23,24 (in the
with their respective strengths and limitations. Understanding what WGIII report and the SYR).
can lead to possible differences in budget estimates is critical to Budget estimates can differ depending on whether a single-sce-
avoid misinterpretation of the numbers. nario multi-model ensemble is used (for example, all CMIP5 runs
The budget type definition, the underlying data and modelling, for RCP8.5) or alternatively a single-model multi-scenario perturbed
the scenario selection, temperature response timescales and the parameter ensemble is used (for example, the IPCC AR5 WGIII
accompanying pathway of CO2 and non-CO2 emissions are identi- approach, which uses MAGICC). The former approach allows us to
fied as possible key drivers of the difference between the various use information from a wide range of the most sophisticated mod-
budget options discussed above. els and incorporate state-of-the-art Earth system interactions in the
That the budget type definition will have an influence on the budget assessment. However, this approach comes at a high compu-
resulting numbers is almost trivial. For example, when defining tational cost, resulting in only a limited ensemble of opportunity of
TABs from 2011 to 2100 instead of until peak warming, the cumu- model runs being available for any assessment. The latter method,
lated net negative emissions that can be achieved until the end of on the other hand, uses a much simpler model, and hence comes
the century will lead to consistently lower 20152100 TABs com- with great computational efficiency, which allows for hundreds if not
pared with TABs defined until peak warming levels. Negative emis- thousands of realizations per scenario. Thus variations in scenario
sions occur when CO2 is actively removed from, instead of emitted assumptions on the pathways and evolution of non-CO2 forcing over
into, the atmosphere by human activities. For instance, for TABs time can be explored in more detail.
compatible with limiting warming to below 2C with >66% chance, These differences in the underlying data and modelling can result
the difference between TABs defined until peak warming and over in changes in the budget estimates. However, although a simple climate

526
a 1.6 b
RCP8.5
100 100
1.4 2575% range
frequency over subset (%)

Non-CO2 radiative forcing (W m2)
80 595% range Median 80

1.2
Count per subset
Cumulative relative
60 60
1
RCP6 40 RCP2.6 40
0.8 RCP4.5
RCP4.5 RCP6
0.6 20 RCP8.5 20
RCP2.6
0.4 0 0
0.2 0.4 0.6 0.8 1 1.2
0.2 Non-CO2 forcing at time of deriving carbon budget (W m2)
0
c
0.2
0 1,000 2,000 3,000 4,000 100 100
153
Cumulative CO2 emissions from 2015 onwards (Gt)
frequency over subset (%)

80 80
Count per subset
Cumulative relative
d
(0.16, 1,890) and (0.05, 2,476) 60 60
Carbon budget (GtCO2)
1,800
40 40
1,400
20 20
1,000 0.70
600 0 0
0.53 0.2 0.3 0.4 0.5 0.6 0.7 0.8
200 Estimated temperature contribution from
0.2 0.3 0.4 0.5 0.6 0.7 0.8 non-CO2 forcing at time of deriving carbon budget (C)
Estimated temperature contribution from
non-CO2 forcing at time of deriving carbon budget (C) TEB TAB
Figure 3 | Non-CO2 forcing and cumulative CO2 emissions. a, Non-CO2 forcing as a function of cumulative CO2 emissions from 2015 onwards for scenarios
of the IPCC AR5 Scenario Database. Scenarios are split up into two subsets: those that limit warming to below 2C relative to pre-industrial levels with
at least 66% probability (yellow, used for TABs), and those that lead to global-mean temperatures exceeding the 2C limit with at least 34% (orange,
used for TEBs). b, Distribution of non-CO2 forcing at the time point critical for deriving TEB (orange) and TAB (yellow) budgets, that is, the moment the
2C limit is exceeded for TEBs and peak warming for TABs. c, Distribution of the estimated temperature contribution from non-CO2 forcing at the same
time point as in b (see Box1). The four RCPs are also included for comparison. d, Variation within the TEB and TAB budget subsets as a function of the
estimated temperature contribution from non-CO2 forcing as in c. Numerical values in d are R2 values for the two linear fits.
model does not provide the detail of ESMs, it can closely emulate Scenario selection. Applying the definitions of TEBs and TABs to
their global-mean behaviour 43 and can represent the uncertainties a large scenario ensemble for the assessment of CO2 budgets in line
in carbon-cycle and climate response in line with the assessment of with a particular temperature limit results in the selection of two
AR57,24,44. Of importance here is that the MAGICC setup applied in disjoint subsets of emission scenarios: a subset of baseline and weak
WGIII and the SYR is consistent with the CMIP5 ensemble for tem- mitigation scenarios that exceed the temperature limit with a given
perature projections and TCRE (Fig.12.8in ref.15 and Fig.6.12in probability in the case of TEB budgets and a disjoint subset of more
ref.35). It is therefore expected that these differences are limited. stringent to very stringent mitigation scenarios that all keep warm-
A final aspect related to the data and modelling is the interpreta- ing to below the specified temperature limit with a given probability
tion of the nature of the uncertainties that accompany the various in the case of TAB budgets.
data. Uncertainty ranges can be the expression of a variety of under-
lying uncertainty sources45, and they can be interpreted in different
ways. In the context of the quantification of carbon budgets, at least Box 1 | Non-CO2 temperature contributions.
three kinds of uncertainty ranges can be distinguished: an uncertainty
range resulting from an in-depth assessment of multiple lines of evi- The estimated temperature contributions of non-CO2 forcing,
dence (a so-called assessed uncertainty range); an uncertainty range shown in Fig. 3c, are derived by the following equation, as
emerging from a sophisticated statistical sampling of the parameter described in the Supplementary Material to the WGI chap-
space; or an uncertainty range that represents the spread across an ter on Anthropogenic and Natural Radiative Forcing53
arbitrary collection of model results (a so-called ensemble of oppor- (equation 8.SM.13).
tunity). Each of these uncertainty ranges can be interpreted in dif-
ferent ways, and they decline in robustness going from an assessed RT (t) =
M cj
j=1 dj
exp
t
dj ) )
uncertainty range over targeted statistical approaches to ensembles of
opportunities. These aspects thus also influence the robustness of any where RT is the climate response to a unit of forcing, cj the
carbon budget estimates based on them. For example, the budget for component of the climate sensitivity, dj the response times, and
CO2-induced warming from WGI is derived from an assessed uncer- t the time. For the two-term approximation (M=2) presented by
tainty range, whereas the WGI budgets that also take into account ref.54, values of c1, c2, d1, and d2 are taken from table8.SM.9in
non-CO2 forcing are based on an ensemble of opportunity, which ref.53. This estimate is to be considered an illustrative approxi-
makes them much less robust (see also Technical Focus Element 8in mation of the temperature effect of non-CO2 forcing.
ref.42).

527
Table 2 | A selection of carbon emission budgets related to a global temperature limit of 2C relative to pre-industrial levels from
various sources.
Source Type Specification Value from 2011 Value from 2015
(GtCO2) (GtCO2)
IPCC AR5 WGI26 CO2-only To limit warming to less than 2C since the period 18611880 with greater than 66% 1,780 (or 2,550) 1,620 (or 2,390)
budget (or 50%) probability
IPCC AR5 WGI26 TEB To limit warming to less than 2C since the period 18611880 in more than 66% (or 1,010 (or 1,120) 850 (or 960)
50%) of the model runs when accounting for the non-CO2 forcing as in the RCP8.5
scenario
IPCC AR5 WGIII35 TAB To limit warming in 2100 to below 2C since 18501900 with a likely (>66%) 630 to 1,180 470 to 1,020
probability, accounting for the non-CO2 forcing as spanned by the subset of stringent
mitigation scenarios in the IPCC AR5 Scenario Database*. (1090% range over
scenarios in IPCC WGIII scenario category 1)
IPCC AR5 WGIII35 TAB To limit warming in 2100 to less than 2C since 18501900 with a more likely than 960 to 1,430 800 to 1,270
not (>50%) probability, accounting for the non-CO2 forcing as spanned by the subset
of stringent mitigation scenarios in the IPCC AR5 Scenario Database*. (1090%
range over scenarios in IPCC AR5 scenario category II without overshoot)
IPCC AR5 SYR33 TEB To limit warming to less than 2C since the period 18611880 in more than 66% (or 1,010 (1,110 or 850 (960 or
50% or 33%) of the model runs of the CMIP5 RCP8.5 ESM and EMIC simulations. 1,410) 1,250)
(These correspond to the IPCC AR5 WGI TEB budgets reported above)
IPCC AR5 SYR33 TAB To limit warming to below 2C since 18611880 with 66100% probability, 750 to 1,400 590 to 1,240
accounting for the non-CO2 forcing as spanned by the subset of stringent mitigation
scenarios in the IPCC AR5 Scenario Database. (1090% range)
IPCC AR5 SYR33 TAB To limit warming to below 2C since 18611880 with 5066% probability, accounting 1,150 to 1,400 990 to 1,240
for the non-CO2 forcing as spanned by the subset of stringent mitigation scenarios in
the IPCC AR5 Scenario Database. (1090% range)
Friedlingstein et al.36 TEB To limit warming to less than 2C since 18501900 with a 66% probability, 1,310 (1,010 to 1,150 (850 to
accounting for the non-CO2 forcing as spanned by the subset of baseline and weak 1,710) 1,550)
mitigation scenarios in the IPCC AR5 Scenario Database*. (595% range)
Friedlingstein et al.36 TEB To limit warming to less than 2C since 18501900 with a 50% probability, 1,610 (1,210 to 1,450 (1,050 to
accounting for the non-CO2 forcing as spanned by the subset of baseline and weak 2,010) 1,850)
mitigation scenarios in the IPCC AR5 Scenario Database*. (595% range)
1,890GtCO2 were already emitted by 2011, and about 2,050GtCO2 by 2015. All values are rounded to the nearest 10. Budget types are defined in Table 1. *The temperature difference between 18611880 and
18501900 is 0.02C, based on ref. 55.
A first implication of the use of these disjoint scenario sets results Recent research indicates that, at current emission rates,
from only very few scenarios being available that have, for example, maximum CO2-induced warming only occurs about a decade after
precisely a 66% probability for limiting warming to below a given a CO2 emission46,47. Thus, even in a CO2-only world, TABs and TEBs
temperature threshold. Although TEBs are consistently computed for with complementary probabilities (for example, a 66% probability
each scenario at the time a scenario exceeds a temperature limit with to limit warming below 2 C and a 34% probability of exceeding
a given probability, the value of TABs is further driven by the choice 2C) would not be entirely identical. In case of the TEB approach,
of the range of probabilities that is used to select appropriate TAB the maximum warming of the CO2 emissions of the last decade
scenarios. For example, the IPCC SYR selected all scenarios that have before the temperature limit was exceeded has possibly not yet
a 66 to 100% probability of limiting warming to below a given thresh- fully occurred. In a TAB approach the emissions in the last dec-
old (compared with exactly 66% for TEBs). This resulted in an average ade would be significantly lower, if not zero, and this would allow
probability of staying below 2C across the subset of TAB scenarios a much larger fraction of the warming to already be realized. The
that comply with the above-mentioned selection criterion of about TEB approach thus leads to a consistent overestimate of the CO2
75%. This can explain about one-third to half of the 260310GtCO2 budget compatible with a given temperature limit, whereas this
difference between the TEB estimates from Friedlingstein et al.36 is not the case with the TAB approach. At least one-third of the
and the IPCC SYR TAB estimates. Moreover, for some temperature approximately 260310 GtCO2 difference between the TEB esti-
levels, for example around 3C, the scenarios available in the IPCC mates from Friedlingstein etal.36 and the IPCC SYR TAB estimates
AR5 Scenario Database do not sample the possible range extensively, can be explained by accounting for the approximately one decade
which can lead to further biases in the numbers obtained. delay between CO2 emissions and their maximum warming.
Temperature response timescales. A second aspect that is different Non-CO2 warming contribution. A third and last aspect that dif-
in the disjoint scenario subsets are the CO2 emission pathways and fers between the two disjoint TEB and TAB scenario subsets is the
hence the annual CO2 emissions at the time the compatible carbon mixture of CO2 and non-CO2 forcers. This mixture differs over time
budget is derived. In the TAB subset, CO2 emissions will typically and therefore, depending on when the compatible carbon budget
approach zero or become negative to stabilize global temperatures, is determined, the TABs and TEBs are derived under possibly very
and will thus be very low at the time of maximum warming dur- different non-CO2 forcing (Fig.3b). The relationship between CO2
ing the twenty-first century. In the TEB subset this is not the case. emissions and non-CO2 forcing is complex, as it covers the total
Because of the timescales of CO2-induced warming 46,47 this leads to non-CO2 forcing that results from both positive and negative cli-
differences in the carbon budget estimates. mate forcers. Climate policy influences these non-CO2 forcers

528
both directly (via abatement measures) and indirectly (via changes allowable carbon budget, that is, a very high risk of exceeding a given
induced in the energy system), and this is captured in different ways climate target when emitting that particular carbon budget.
in IAMs. For example, stabilizing and peaking global temperatures The strength of TEBs is that they are easily comparable to TCRE-
requires global CO2 emissions to be reduced to close to net zero. based budgets for CO2-induced warming only. Hence the influence
Such very low CO2 emissions are achieved through a fundamental of non-CO2 forcing on the size of carbon budgets can be assessed.
transformation of the global energyeconomyland system35, which However, because of the limitations related to scenario selection
in turn leads to changes in non-CO2 emissions because of the phase- (TEBs are derived from scenarios that fail in limiting warming to
out of common sources of CO2 and non-CO2 emissions14,48. This the temperature level of interest) and the timescales of the tempera-
can lead to important differences in non-CO2 forcing as a function ture response, TABs are preferred over TEBs. The strength of TABs
of total cumulative CO2 emissions (Fig.3a). Figure3b shows that lies exactly in their use of scenarios that represent our best under-
median non-CO2 forcing at the time that is of importance for deriv- standing of how CO2 and other radiatively active species would
ing the carbon budget (that is, the time of exceedance for TEBs, and evolve over time when CO2 emissions are stringently reduced.
peak warming for TABs) is about 0.2Wm2 higher in the subset of
scenarios used for TEBs compared with the subset used for TABs. Conclusions
However, the non-CO2 forcing at either peak warming or the Several possibilities are available to compute cumulative carbon
time of exceeding a given temperature threshold does not tell the budgets consistent with a particular temperature limit. We have
entire story. When estimating the actual non-CO2-induced warm- shown that each of the carbon budget approaches has strengths but
ing at these time points of interest (see Box1), very little difference also comes with important limitations. The devil is in the detail here.
can be found between the TEB and TAB scenario subsets (Fig.3c). The most scientifically robust number the budget for CO2-induced
This suggests that when a sufficiently large scenario sample is avail- warming is also the least practical in the real world. Selecting
able, variations in non-CO2 forcing cannot be used to explain the budgets based on multi-gas emission scenarios that actually restrict
variations between TEB and TAB estimates for limiting warming warming to below a given temperature threshold, results in the low-
to below 2C. The precise influence of this difference on the carbon est, but most relevant CO2 emission budgets in a real-world multi-gas
budgets has not been quantified. setting. Any practical implementation of a carbon budget mitigation
Incidentally, this feature is not obviously visible when looking at strategy would require parallel mitigation efforts for non-CO2 agents.
the four RCPs only, because both the lowest, RCP2.6, and the high- At the time of the IPCC AR5, no established methodologies were
est, RCP8.5, are outliers in terms of non-CO2 warming, at opposite available to ensure easy comparability of carbon budget estimates
sides of the scenario distribution (Fig.3b,c). across working groups. In hindsight and anticipating future assess-
Finally, although non-CO2 forcing does not fully explain the var- ments, three recommendations can be formulated. First, insofar
iations between TEB and TAB estimates, it plays an important role as important topics can already be identified, coordinated model
for the variation within the TEB and TAB subsets. Figure3d shows simulations, intercomparisons, and methods could be initiated at
that respectively 70% and 50% of the variance within the TEB and an early stage to ensure consistency and traceability. Second, con-
TAB subsets can be explained by non-CO2 warming at the time of sistency across and collaboration and integration between the
determining the carbon budget. IPCC working groups could be improved by setting up stronger ties
Future non-CO2 warming under stringent mitigation remains between them. And third, IPCC reports should be clearer about the
nonetheless very uncertain at present. Its magnitude will depend policy-applicability of the numbers they provide, without being pol-
on the extent to which society will be successful in bringing about icy prescriptive.
assumed future improvements in agricultural yields and practices For limiting warming to below 2 C relative to pre-industrial
or dietary changes49, amongst many other factors. These are very levels with greater than 66% probability, the remaining CO2 budget
uncertain. Furthermore, how much non-CO2 forcing is reduced from 2015 onwards for CO2-induced warming only is 1,620GtCO2.
compared with CO2 depends on the relative weight that is given The corresponding TAB budget would be 5901,240 GtCO2. The
to CO2 and non-CO2 emissions in mitigation scenarios, and also latter is equivalent to about 15to 30 years of CO2 emission at cur-
on other mitigation choices50. These weights are mostly constant in rent (2014) levels (about 40 GtCO2 yr1, ref. 52). No matter which
IAMs (for example, by using global warming potentials as a fixed approach is taken, the CO2 budget for keeping warming to below
exchange rate), but can also change over time and depend on the 2C always implies stringent emission reductions over the coming
question posed. decades and net zero CO2 emissions in the medium to long term. For
Air pollution controls can influence the rate of near-term policy-making in the context of the UNFCCC, we suggest using the
warming and, depending on the precise mix of air pollutants that 5901240GtCO2 estimate from 2015 onwards for a likely chance of
is reduced by air pollution controls, non-CO2 warming can be limiting warming to below 2 oC, as this is derived from an assessment
increased, decreased or stay constant 14. The estimated effect of air of scenarios that effectively limit warming to below the 2 C limit.
pollution controls on carbon budgets, in particular on TABs, is very Further efforts will be required to limit warming below 1.5 C.
small51. This is important information for policy-making, as it can
be used to consider trade-offs between the uncertainty in non-CO2 Received 10 April 2015; accepted 15 October 2015;
mitigation, possibly larger CO2 budgets, and a larger amount of published online 24 February 2016
committed warming at the multi-century scale due to larger cumu-
lative CO2 emissions. References
1. United Nations Framework Convention on Climate Change (UN, 1992).
2. Adoption of the Paris Agreement FCC/CP/2015/L.9/Rev.1 (UNFCCC, 2015).
Applicability. Earlier we indicated that budgets that only take into 3. Andrew, H. M., Kirsten, Z., Reto, K. & Matthews, H. D. Sensitivity of
account CO2-induced warming are scientifically best understood carbon budgets to permafrost carbon feedbacks and non-CO2 forcings.
as per definition they do not depend on extra uncertainties Environ. Res. Lett. 10, 125003 (2015).
associated with other forcings. However, at the same time, they are 4. Matthews, H. D. & Caldeira, K. Stabilizing climate requires near-zero
impractical and largely irrelevant for use in the real world, because emissions. Geophys. Res. Lett. 35, L04705 (2008).
5. Matthews, H. D, Gillett, N. P., Stott, P. A. & Zickfeld, K. The proportionality of
of their obvious limitation of neglecting any contribution other than global warming to cumulative carbon emissions. Nature 459, 829832 (2009).
CO2. The other approaches that go beyond this CO2-only approach, 6. Zickfeld, K., Eby, M., Matthews, H. D. & Weaver, A. J. Setting cumulative
might therefore be more practical. Using a CO2-only estimate for emissions targets to reduce the risk of dangerous climate change.
real-word decision-making would lead to an overestimation of the Proc. Natl Acad. Sci. USA 106, 1612916134 (2009).

529
7. Meinshausen, M. etal. Greenhouse-gas emission targets for limiting global 36. Friedlingstein, P. etal. Persistent growth of CO2 emissions and implications for
warming to 2C. Nature 458, 11581162 (2009). reaching climate targets. Nature Geosci. 7, 709715 (2014).
8. Allen, M. R. etal. Warming caused by cumulative carbon emissions towards 37. Schaeffer, M. et al. Mid- and long-term climate projections for fragmented and
the trillionth tonne. Nature 458, 11631166 (2009). delayed-action scenarios. Technol. Forecast. Soc. Change 90, 257268 (2015).
9. Gillett, N. P., Arora, V. K., Zickfeld, K., Marshall, S. J. & Merryfield, W. J. 38. Rogelj, J. etal. Energy system transformations for limiting endofcentury
Ongoing climate change following a complete cessation of carbon dioxide warming to below 1.5C. Nature Clim. Change 5, 519527 (2015).
emissions. Nature Geosci. 4, 8387 (2011). 39. Riahi, K. etal. RCP 8.5 A scenario of comparatively high greenhouse gas
10. Gillett, N. P., Arora, V. K., Matthews, D. & Allen, M. R. Constraining the ratio emissions. Climatic Change 109, 3357 (2011).
of global warming to cumulative CO2 emissions using CMIP5 simulations. 40. Jones, C. etal. Twentyfirstcentury compatible CO2 emissions and airborne
J.Clim. 26, 68446858 (2013). fraction simulated by CMIP5 earth system models under four representative
11. Knutti, R. & Rogelj, J. The legacy of our CO2 emissions: a clash of scientific concentration pathways. J.Clim. 26, 43984413 (2013).
facts, politics and ethics. Climatic Change 133, 361373 (2015). 41. Zickfeld, K. etal. Long-term climate change commitment and reversibility:
12. Smith, S. M. et al. Equivalence of greenhouse-gas emissions for peak an EMIC intercomparison. J.Clim. 26, 57825809 (2013).
temperature limits. Nature Clim. Change 2, 535538 (2012). 42. Stocker, T. F. etal. in Climate Change 2013: The Physical Science Basis
13. Bowerman, N. H. A. et al. The role of short-lived climate pollutants in (eds Stocker, T. F. et al.) 33115 (Cambridge Univ. Press, 2013).
meeting temperature goals. Nature Clim. Change 3, 10211024 (2013). 43. Meinshausen, M., Wigley, T. M. L. & Raper, S. C. B. Emulating atmosphere
14. Rogelj, J. etal. Disentangling the effects of CO2 and short-lived climate forcer ocean and carbon cycle models with a simpler model, MAGICC6 Part 2:
mitigation. Proc. Natl Acad. Sci. USA 111, 1632516330 (2014). Applications. Atmos. Chem. Phys. 11, 14571471 (2011).
15. Collins, M. R. etal. in Climate Change 2013: The Physical Science Basis 44. Rogelj, J., Meinshausen, M., Sedlek, J. & Knutti, R. Implications of potentially
(eds Stocker, T. F. et al.) 10291136 (IPCC, Cambridge Univ. Press, 2013). lower climate sensitivity on climate projections and policy. Environ. Res. Lett.
16. Frolicher, T. L., Winton, M. & Sarmiento, J. L. Continued global warming 9, 031003 (2014).
after CO2 emissions stoppage. Nature Clim. Change 4, 4044 (2014). 45. Smith, L. A. & Stern, N. Uncertainty in science and its role in climate policy.
17. Mastrandrea, M. D. etal. Guidance Notes for Lead Authors of the IPCC Fifth Phil. Trans. R.Soc. A 369, 48184841 (2011).
Assessment Report on Consistent Treatment of Uncertainties 5 (IPCC, 2010). 46. Ricke, K. L. & Caldeira, K. Maximum warming occurs about one decade after a
18. Zickfel, K., Arora, V. K. & Gillett, N. P. Is the climate response to CO2 carbon dioxide emission. Environ. Res. Lett. 9, 124002 (2014).
emissions path dependent? Geophys. Res. Lett. 39, L05703 (2012). 47. Zickfeld, K. & Herrington, T. The time lag between a carbon dioxide
19. Van Vuuren, D. P. etal. Temperature increase of 21st century mitigation emission and maximum warming increases with the size of the emission.
scenarios. Proc. Natl Acad. Sci. USA 105, 1525815262 (2008). Environ. Res. Lett. 10, 031001 (2015).
20. Obersteiner, M. etal. Managing climate risk. Science 294, 786787 (2001). 48. Rogelj, J. etal. Air-pollution emission ranges consistent with the representative
21. Azar, C. etal. The feasibility of low CO2 concentration targets and the role concentration pathways. Nature Clim. Change 4, 446450 (2014).
of bio-energy with carbon capture and storage (BECCS). Climatic Change 49. Gernaat, D. E. H. J. etal. Understanding the contribution of non-carbon dioxide
100, 195202 (2010). gases in deep mitigation scenarios. Glob. Environ. Change 33, 142153 (2015).
22. Tavoni, M. & Socolow, R. Modeling meets science and technology: an 50. Rogelj, J. et al. Mitigation choices impact carbon budget size compatible with
introduction to a special issue on negative emissions. Climatic Change low temperature goals. Environ. Res. Lett. 10, 075003 (2015).
118, 114 (2013). 51. Rogelj, J., Meinshausen, M., Schaeffer, M., Knutti, R. & Riahi, K. Impact of
23. Meinshausen, M., Raper, S. C. B. & Wigley, T. M. L. Emulating coupled short-lived non-CO2 mitigation on carbon budgets for stabilizing global
atmosphereocean and carbon cycle models with a simpler model, warming. Environ. Res. Lett. 10, 075001 (2015).
MAGICC6 Part 1: Model description and calibration. Atmos. Chem. Phys. 52. Le Qur, C. etal. Global carbon budget 2014. Earth Syst. Sci. Data Discuss
11, 14171456 (2011). 7, 521610 (2014).
24. Rogelj, J., Meinshausen, M. & Knutti, R. Global warming under old and new 53. Myhre, G. etal. in Climate Change 2013: The Physical Science Basis
scenarios using IPCC climate sensitivity range estimates. Nature Clim. Change (eds Stocker, T. F. et al.) 659740 (IPCC, Cambridge Univ. Press, 2013).
2, 248253 (2012). 54. Boucher, O. & Reddy, M. S. Climate trade-off between black carbon and carbon
25. Matthews, H. D., Solomon, S. & Pierrehumbert, R. Cumulative carbon as a dioxide emissions. Energy Policy 36, 193200 (2008).
policy framework for achieving climate stabilization. Phil. Trans. R.Soc. A 55. Brohan, P., Kennedy, J. J., Harris, I., Tett, S. F. B. & Jones, P. D. Uncertainty
370, 43654379 (2012). estimates in regional and global observed temperature changes: a new data set
26. IPCC Summary for Policymakers in Climate Change 2013: The Physical from 1850. J.Geophys. Res.-Atmos. 111, D12106 (2006).
Science Basis (eds Stocker, T. F. et al.) 129 (Cambridge Univ. Press, 2013).
27. Clarke, L. et al. International climate policy architectures: overview of the Acknowledgements
EMF 22 International Scenarios. Energy Econ. 31, S64S81 (2009). We acknowledge the work by IAM modellers that contributed to the IPCC AR5 Scenario
28. Riahi, K. etal. in Global Energy AssessmentToward a Sustainable Database and the climate modelling teams contributing to CMIP5. We thank IIASA for
Future (eds Johansson, T. B., Patwardhan, A., Nakicenovic, N. & hosting the IPCC AR5 Scenario Database, and M. Meinshausen for detailed comments
Gomez-Echeverri, L.) 12031306 (Cambridge Univ. Press and IIASA, 2012). and feedback on the manuscript.
29. Kriegler, E. etal. The role of technology for achieving climate policy
objectives: overview of the EMF 27 study on global technology and climate
policy strategies. Climatic Change 123, 353367 (2014).
Author contributions
All authors contributed to parts of the underlying research during the writing process
30. Moss, R. H. etal. The next generation of scenarios for climate change research
of the IPCC AR5. J.R. coordinated the conception and the writing of the paper. J.R.
and assessment. Nature 463, 747756 (2010).
carried out the research with significant contributions from M.S., and developed the TEB
31. Van Vuuren, D. P. etal. The representative concentration pathways:
and TAB conceptual framework. J.R. produced the figures and wrote the first draft of
an overview. Climatic Change 109, 531 (2011).
the manuscript. All authors contributed to interpreting and discussing the results, and
32. Taylor, K. E., Stouffer, R. J. & Meehl, G. A. An Overview of CMIP5 and the writing the paper.
Experiment Design. Bull. Am. Meteorol. Soc. 93, 485498 (2011).
33. IPCC Summary for Policymakers in Climate Change 2014: Synthesis Report
(eds Pachauri, R. K. et al.) 133 (Cambridge Univ. Press, 2014). Additional information
34. IPCC Summary for Policymakers in Climate Change 2014: Mitigation Reprints and permissions information is available online at www.nature.com/reprints.
of Climate Change (eds Edenhofer, O. et al.) 133 (Cambridge Univ. Correspondence should be addressed to J.R.
Press, 2014).
35. Clarke, L. etal. in Climate Change 2014: Mitigation of Climate Change Competing financial interests
(eds Edenhofer, O. et al.) 413450 (Cambridge Univ. Press, 2014). The authors declare no competing financial interests.

530
LI
This is Exhibit ".l..l' relerred to in the
I affrdavit of Kirsten Zickfeld,
before me on the
day of April, 2017.
A commissioner for
British Columbia
531
PERSPECTIVE
PUBLISHED ONLINE: 21 SEPTEMBER 2014|DOI: 10.1038/NCLIMATE2384
Sharing a quota on cumulative carbon emissions

Michael R. Raupach1*, Steven J. Davis2, Glen P. Peters3, Robbie M. Andrew3, Josep G. Canadell4,
Philippe Ciais5, Pierre Friedlingstein6, Frank Jotzo7, Detlef P. van Vuuren8,9 and Corinne Le Qur10
Any limit on future global warming is associated with a quota on cumulative global CO2 emissions. We translate this global
carbon quota to regional and national scales, on a spectrum of sharing principles that extends from continuation of the present
distribution of emissions to an equal per-capita distribution of cumulative emissions. A blend of these endpoints emerges as
the most viable option. For a carbon quota consistent with a 2C warming limit (relative to pre-industrial levels), the necessary
long-term mitigation rates are very challenging (typically over 5% per year), both because of strong limits on future emissions
from the global carbon quota and also the likely short-term persistence in emissions growth in many regions.
C
limate modelling studies16 have established a robust near-linear given adequate social capital35 to support a framework for adaptive
relationship between global warming and cumulative CO2 governance33. When the sharing challenge is framed in this way, the
emissions since industrialization. This implies that a carbon emphasis shifts away from questions about global rules (What shares
quota or cap on future cumulative CO2 emissions is required if global of the carbon quota should be allocated to every country?) to ques-
warming is to be kept below any nominated limit (such as 2C above tions about consistent local behaviours (If others acted consistently
pre-industrial temperatures7) with a nominated chance of success810. with our proposed share of the carbon quota, would the global out-
Estimated carbon quotas are significantly smaller than the known come be acceptable to us?). This view further motivates a direct con-
global fossil-fuel reserves2,11,12. nection between the global carbon quota and effort-sharing analyses,
The carbon quota implies that future cumulative CO2 emissions to explore frameworks that can infer the global implications of a pro-
consistent with a given warming limit are a finite common global posed share of the carbon quota by any one country, were others to
resource that must necessarily be shared among countries, whether act on similar principles.
through prior agreement or as an emergent property of individually To establish principles, we focus on the sharing of fossil-fuel CO2
determined national efforts. The problem of sharing the global miti- emissions, the largest single contributor to radiative forcing and cli-
gation effort is addressed in an extensive literature, from the perspec- mate change36. Emissions of CO2 from land-use change are now a
tives of equity, international policy and institutions, and economics relatively small fraction of total CO2 emissions (83%)37, declining
and financing 1322. Here, we combine perspectives from two previ- with time, and subject to significant uncertainty at the global scale
ously distinct strands of analysis the global carbon quota and effort and even more at regional scales37,38. Inclusion of CO2 emissions from
sharing to infer the regional and national implications of global land-use change is straightforward in principle, though data uncer-
carbon quotas under a wide range of possible sharing strategies. tainties would require careful assessment. In the absence of historic
The need for multiple approaches is heightened by the pre- attribution, the effects would be small for most regions, but large for
sent impasse in the search for long-term climate safety. Two broad tropical forest countries where land clearing is still ongoing. More sig-
approaches have been pursued hitherto in international negotia- nificant at the global scale is the role of non-CO2 forcing agents, both
tions: top-down international agreements, such as the 1997 Kyoto those accounted (the major non-CO2 greenhouse gases) and unac-
Protocol23, and bottom-up nationally determined contributions to a counted (aerosols) in inventories. However, full inclusion of these
global outcome. The top-down approach has made little progress over forcing agents in extensions to carbon quotas at regional and national
the last two decades24. An approach based on nationally determined scales is beyond the present scope, requiring more complex climate
contributions is now being explored25, but current commitments in modelling to resolve issues such as local impacts of short-lived cli-
sum are far short of what is needed to meet internationally agreed mate forcers39,40, nonlinear forceresponse relationships41 and cooling
climate goals2629. by some aerosol species36.
The present impasse arises in part because the sharing challenge
forms a tragedy of the commons30 or collective-action dilemma31. Ways of sharing a cumulative emissions quota
The challenge of governing common natural resources has devel- A common-pool resource can be shared objectively among
oped a rich literature encompassing issues of governance, institu- participants in a socialecological system by distributing the resource
tions, communities and ethics3234. In broad terms, this literature according to a set of observable metrics. In the case of the carbon
suggests that solutions to the underlying problem of collective action quota, two generic metrics are measures of inertia (also known
can emerge from individual actions by participants (here, countries), as grandfathering22) and equity, the inertia metric reflecting
1
Climate Change Institute, Australian National University, Canberra, Australian Capital Territory 0200, Australia, 2Department of Earth System Science,
University of California, Croul Hall, Irvine, California 92697, USA, 3Center for International Climate and Environmental Research Oslo (CICERO), PO Box
1129 Blindern, N-0318 Oslo, Norway, 4Global Carbon Project, Oceans and Atmosphere Flagship, Canberra, Australian Capital Territory2601, Australia,
5
IPSL-LSCE, CEA CNRS UVSQ, Centre dEtudes Orme des Merisiers, 91191Gif sur Yvette, France, 6College of Engineering, Mathematics and Physical
Sciences, University of Exeter, Exeter EX44QF, UK, 7Crawford School of Public Policy, Australian National University, Canberra, Australian Capital
Territory0200, Australia, 8PBL Netherlands Environmental Assessment Agency, Postbus 303, 3720 AHBilthoven, The Netherlands, 9Copernicus Institute
of Sustainable Development, Faculty of Geosciences, Utrecht University, Postbus 80.115, 3508TC Utrecht, The Netherlands, 10Tyndall Centre for Climate
Change Research, University of East Anglia, Norwich Research Park, NorwichNR47TJ, UK. *e-mail: michael.raupach@anu.edu.au

532
19 5 9 12 18 8
11 7 3 6
4
8 5
11 2
11
10 5
4
5
27 8 22
4 17
5 25
6 7 16
Inertia: allocated by current emissions (w = 0) Blended allocation (w = 0.5) Equity: allocated by population (w = 1)
Europe Pacific China+ Rest of Asia Africa

North America Reforming economies India+ Middle East Latin America
Figure 1 | Sharing the carbon-quota pie. The share of an available carbon quota allocated to 10 regions (Europe, North America, Pacific Organisation for
Economic Co-operation and Development countries, reforming economies, China+, India+, Rest of Asia, Middle East, Africa, Latin America) under three
sharing principles based on equation(2), with sharing index w=0, 0.5and 1. Numbers give the percentage share of the global quota allocated to each
region, summing to 100for each chart. Shares are calculated using equation (2) with emissions (fj) averaged over last fiveyears of data, and population
(pj) averaged over a five-year period centred on the time at which world population reaches nine billion. See Supplementary Text 1 for details.
the distribution of emissions and the equity metric reflecting the global average per-capita emissions (F/P) (Supplementary Text 2);
population distribution. These metrics would suggest two alterna- the last four regions have below-global-average per-capita emissions
tive sharing principles: (Supplementary Fig. 1).
fj pj The concept of a blended sharing principle can potentially be
sj (emissions) = ; sj (population) = (1) generalized to include additional metrics of responsibility and/or
F P capability 19,21 for example, the distribution of gross domestic
where sj is the share of the quota to country j; fj and pj are the emissions product (GDP) as a measure of capability to undertake mitigation
(current or cumulative) and population (present or future), respec- efforts (Supplementary Text 2). The influences of emissions and
tively, for country j; and F and P are the corresponding emissions GDP on sharing are broadly similar because both are correlated
and population for the world. Shares sum to one over all countries with development status, and both are very different to the influence
(sj=1) because fj=F and pj=P. Depending on the choice of refer- of population (Supplementary Fig.1). Therefore, we focus mainly
ence times for emissions and population, this formulation can accom- on emissions and population using equation (2), and briefly explore
modate sharing by current or historic emissions, and can account for later the effect of also including GDP in the sharing principle. We
expected future changes in population. also note that allocated shares and quotas are not the same as actual
In their simplest forms, both options in equation(1) face major future cumulative emissions if emissions are tradable between
difficulties. Sharing by present emissions (inertia) would leave devel- countries, as discussed later.
oping countries with little access to the energy and development
opportunities embodied in remaining future carbon emissions, Regional carbon quotas
whereas sharing by population (equity) would impose extremely high The global carbon quota from the past to the long-term future
mitigation demands on many developed countries. This has moti- (when emissions fall to zero) is:
vated the analysis of blended sharing principles1618 that can compro-
mise between the endpoint positions. One possibility (among others Qtot = Qpast(FFI) + Qpast(LUC) + Qfuture(FFI) + Qfuture(LUC)(3)
explored below) is that the share of the quota to country j is:
fj pj where Qtot is the quota for anthropogenic CO2 emissions from a
sj(w) = (1 w) +w (2) reference time (here 1870) to the far future, including contributions
F P from fossil-fuel combustion and industrial processes (FFI) and net
where w is a sharing index between 0and 1, weighting between the land-use change (LUC); Qpast is the past emissions and Qfuture is the
endpoints of sharing by inertia (w=0) and by equity (w=1). This shared available future emissions. Past cumulative CO2 emissions
principle also satisfies the requirement sj=1.It can be regarded as from 1870 to the end of 2012 were 1,922 Gt CO2 (1,396 Gt CO2
a simplified form of the contraction-and-convergence algorithm1618, from FFI and 526GtCO2 from LUC)37. Global LUC emissions have
applied to a total carbon quota rather than to emissions trajectories decreased since 2000 to 83% of total emissions in 201337 and are
specified through time; the key simplification is independence from expected to continue to decrease; a linear decrease to zero in 2100
specific assumptions about emissions pathways through time. would imply Qfuture(LUC)=137GtCO2.
Using equation(2), Fig.1 shows how w influences the share of Estimation of the global carbon quota Qtot is an ongoing scientific
a global carbon quota assigned to 10 regions that span the world issue. The estimates used here10 are Qtot=3,500, 4,400 and 5,300GtCO2
(Europe, North America, Pacific Organisation for Economic for warming limits of 2, 2.5 and 3C, respectively, with 50% chance of
Co-operation and Development countries, reforming economies, success and accounting for both CO2 and non-CO2 forcings (all quota
Middle East, China+, India+, Rest of Asia, Africa, Latin America; estimates are rounded to nearest 100GtCO2). These are larger (more
Supplementary Text 1). The last four regions receive a greater conservative) quotas than estimated elsewhere8.
share with increasing weighting of equity (increasing w), while We consider sharing of the available quota of future fossil-fuel
the share for the other six regions decreases. This occurs because emissions Qfuture(FFI) from equation (3), henceforth Qavail. The above
the response to increasing w of the share sj for a region j hinges on estimates for Qtot imply that Qavail=1,400, 2,300 and 3,200GtCO2, for
whether its per-capita emissions (fj/pj) are less or greater than the warming limits of 2, 2.5 and 3C at 50% chance of success.

533
22 81 118 28 40 57
Fossil reserves 98 637 231
Inertia 46
176 171 171 46
w=0
Blended Pacific
394 99 72 107 107 North America 68 31
w = 0.5 31
16
Equity
17
w=1
31 42 42 -7 34 17
33
7 7 101 17 64 105
Fossil reserves 487 99
Inertia 39
120 240 240 4848
w=0
Latin America
Blended 160 267 52 199 199 China+ 55 60 61 61
w = 0.5
Equity 157 157 81 74 74
w=1
15 38
6 8 4 7 16 9
Fossil reserves 69 62
33
Inertia 77 99 99 52 52
w=0
Rest of Asia
Blended 78 78 4651 65 72 72
268 82 Europe
w = 0.5
Equity 56 56 98 92 92
w=1
43 8
3 12
100
Fossil reserves 53 130 545 330 169

29
Inertia 130 98 98 4949
w=0 Middle East
Reforming economies 17
Blended 222 87 71 71 50 41
w = 0.5 41 Unconventional gas
4
Equity 44 34
Unconventional oil
w=1
27 44 44 40 34
Coal
35 10 30 83 Gas
Fossil reserves 220 52 93 Oil
27
Inertia India+ 33 Africa 2 C 2.5C 3 C
48 63 63 Quota remaining
w=0 33
Blended 47 53 173 141 141 131 98 98 Committed
w = 0.5
Past cumulative
Equity 299 219 219 235 162 162
w=1
36 26
0 200 400 600 800 1,000 1,200 1,400 0 200 400 600 800 1,000 1,200 1,400
Cumulative emissions (Gt CO2) Cumulative emissions (Gt CO2)
146 599
Fossil reserves 658 412 2381
Inertia
w=0
Blended 1335 729 714 898 898
w = 0.5 Global
Equity
w=1
0 1,000 2,000 3,000 4,000 5,000

Cumulative emissions (Gt CO2)
Figure 2 | Quotas, cumulative committed emissions and fossil-fuel reserves. Past cumulative fossil-fuel CO2 emissions (purple), future committed
emissions42,43 (orange) and available fossil-fuel carbon quotas to meet warming limits of 2, 2.5and 3C with 50% probability (green), for 10 regions and
the world, under inertia, blended and equity sharing principles. Stacked bars are cumulative; numbers give the contribution of each increment in GtCO2.
Negative increments are shown below the zero axis. Also shown are fossil-fuel reserves (coal, oil, gas, unconventional oil, unconventional gas)12.

534
The available carbon quota for country j is a share sjQavail of the where f(t) is the emissions at time t, f0 is the emissions at the start of
global quota. Figure 2 shows the resulting quotas for 10 regions mitigation (t = 0), and r and m both have units of per year. When
(Fig.1) and for the world, with shares sj from equation(2) for three mitigation is started at t=0 (with a positive initial growth rate r), the
values of w (0, 0.5 and 1, corresponding to inertia-based, blended and resulting emissions trajectory increases, peaks and eventually declines
equity-based sharing, respectively), and with available quotas Qavail exponentially at the rate m (Supplementary Fig.4). A possible delay
corresponding to warming limits of 2, 2.5 and 3C at 50% chance of in starting mitigation can also be included (Supplementary Text3).
success. Global quotas are independent of w, but the regional quo- By incorporating such a delay, equation(4) can provide a good rep-
tas depend strongly on w, with increasing w causing the quotas to resentation of the trajectories of CO2 emissions from fossil fuels in
increase for regions with low per-capita emissions, and vice versa the four representative concentration pathway scenarios45 before any
(Supplementary Text2). introduction of negative emissions (Supplementary Fig.5). This indi-
The regional quotas can be assessed against two independently cates that equation (4) is suitable for empirically describing persis-
determined quantities. First, committed emissions (orange bars in tence effects in emissions trajectories.
Fig.2) are estimates of future emissions from existing CO2-emitting To meet a specified available cumulative emission quota, persis-
infrastructure that will continue for infrastructure lifetimes without tence in emissions growth causes the necessary eventual characteristic
early retirement4244 (Supplementary Text1). Committed emissions in rate of decline in emissions (m) to be typically more than twice the rate
North America, Europe and China exceed quotas for a 2C warming 1/T that would be required if exponential decline could commence
limit under equity sharing (w=1), implying a requirement to either immediately (Supplementary Text3, Supplementary equation(8)).
retire or improve such infrastructure before its design lifetime, or to
compensate these emissions by negative emissions later in the century
a
or by some form of offset such as emissions trading (see below). For
16
the world, committed emissions are about half of the available quota
Qavail for a 2C warming limit. Inertia
14 Equity
w=0
Second, quotas can be compared with fossil-fuel reserves of coal, Blended w = 1
oil, gas, unconventional oil and unconventional gas (Fig.2). Reserves 12
Mitigation rate (% per yr)

w = 0.5
are the part of total resources currently identified as economically via-
10
ble for extraction. Globally and in most regions, estimated reserves12
substantially exceed the global quota Qavail for warming limits up to 8
and beyond 3C, in agreement with other assessments2,11. Estimates
of total fossil-fuel resources are even larger. 6
4
The distribution of the mitigation challenge
A simple measure of the challenge implied by the available quota 2
for any region or country (before any possible redistribution by
emissions trading) is the time for which the quota would last if 0
emissions were held steady at current levels until the quota is b
exhausted, Tj = sjQavail/fj. This emission time10 depends strongly
on the sharing index w (Supplementary Fig. 2). With pure emis- 14
sion-based sharing (w = 0), the emission time for all countries 66% chance of 2 C 50% chance of 3 C
12 50% chance of 2 C 50% chance of 2.5 C
is the same and equals the global emission time Qavail/F. As w
increases to yield pure population-based sharing at w = 1, the 10

emission time increases (decreases) for regions with per-capita
emissions less (greater) than the global average. The response 8
of emission times to w is the same as, and is determined by, the
response of shares (sj) to w (Supplementary Text 2). 6
If emissions were to decrease at a steady exponential rate
4
starting immediately, an emission time T would correspond to a
decrease in emissions at a fractional rate 1/T (or 100/T per cent per 2
year). However, this estimate of a required reduction rate to meet
a given quota is too low if the mitigation effort must first over- 0
Reforming economies
North America
Europe
Pacific
India+
Rest of Asia
Middle East
Africa
Latin America
Global
China+
come existing emissions growth, because of persistence effects.

Persistence in emissions growth arises from the time needed to
implement new low-emission energy technologies on the energy
supply side, and to adopt energy-efficiency measures and make
behavioural changes in energy consumption on the demand side.
Persistence is evident in emissions data (Supplementary Fig.3).
The supply-side aspects of this persistence arise mainly from the
committed emissions in existing, long-lived energy infrastruc- Figure 3 | Dependence of the regional mitigation challenge on the sharing
ture42,43 (Fig.2). index (w) and the warming limit. a,Mitigation rates for 10 regions and
We account for persistence in emissions growth by represent- the world at w=0, 0.5and 1. Available global fossil-fuel combustion and
ing the future emissions of a country, region or the world with industrial processes (FFI) carbon quota from 2013 is Qavail=1,400GtCO2,
an analytic capped-emissions trajectory that blends an initially corresponding to a 2C warming limit with 50% success probability.
linear growth at rater with eventual exponential decline at a miti- b,Mitigation rates under a blended sharing principle (w=0.5) in four cases
gation ratem. Continuity requirements determine this trajectory with warming limit and success probability, respectively, equal to 2C and
uniquely (Supplementary Text3): 50%; 2.5C and 50%; 3C and 50%; and 2C and 66%. The available
global FFI carbon quotas for these four cases are Qavail=1,400, 2,300,
) ) )) ) )
f (t) = f0 1 + r + m t exp mt (4) 3,200 and 1,100GtCO2, respectively10.

535
This occurs because the persistence in emissions growth in the early 20
phase of the mitigation effort has to be compensated by more rapid Equity, w = 1
later decline (larger m). Blended, w = 0.5
Figure3a shows the mitigation rates m needed to meet an available Inertia, w = 0
carbon quota Qavail=1,400GtCO2 (a 2C warming limit at 50% suc-
Russia
cess probability), for the world and in 10 regions, with sharing index 15
USA
South Korea
w=0, 0.5 and 1. The required global mitigation rate to meet the quota
Australia
is 5.5% per year (independent of sharing index) more than twice
the reduction rate 1/T if exponential decline could start immediately,
Canada
because of persistence in emissions growth. This result is consistent
Japan
with scenario-based analyses that account for policy delay 46,47. 10
China
With pure emissions-based sharing (w=0), m varies little among
South Africa
regions (Fig.3a); it is not identical across regions (in contrast with
Brazil
India
World
the emission time T; Supplementary Fig.2) because of regional varia-
tions in the initial growth rate r. With pure population-based sharing
5
Indonesia
(w=1), m varies greatly among regions, from 1.4% per year (Africa)
EU28
Ethiopia
to over 15% per year (North America). With a blended sharing prin-
Mexico
ciple (w=0.5), required mitigation rates are intermediate between the
ROW
endpoint options w=0and w=1, but very different in most cases
from a simple average of the endpoints. For North America, the 0
required mitigation rate at w=0.5 is about 30% more than with emis- 0 5 10 15
sions-based sharing (w=0); for Africa, it is about 70% less. Therefore, Per capita emissions in 2012 (tCO2 per person per yr)
a shift from an emissions-based to blended sharing principle leads
to large benefits for developing regions at the cost of a much smaller Figure 4 | Distribution of the mitigation challenge among countries.
increase in the demands on developed regions, as quantified by frac- Mitigation rates (m) for 14 countries and regions spanning the
tional changes in the required mitigation rates m. development spectrum and for the whole world, with sharing index w=0
Regional mitigation rates are also strongly sensitive to the global (open squares),1 (filled squares) and0.5 (half-open squares). Horizontal
carbon quota, determined by the warming limit and probability axis is 2012per-capita fossil-fuel CO2 emissions. Available global fossil-fuel
of success. If the required probability of success for a 2 C limit is carbon quota from 2013is Qavail=1,400GtCO2, corresponding to a 2C
increased from 50% (as in Fig.3a) to 66%, then the required global warming limit with 50% success probability. With increasing equity in the
mitigation rate m increases from 5.5 to 7% per year, with commensu- sharing principle, the mitigation challenge increases for countries to the
rate increases for regions and countries (Fig.3b). For warming limits right of the point for the world (the pivot for the see-saw) and decreases
of 2.5 and 3C at 50% success probability, the required global mitiga- for countries to the left. Mexico is so close to the pivot that symbols are
tion rates fall to 3.7 and 2.9% per year, respectively, with commensu- indistinguishable. ROW, rest of world; EU28, the 28 member states of the
rate decreases at regional and national levels. Even a 3C limit (with a European Union.
much higher risk of dangerous climate change48) requires significant
global and national mitigation. but small effects for countries close to the pivot point defined by
To explore further the distribution of the mitigation chal- global-average per-capita emissions (Fig.4). In particular, China has
lenge at national level, we focus on a set of 14representative coun- a high required mitigation rate (because of currently high emissions
tries (Supplementary Text 1) that span the development spectrum growth) that is not strongly sensitive to the choice of w.
in terms of both per-capita emissions and rates of development The regional mitigation rates in Figs 3a and 4 pivot around a very
(Supplementary Fig.6; a national-level counterpart of Fig.2 is given challenging global mitigation rate of 5.5% per year, for a warming
in Supplementary Fig.7). The required mitigation rates m for these limit of 2 C with 50% chance of success. If the associated global
countries, before any possible emissions trading, are plotted against carbon quota from ref. 10 (Qtot = 3,500 Gt CO2) is reduced to a
present per-capita emissions in Fig. 4. Increasing equity (larger w) more stringent 3,000 GtCO2 (ref. 8), then the required global miti-
causes the mitigation challenge to respectively increase and decrease gation rate increases further to 7.9% per year, and regional rates
for countries with per-capita emissions above and below the world increase correspondingly.
average, pivoting about that point. For least-developed countries
with very low per-capita emissions, a shift from w=0 (inertia) to 0.5 Additional factors
(blended) reduces the mitigation challenge from near the world aver- To this point, we have not yet considered several additional factors
age to near zero, because these countries collectively account only for that can be assessed within the framework of equation (2) or its
a small share of global emissions. generalizations (Supplementary Text 3) as part of future climate
The implication is that a blended sharing principle can ameliorate policy regimes.
the opposite difficulties associated with the endpoint sharing principles
at w=0 (which would be prohibitive for least-developed countries) or Extent of inclusion of historic emissions. It has been sug-
w=1 (which would be prohibitive for developed countries because gested41,49,50 that historic responsibility for climate change be
of required mitigation rates exceeding 15% per year). Such compro- included in principles for sharing the mitigation challenge. In a
mises will be necessary in applying equity principles of responsibility carbon-quota approach, this involves defining an attribution start
and capability to apportion the burden of emissions reductions [and] date in the past and then sharing cumulative global emissions from
address concerns of both the global North and South24. that time onwards rather than from the present (Supplementary
Together, Figs3and 4 demonstrate the interplay between the three Text4). A shift to this sharing principle has no effect on the required
major factors determining required regional and national mitigation global mitigation rates, but has large implications for regions and
rates: the warming limit, the nominated chance of success and the countries. With historic attribution, required mitigation rates for
sharing principle (here w). The first two are comparably important developed regions become very large because attributed historic
everywhere. The sharing principle (w) has dominant but opposite emissions approach (or even exceed) allocated shares for future
effects for countries at opposite ends of the development spectrum, emissions (Supplementary Fig. 8). The corresponding benefits for

536
developing regions are not as large as might be expected because key requirement for a robust solution to the climate change challenge
historic emissions for these regions are low. through adaptive governance. As a methodological contribution to
assist in this kind of testing, the present work combines existing analy-
Effect of delaying mitigation. It is well known that a delay in starting ses of the global carbon quota and effort sharing. The carbon-quota
mitigation has a profound effect on the steepness of the mitigation approach offers the important simplification of independence from
challenge51,52. Noting that our analysis already includes persistence assumptions about emissions pathways in time, yielding a transparent
before a peak in emissions is reached, an additional 10-year delay methodology for translating global to national carbon quotas under a
would increase the required global mitigation rate m from 5.5 to wide range of possible sharing principles.
over 9% per year (with global quota Qavail = 1,400 Gt CO2), with The question of achievability is clearly central13. The required global
commensurate increases in regions (Supplementary Fig.9). mitigation rates emerging from our analysis are high, typically over 5%
per year for a 2C limit at 50% success probability (and 8% per year
Consumption-based and territorial emissions accounting. for China, a rate that remains very high under any sharing principle;
Consumption-based inventories for national CO2 emissions53 aug- Fig.4). These rates can be compared with the distribution of maximum
ment established territorial inventories54 by attributing emissions to mitigation rates in the ensemble of scenarios in the Intergovernmental
countries where products are consumed rather than where emissions Panel on Climate Change Fifth Assessment Report (Supplementary
of manufacture occur 5557. Under consumption-based accounting, Fig.13). For scenarios with CO2 peaking below 530ppm, the median
the emissions of manufacturing-export countries, such as China, of the distribution of maximum rate of emissions decline is approxi-
are reduced by up to 20% in recent years (relative to territorial mately consistent with the required rates from our analysis if there is
accounting) and emissions of importing countries are correspond- no delay in starting mitigation, but even a five-year delay causes the
ingly increased37,58. Use of consumption-based rather than territo- required rate to approach the upper edge of the distribution.
rial emissions leads to only a small change in shares and mitigation Although the global quota is determined biophysically, the result-
rates for regions and countries (Supplementary Fig.10), because the ing distribution of effort among countries can be made more achiev-
favourable effects of consumption-based accounting for manufac- able by emissions or quota trading and related instruments. These can
turing-export countries are offset by the effects of their typically high help to make very high national mitigation targets achievable, given
persistence in emissions growth. Still, consumption-based emissions sufficient globally connected trading systems and an effective price on
accounting may contribute to negotiation of agreements24. emissions. Quota trading means that an initial quota does not deter-
mine the actual future cumulative emissions for a country; it also can
Effect of timing of population distribution. Sharing by population improve the overall cost-effectiveness of the global mitigation effort,
can be based on a future population forecast (the default for this and facilitate transfer payments between countries to help achieve
Perspective; Supplementary Text 1), or on the present population desired distributional outcomes. It is an open question whether such
distribution. There is only a small sensitivity of required mitiga- payments can be actually implemented on a large scale.
tion rates to whether sharing occurs on the population distribution For the emergence of long-term, cooperative solutions to anthro-
at a future time when the global population is nine billion, or on pogenic climate change33,35, one essential element is an ability to per-
the distribution in 2013 with a global population of seven billion ceive the consistent global consequences of local actions, given great
(Supplementary Fig. 11). differences in national economies and histories. The social capital
that underpins cooperative governance of the commons takes time
Effect of including GDP in the sharing principle. Equation (2) to evolve, but the biophysical realities of climate change demand solu-
can be generalized to include additional metrics such as GDP tions within decades. This is why the development of new perspec-
(Supplementary Text 2). If the emissions distribution in equa- tives on the sharing challenge is vital.
tion (2) is replaced completely with the GDP distribution, the
resulting effect on shares and mitigation rates is moderate, but not Received 30 June 2014; accepted 27 August 2014;
large (Supplementary Fig. 12), because both emissions and GDP are published online 21 September 2014
correlated with development status. Sharing principles combining
three or more metrics (emissions, GDP, population, and so on) can References
1. Allen, M.R. etal. Warming caused by cumulative carbon emissions: towards
be constructed (Supplementary Text 2). The most important clus-
the trillionth tonne. Nature 458, 11631166 (2009).
ters of metrics are those that represent development status (such as 2. Meinshausen, M. etal. Greenhouse gas emission targets for limiting global
emissions and GDP) and those representing population. warming to 2C. Nature 458, 11581162 (2009).
3. Zickfeld, K., Eby, M., Matthews, H.D. & Weaver, A.J. Setting cumulative
Negative emissions. Model-based scenario studies indicate pathways emissions targets to reduce the risk of dangerous climate change.
to a range of warming limits, through transformations in energy sys- Proc. Natl Acad. Sci. USA 106, 1612916134 (2009).
4. Matthews, H.D., Gillett, N.P., Stott, P.A. & Zickfeld, K. The proportionality of
tems and other mitigation measures13. For limits around 2C or less,
global warming to cumulative carbon emissions. Nature 459, 829833 (2009).
such scenarios often depend on the use of negative emissions13,5961 5. Raupach, M.R. etal. The relationship between peak warming and cumulative
through strategies such as land-based biosequestration or bioenergy CO2 emissions, and its use to quantify vulnerabilities in the carbonclimate
with carbon capture and storage. Most 2C scenarios propose signifi- human system. Tellus B 63, 145164 (2011).
cant gross negative emissions to offset gross positive emissions that 6. Raupach, M.R. The exponential eigenmodes of the carbonclimate system,
may be difficult or impossible to avoid, and many propose net negative and their implications for ratios of responses to forcings. Earth Syst. Dynam.
4, 3149 (2013).
emissions from the late twenty-first century. To the extent that gross
7. EuropeanCommission The 2C Target: Background on Impacts, Emission
negative emissions offset gross positive emissions, they are handled Pathways, Mitigation Options and Costs (European Commission, 2008).
naturally by the cumulative carbon quota approach because the car- 8. Collins, M. et al. in IPCC Climate Change 2013: The Physical Science Basis
bon quota applies to net (gross positive less gross negative) emissions. (eds Stocker, T.F. et al.) Ch. 12, 10291136 (Cambridge Univ. Press, 2013).
This applies at regional and national scales just as at the global scale. 9. Matthews, H.D., Solomon, S. & Pierrehumbert, R. Cumulative carbon as a
policy framework for achieving climate stabilization. Phil. Trans. R.Soc. A
370, 43654379 (2012).
Shared responsibility and collective achievability 10. Friedlingstein, P. etal. Persistent growth of CO2 emissions and implications for
A longstanding idea in analyses of burden sharing has been that coun- reaching climate targets. Nature Geosci. (in the press).
tries need to test and explain how their own nominated climate goals 11. GEA Global Energy Assessment Toward a Sustainable Future (Cambridge
fit with a global outcome16,20,22,62,63. Engagement in such testing is a Univ. Press and International Institute for Applied Systems Analysis, 2012).

537
12. BGR Energy Study 2013: Reserves, Resources and Availability of Energy 47. Riahi, K. et al. Locked into Copenhagen pledges implications of short-term
Resources (Federal Institute for Geosciences and Natural Resources, 2013). 386 emission targets for the cost and feasibility of long-term climate goals.
13. Clarke, L. etal. in IPCC Climate Change 2014: Mitigation of Climate Change Technol. Forecast. Soc. http://dx.doi.org/10.1016/j.techfore.2013.09.016
(eds Edenhofer, O. etal.) Ch. 6 (Cambridge Univ. Press, 2014). (in the press).
14. Stavins, R. etal. in IPCC Climate Change 2014: Mitigation of Climate Change 48. Schellnhuber, H.J., Cramer, W., Nakicenovic, N., Wigley, T.M. L. & Yohe, G.
(eds Edenhofer, O. etal.) Ch. 13 (Cambridge Univ. Press, 2014). Avoiding Dangerous Climate Change (Cambridge Univ. Press, 2006).
15. Hhne, N., den Elzen, M.G. J. & Escalante, D. Regional GHG reduction 49. Den Elzen, M.G. J. etal. The Brazilian Proposal and Other Options for
targets based on effort sharing: a comparison of studies. Clim. Policy International Burden Sharing: An Evaluation of Methodological and Policy
14, 122147 (2014). Aspects using the FAIR Model (National Institute of Public Health and the
16. Meyer, A. Contraction and Convergence. The Global Solution to Climate Change Environment, 1999).
Schumacher Briefings 5 (Green Books, 2000). 50. Den Elzen, M.G. J., Schaeffer, M. & Lucas, P.L. Differentiating future
17. Hohmeyer, O. & Rennings, K. Man-made Climate Change: Economic Aspects commitments on the basis of countries relative historical responsibility for
and Policy Considerations (ZEW Economic Studies, Centre for European climate change: uncertainties in the Brazilian proposal in the context of a
Economic Research, 1999). policy implementation. Climatic Change 71, 277301 (2005).
18. Bows, A. & Anderson, K. Contraction and convergence: an assessment of the 51. Den Elzen, M.G. J., van Vuuren, D.P. & van Vliet, J. Postponing emission
CCOptions model. Climatic Change 91, 275290 (2008). reductions from 2020to 2030 increases climate risks and long-term costs.
19. Dellink, R. etal. Sharing the burden of financing adaptation to climate change. Climatic Change 99, 313320 (2010).
Glob. Environ. Change 19, 411421 (2009). 52. Stocker, T.F. The closing door of climate targets. Science
20. Bartsch, U. & Muller, B. Fossil Fuels in a Changing Climate Impacts of the 339, 280282 (2013).
Kyoto Protocol and Developing Country Participation (Oxford Univ. Press, 2000). 53. Peters, G.P., Minx, J.C., Weber, C.L. & Edenhofer, O. Growth in emission
21. Fssel, H-M. How inequitable is the global distribution of responsibility, transfers via international trade from 1990to 2008. Proc. Natl Acad. Sci. USA
capability, and vulnerability to climate change: a comprehensive indicator- 108, 89038908 (2011).
based assessment. Glob. Environ. Change 20, 597611 (2010). 54. Andres, R.J. etal. A synthesis of carbon dioxide emissions from fossil-fuel
22. Ringius, L., Torvanger, A. & Underdal, A. Burden sharing and fairness principles combustion. Biogeosciences 9, 18451871 (2012).
in international climate policy. Int. Environ. Agreem. P. 2, 122 (2002). 55. Peters, G.P. & Hertwich, E.G. CO2 embodied in international trade
23. UNFCCC Kyoto Protocol to the United Nations Framework Convention with implications for global climate policy. Environ. Sci. Technol.
on Climate Change (United Nations Framework Convention on Climate 42, 14011407 (2008).
Change, 1997). 56. Hertwich, E.G. & Peters, G.P. Carbon footprint of nations: a global, trade-
24. Grasso, M. & Roberts, T. A compromise to break the climate impasse. linked analysis. Environ. Sci. Technol. 43, 64146420 (2009).
Nature Clim. Change 4, 543549 (2014). 57. Davis, S.J. & Caldeira, K. Consumption-based accounting of CO2 emissions.
25. EuropeanClimateFoundation Taking Stock The Emission Levels Implied by Proc. Natl Acad. Sci. USA 107, 56875692 (2010).
the Pledges to the Copenhagen Accord (European Climate Foundation, 2010). 58. Peters, G.P. etal. Rapid growth in CO2 emissions after the 20082009 global
26. Den Elzen, M.G. J. etal. The Copenhagen Accord: abatement costs and carbon financial crisis. Nature Clim. Change 2, 24 (2011).
prices resulting from the submissions. Environ. Sci. Policy 14, 2839 (2011). 59. Azar, C., Lindgren, K., Larson, E. & Mollersten, K. Carbon capture and storage
27. Rogelj, J. etal. Analysis of the Copenhagen Accord pledges and its global from fossil fuels and biomass costs and potential role in stabilizing the
climatic impacts a snapshot of dissonant ambitions. Environ. Res. Lett. atmosphere. Climatic Change 74, 4779 (2006).
5, 034013 (2010). 60. Van Vuuren, D.P. & Riahi, K. The relationship between short-term
28. Rogelj, J. etal. Copenhagen Accord pledges are paltry. Nature emissions and long-term concentration targets. Climatic Change
464, 11261128 (2010). 104, 793801 (2011).
29. UNEP The Emissions Gap Report 2013 (United Nations Environment 61. Fuss, S. et al. Betting on negative emissions. Nature Clim. Change
Program, 2013). 4, 850853 (2014).
30. Hardin, G. The tragedy of the commons. Science 162, 12431248 (1968). 62. Garnaut, R. Garnaut Climate Change Review: Final Report
31. Kolstad, C. etal. in IPCC Climate Change 2014: Mitigation of Climate Change (Cambridge Univ. Press, 2008).
(eds Edenhofer, O. etal.) Ch. 3 (Cambridge Univ. Press, 2014). 63. Garnaut, R. The Garnaut Review 2011: Australia in the Global Response to
32. Ostrom, E. Governing the Commons: The Evolution of Institutions for Collective Climate Change (Cambridge Univ. Press, 2011).
Action (Cambridge Univ. Press, 1990).
33. Dietz, T., Ostrom, E. & Stern, P.C. The struggle to govern the commons.
Science 302, 19071912 (2003). Acknowledgements
34. Lejano, R.P., Araral, E. & Araral, D. Interrogating the commons: introduction M.R.R. and J.G.C. acknowledge support from the Australian Climate Change
to the Special Issue. Environ. Sci. Policy 36, 17 (2014). Science Program of the Department of Environment, Australian Government.
35. Pretty, J. Social capital and the collective management of resources. Science G.P.P. and R.M.A. were supported by the Norwegian Research Council (236296).
302, 19121914 (2003). P.C. acknowledges support from the European Commissions 7th Framework
36. Myhre, G. et al. in IPCC Climate Change 2013: The Physical Science Basis Programme under Grant Agreements 603864 (HELIX) and the ERC Synergy Project
(eds Stocker, T.F. et al.) Ch. 8, 659740 (Cambridge Univ. Press, 2013). P-IMBALANCE. P.F. was supported by the European Commissions 7th Framework
37. Le Qur, C. etal. Global carbon budget 2013. Earth Syst. Sci. Data Programme under Grant Agreements 282672 (EMBRACE) and 603864 (HELIX). F.J.
6, 235263 (2014). was supported by the Australian Research Council (grant DP110102057). C.L.Q. was
38. Gasser, T. & Ciais, P. A theoretical framework for the net land-to-atmosphere supported by the UK Natural Environment Research Council (NERC)s International
Opportunities Fund (project NE/103002X/1) and EU/FP7 project GEOCarbon
CO2 flux and its implications in the definition of emissions from land-use
(283080). The authors are grateful to C. Wilson and H. Ransan-Cooper for insightful
change. Earth Syst. Dynam. 4, 171186 (2013).
comments. This work is a contribution to the Global Carbon Project
39. Shindell, D. & Faluvegi, G. Climate response to regional radiative forcing
(www.globalcarbonproject.org).
during the twentieth century. Nature Geosci. 2, 294300 (2009).
40. Berntsen, T.K. etal. Response of climate to regional emissions of ozone
precursors: sensitivities and warming potentials. Tellus B 57, 283304 (2005).
Author contributions
41. Trudinger, C.M. & Enting, I.G. Comparison of formalisms for attributing M.R.R. designed the study, carried out calculations and coordinated the conception and
responsibility for climate change: non-linearities in the Brazilian proposal writing of the paper. S.J.D. contributed data on committed emissions and drew figures.
approach. Climatic Change 68, 6799 (2005). G.P.P. and R.M.A. contributed data on committed emissions and resources. All authors
42. Davis, S.J., Caldeira, K. & Matthews, H.D. Future CO2 emissions and climate contributed to the writing of the paper.
change from existing energy infrastructure. Science 329, 13301333 (2010).
43. Davis, S.J. & Socolow, R.H. Commitment accounting of CO2 emissions.
Environ. Res. Lett. 9, 084018 (2014). Additional information
44. Davis, S.J., Peters, G.P. & Caldeira, K. The supply chain of CO2 emissions. Supplementary information is available in the online version of the paper. Reprints
Proc. Natl Acad. Sci. USA 108, 1855418559 (2011). and permissions information is available online at www.nature.com/reprints.
45. van Vuuren, D.P. etal. The representative concentration pathways: an overview. Correspondence and requests for materials should be addressed to M.R.R.
Climatic Change 109, 531 (2011).
46. Kriegler, E. etal. The role of technology for achieving climate policy objectives:
overview of the EMF 27 study on global technology and climate policy Competing financial interests
strategies. Climatic Change 123, 353367 (2014). The authors declare no competing financial interests.

538
This is Exhibit ".*" refer.red to in the

l't affidavit of Kirsten Zickfeld,
before me on the
of Aoril,2017.
A commissioner 1'or taking affidavits in

British Columbia
539
CANADAS CONTRIBUTION TO MEETING THE TEMPERATURE LIMITS IN THE PARIS

CLIMATE AGREEMENT
Dr. Simon Donner, University of British Columbia

Dr. Kirsten Zickfeld, Simon Fraser University
At the 2015 United Nations Climate Change Conference (COP 21) in Paris, the worlds
governments agreed to limit global warming to well below 2C above pre-industrial levels, and
recognized that avoiding 1.5C of warming would significantly reduce the risks and impacts of
climate change. Each of the parties to the United Nations Framework Convention on Climate
Change (UNFCCC), including Canada, are responsible for setting national emissions targets as
part of the global effort to avoid passing the stated temperature limits.
This brief report evaluates future CO2 emissions trajectories for Canada that are
consistent with the global temperature limits in the Paris Agreement, using a cumulative
emissions framework. The cumulative CO2 emissions over time can roughly predict the amount
of long-term warming of the climate system1,2. According to the Intergovernmental Panel on
Climate Change (IPCC), the ratio of global temperature change to total anthropogenic CO2
emissions over time is likely to be 0.8C to 2.5C per 1000 PgC (for emissions < 2000 PgC)3.
Here, a range of CO2 emissions trajectories for Canada which are consistent with the proposed
global temperature limits are estimated by scaling the global cumulative CO2 emissions budgets
reported by the IPCC4 to Canada, and combining them with historical Canadian emissions data.
Due to the cumulative carbon budget methodology, the trajectories do not explicitly
include emission of non-CO2 gases, which comprises 22% of greenhouse gas emissions from
Canada over the past decade. Though this analysis is based on the relationship between
temperature and cumulative CO2-only emissions, the computed percent reduction targets could
be applied to CO2 and non-CO2 gases.
KEY FINDINGS
1. Canadas portion of the global carbon budget is substantially lower if based on a

principle of international equity. The analysis assumes that the remaining carbon budget is
distributed according to Canadas fraction of the worlds CO2 emissions (1.6-1.8%, see
Methods), excluding land use change, land cover change, and forestry (referred to here as land
use). An equity-based carbon budget, determined based on Canadas fraction of the worlds
population, is one quarter or less of the size of an emissions-based budget.
Remaining Cumulative Canadian CO2 Emissions (Gt CO2) from 20161

Net warming <1.5C <2C <3C
Probability 66% 50% 33% 66% 50% 33% 66% 50% 33%
2
Emissions-based budget 3.9 6.6 11.8 14.4 19.7 23.2 38.9 45.9 53.7
2
Including land use 6.3 9.8 16.7 20.2 27.2 31.8 52.8 62.1 72.5
Equity-based budget3 0.7 1.4 2.9 3.6 5.1 6.0 10.4 12.4 14.6
1
Calculated following Table 2.2 in IPCC (2014), which assumes non-CO2 forcings follow RCP8.5.
Cumulative emissions are based on the time that the temperature limit is exceeded for the stated %
of model simulations in the CMIP5 archive
2
based on Canadas fraction of global CO2 emissions, using the last ten years of available data
3
based on Canadas fraction of worlds population, excluding land use
540
2. A carbon budget for

Canada with a likely
(66%) chance of
avoiding 1.5C of
warming globally is
extremely limited given
current emissions, even
with the more generous
emissions-based
budget. It would require a
90% to 99% reduction in
emissions below 2005
levels by 2030. The
budget is equivalent to
less than seven years of
emissions at current
(2013, year with most
recent data) levels
without efforts to expand
the budget via negative
emissions or international
credits.
3. The current Canadian target of a 30% reduction below 2005 levels by 2030 is
consistent with maintaining a likely chance of limiting warming to less than 2C globally,
only if the generous emissions-based budget is used. The emissions-based budget analysis
indicates that emissions must be 15% to 41% below 2005 levels by 2030 to remain consistent
with the 2C warming limit. A target of 29% below 2005 levels by the year 2030 is consistent
with the mean emissions pathway. Conversely, if an equity-based allocation is used, emissions
would need to be reduced to effectively zero before 2030 to be consistent with the 2C warming
limit.
4. CO2 emissions must drop to one-fifth of 2005 levels by 2050 to be consistent with
maintaining a likely chance of limiting warming to less than 2C globally in the
emissions-based case. The budget analysis indicates that emissions must be 73% to 91%
(mean of 81%) below 2005 levels by 2050 to remain consistent with the 2C warming limit. To
maintain a likely chance of avoiding 1.5C of warming globally, net CO2 emissions must decline
to effectively zero before 2050. To maintain the < 2C trajectory, emissions must be decreasing
at a rate of 2% to 7% per year in 2030, suggesting that measures to accelerate decarbonisation
of the economy would need to be in place before 2030.
5. The estimated remaining carbon budget for Canada could be increased through
negative emissions, including land carbon uptake or direct air capture technology,
through reducing non-CO2 greenhouse gas emissions, or through securing credits from
other nations. For example, to maintain a likely chance of limiting warming to less than 1.5C
globally, but with the larger domestic carbon emissions budget of a < 2C target, would require
9.8-10.5 Gt CO2, or 196-210 Mt CO2/year over fifty years, of negative emissions or credits
6. The size of Canadas remaining emissions budget consistent with avoiding the
different temperature limits could be sensitive to the treatment of land use. Net CO2
emissions from land use (land use change, land cover change, and forestry) are highly
541
variable from year to year nationally and globally due to climate and management. If an
emissions-based approach is used and land use is included in this analysis, the remaining
budget increases by 0% to 62% and 0% to 43% in the 1.5C and 2C cases respectively,
depending only on the choice of baseline year. To avoid this uncertainty, future carbon uptake
by land may best be treated as a means to expand the remaining carbon budget (finding #5)
rather than as a part of the historical budget. This approach is consistent with Canadas
Intended Nationally Determined Contribution which includes land use as a possible mitigation
measure but excludes land use from the historical emissions data.
METHODS
The Canadian cumulative CO2 emissions budgets consistent with a likely (66%) chance of
avoiding 1.5C, 2C, and 3C of warming globally were computed as the fraction of the global
cumulative CO2 emissions from 2011 in Gt CO2 from complex models, reported on Table 2.2,
IPCC (2014). The fraction is based on Canadas percent of the worlds CO2 emissions over the
last ten years as reported in the Global Carbon Project5 (national emissions) and Canadas
2015 National Inventory Report. The budgets remaining from 2016 onwards in Canada, as
reported in the Table above, were estimated by subtracting emissions for the past five years. To
assess the sensitivity of the budget to the assumed baselines, the fraction was also calculated
using the last five years of data. Emissions since 2013, the last year reported in the National
Inventory Report, were assumed to be stable. For comparison, the Canadian cumulative
emissions budgets were also calculated on the basis of Canadas proportion of the worlds
population, using 2014 population estimates from Statistics Canada and the CIA World
Factbook.
The emissions trajectories were computed using logistic curves, as in other studies of future
emissions trajectories6. The one key difference is that other studies compute emissions
trajectories with an assigned rate of increase until the year mitigation begins5; here it is
assumed, based on the past few years, that Canadas emissions will stay level or decline in the
future.
The multiple possible trajectories consistent with each emissions budget were computed by
varying the midpoint year (2030, 2070, 2100, 2200) and the year the emissions decline (lag of 0,
2, 6, 8, 10 years from 2016) in the logistic equation and solving for k7. This method created 56
possible trajectories for the < 2C and < 3C; only 20 trajectories are presented for < 1.5C,
because it is not possible for Canada to stay within the 1.5C budget with six or more years of
emissions at present levels.
The entire analysis was conducted twice, once excluding land use and once including land use
in national and global historical emissions data. Canadas fraction of the worlds emissions
including land use varies from 1.6% to 3.2% over the past ten years.
1
Zickfeld K et al. (2009) PNAS 106, 16129-16134
2
Matthews HD et al. (2009) Nature 459, 829-832
3
IPCC (2014) Climate Change 2014: Synthesis Report
4
Table 2.2, IPCC (2014)
5
Data available at http://www.globalcarbonproject.org/carbonbudget/15/data.htm
6
Raupach MR et al. (2011) Tellus B 63, 145-164
7
F (yr) = Fpeak / (1+ e^[k*(yr - yrmidpoint)]
542
This is the 1st affidavit of

Maximilian Kniewasser
made April 27, 2017
FEDERAL COURT
BETWEEN:
Applicant
and

MINISTER OF ENVIRONMENT AND CLIMATE CHANGE,
CANADIAN ENVIRONMENTAL ASSESSMENT AGENCY, and
PACIFIC NORTHWEST LNG LIMITED PARTNERSHIP
Respondents
APPLICATION UNDER SECTIONS 18 & 18.1 OF THE FEDERAL COURTS ACT,

R.S.C. 1985, c. F-7
AFFIDAVIT
I, Maximilian Kniewasser, Analyst at the Pembina Institute, of 2-2153 Timber Ridge,

Whistler BC, AFFIRM THAT:
1. I am a policy and technical analyst who has reviewed the Canadian

Environmental Assessment Agencys environmental assessment report
regarding the Pacific NorthWest LNG Project (the CEAA Report), and as
such I have personal knowledge of the facts and matters hereinafter deposed to
save and except where the same are stated to be based upon information or
belief, and where so stated I verily believe the same to be true.
2. Mr. Anthony Ho, counsel to SkeenaWild Conservation Trust in this matter, has
provided me with a selection of documents that he tells me, and I verily believe
to be true, are some of the documents that have been disclosed by the Crown
pursuant to Rule 318. I have reviewed these documents. Where I refer to such
543
2
documents in this affidavit, I have identified them using the identification

number that the Crown uses for these documents (R318 ID#).
Qualification
3. I have been retained by SkeenaWild Conservation Trust as an expert in these

proceedings. The retainer between SkeenaWild Conservation Trust and myself
is attached to this affidavit as Exhibit A.
4. I am an expert in the following areas:
a. Domestic and international climate policy;
b. Impact of climate policy on the economic viability of major fossil fuel

projects, including climate-energy modelling;
c. Modelling of climate impacts from shale and liquefied natural gas

(LNG) developments; and,
d. Technical review of the environmental impacts from, and alternatives to,

major fossil fuel projects.
5. I am a co-creator of the B.C. Shale Tool (2015), which is a scenario-planning

tool for shale gas and LNG development in B.C. that quantifies environmental
impacts under different development, technology and policy scenarios. 1
6. This tool was extensively used by Environment and Climate Change Canada to
assess the upstream greenhouse gas impacts from the Woodfibre LNG project
and the Pacific NorthWest LNG project, which informed the projects
respective environmental assessments. 2,3
1
Pembina Institute, B.C. Shale Scenario Tool, 2015. http://www.pembina.org/pub/BCShaleTool
2
CEAA, Pacific NorthWest LNG Project Review of Related Upstream Greenhouse Gas Emissions
Estimates, 2016. http://www.ceaa.gc.ca/050/documents/p80032/115673E.pdf
544
3
7. My current curriculum vitae is attached to this affidavit as Exhibit B.
Expert Opinion
8. You have asked me to provide my opinion to the following: Please offer your
opinion on the CEA Agencys assessment of alternative means of powering
the Project as set out in CEAA Report.
9. Please find my response to this question below. My opinion is informed by my

experience in working on climate policy in British Columbia and Canada,
climate-energy modelling, and the assessment of greenhouse gas emissions of
LNG and upstream shale gas development under different development,
technology and policy scenarios.
A. Background information
10. Before I provide my opinion, below I set out some background information that
is relevant to understanding my opinion on the Agencys assessment of
alternative means of powering the Pacific NorthWest LNG project (the Project).
11. The liquefaction of natural gas at liquefied natural gas (LNG) facilities is an
energy and emissions intensive process. However, there are several different
ways to power LNG facilities, the choice of which can have significant
implications on the Greenhouse Gas (GHG) emissions of a project. The
Proponent is planning to power the Project almost entirely with on-site natural
gas turbines. Less than 1% of the total power requirements of the Project are
expected to be serviced with grid electricity. 4
3
CEAA, Woodfibre Liquefied Natural Gas (LNG) Project Review of Related Upstream Greenhouse
Gas (GHG) Emissions Estimates, 2016. http://www.ceaa.gc.ca/050/documents/p80060/104688E.pdf
4
KPMG states that for two trains (Phase 1) the power requirement for non-critical load is expected to
be less than 5 MW. For three trains (Phase 2), this is less than 7.5 MW. Total power needs for the
545
4
12. There are two main alternatives of using grid electricity to power LNG projects
and improve their GHG emissions performance, compared to using on-site
natural gas turbines. The first alternative is to use grid electricity to power non-
compression load. Non-compression load describes the power needed for
pumps, air coolers, lighting and space heating at the LNG facility. This load
accounts for approximately 25-30% of the total power needs from an LNG
facility. 5 The majority of non-compression load is critical for the LNG
operations and therefore requires stable and reliable power. The second
alternative is to use grid electricity to power the compression process. In this
alternative, electric motors drive the main liquefaction process. 6 Compression
load accounts for approximately 70-75% of the total power needs from an LNG
facility. The compression process also is critical to LNG operations, and
therefore requires stable and reliable power. In the CEAA Report on the
Project, the Agency only assessed the first alternative described above (using
grid electricity to meet non-compression load). 7 The Agency did not discuss the
second alternative (using grid electricity for compression load). The conditions
imposed on the Project would not require either of these alternative methods of
powering the Project to be implemented.
13. The Project is expected to proceed in two phases. Phase 1 consists of two
liquefaction trains, each with a capacity of 6.84 million tonnes of LNG per year
Project are expected at around 1,100 MW. As such, non-critical load represents less than 1% of total
energy needs. The Proponent has ruled out using grid electricity for critical load, but may use it to
meet non-critical load. KPMG, Pacific NorthWest LNG Limited Partnership Independent review of
power options and selection process, 2014,
http://www.pacificnorthwestlng.com/media/PNW_Partnership%20report_v.6.1(1).pdf
5
This assumes that the power capacity for compression load is between ~40-45 MW per mtpa of LNG
capacity, and that total load for both non-compression and compression load is ~55 MW per mtpa of
LNG capacity. The exact share of power required by the two loads is difficult to pinpoint precisely as
they are interrelated (i.e., waste heat from the compression process can be used to power the non-
compression process). KPMG, Pacific NorthWest LNG Limited Partnership Independent review of
power options and selection process, 2014,
http://www.pacificnorthwestlng.com/media/PNW_Partnership%20report_v.6.1(1).pdf
6
In circumstances in which a project uses grid electricity for the compression process, the project
would likely also service the non-compression load with grid electricity.
7
CEAA Report, Pacific NorthWest LNG Project Environmental Assessment Report, 2016.
http://www.ceaa.gc.ca/050/documents/p80032/115668E.pdf
546
5
(mtpa), while Phase 2 would see the addition of a third train at a later date. 8 As
such, the LNG export capacity of the Project would be 13.7 mtpa in Phase 1,
and would increase to 20.5 mtpa in Phase 2. Due to the different timelines, the
two phases offer different opportunities to implement alternative power supply
options as the extra time could allow for technology improvements and electric
infrastructure upgrades. Table 1 shows five potential project configurations and
the resulting GHG emissions, with Scenario 1 representing the project as
approved based on conditions in the CEAA Decision Statement. 9
Table 1: Overview of alternate project power scenarios

Scenario Phase 1 (13.6 Phase 2 (6.8 GHG Improvement in Cumulative GHG
mtpa) mtpa) emissions GHG emissions emissions
from the performance over savings over 30
Project (Mt Scenario 1 (%) year operating
CO2e) life (Mt CO2e)
1 As per Project As per Project
4.3 0% 0.0
conditions conditions
2 As per Project Grid power for
conditions non-
4.0 6% 8.2
compression
load
3 As per Project Grid power for
conditions non-
compression
3.4 22% 27.9
and
compression
load
4 Grid power for Grid power for
non- non-
3.1 29% 36.9
compression compression
load load
5 Grid power for Grid power for
non- non-
compression compression
2.4 44% 56.6
load and
compression
load
Note: As per Project conditions refers to the Ministers conditions in the CEAA Decision
Statement. These conditions include a GHG emissions intensity limit of 0.21 t-CO2e/t-LNG for
Phase 2 (Trains 1,2,3) of the Project. This is seen as a best-case scenario, as if only Phase 1
proceeds, the emissions intensity will be higher at 0.22 t-CO2e/t-LNG. Each train is assumed to
have a capacity of 6.84 mtpa. This results in annual production in 20.5 mtpa, and total
8
9
CEAA, Decision Statement for the Pacific NorthWest LNG Project, 2016.
547
6
emissions of 4.31 Mt CO2e, almost matching the emissions cap set by the Project conditions.
The alternative technologies are based on the stated emissions intensity of projects that plan to
employ these alternative technologies in B.C. Grid power for non-compression load is based on
the stated emissions intensity of LNG Canada of 0.15 t-CO2e/t-LNG 10, and the grid power for
non-compression and compression load is based on the emissions intensity of the Woodfibre
LNG project of 0.054 t-CO2e/t-LNG. 11
14. As can be seen in Table 1, it can make a material difference to GHG emissions
performance whether and when the Project deploys grid power for either non-
compression load, or for both non-compression and compression loads,
especially over the lifecycle of the Project. Scenarios two to five could reduce
emissions by 6% to 44% compared to the emissions cap set by the Minister.
Over the expected 30-year life span, this could reduce cumulative emissions
from the Project by between 8 and 57 Mt CO2e. 12 Other LNG projects proposed
for B.C. have committed to using these two alternative grid-based power
options (see paragraphs 15 to 18). 13
15. The Project went through the environmental assessment process in parallel with
two other LNG projects, the LNG Canada project and the Woodfibre LNG
project. These two comparator projects received their respective provincial
environmental certificates and federal environmental assessment approvals
before the Project. 14,15
16. LNG Canada is planning to use grid electricity for non-compression load for
both planned phases (similar to Scenario 4 in Table 1). 16 It is a project of
10
BCEAO, LNG Canada Export Terminal Project Assessment Report, 2015. Pg. 61.
https://www.ceaa-acee.gc.ca/050/documents/p80038/101852E.pdf
11
CEAA, Woodfibre Liquefied Natural Gas (LNG) Project Review of Related Upstream Greenhouse
Gas (GHG) Emissions Estimates, 2016. http://www.ceaa.gc.ca/050/documents/p80060/104688E.pdf
12
The Project is expected to be operational, and produce at full capacity, for over 30 years. CEAA
Report, Pacific NorthWest LNG Project Environmental Assessment Report, 2016.
http://www.ceaa.gc.ca/050/documents/p80032/115668E.pdf EA report
13
The two projects that are using the two alternative means of powering the project are the LNG
Canada project, which is planning to use grid power to meet non-compression load, and the Woodfibre
LNG project which is planning to use grid power to meet both non-compression and compression load.
These two projects are the two main comparator projects to the Project.
14
CEAA, Decision Statement to LNG Canada, 2015. http://www.ceaa.gc.ca/050/document-
eng.cfm?document=101851
15
CEAA, Decision Statement to Woodfibre LNG, 2016. http://www.ceaa.gc.ca/050/document-
16
BCEAO, LNG Canada Export Terminal Project Assessment Report, 2015. https://www.ceaa-
acee.gc.ca/050/documents/p80038/101852E.pdf
548
7
similar size (LNG Canada will have a capacity between 24 mtpa and 26 mtpa)
and timeline compared to the Project and is located in nearby Kitimat.
17. The Woodfibre LNG project is a smaller project (2.1 mtpa) located close to
Squamish. It is planning to, and is required to, use grid electricity for both the
compression and non-compression loads (or a similar technology with equal or
better GHG emissions performance). 17 It is expected to have an earlier start up
date than the Project. 18
18. Considering that both the LNG Canada and Woodfibre LNG project received
their environmental certificates before the Project suggests that there was
sufficient time to assess the feasibility of the two grid powered alternatives for
powering the Project.
19. Alternatives for powering the Project were explored by the Proponent and by
the Agency. Furthermore, Natural Resources Canada (NRCan) undertook
economic analysis of using grid electricity to power non-compression load 19
and compression load 20, and provided these analyses to the Agency. The
Agency did not refer to the NRCan economic modelling in the CEAA Report.
20. In draft conditions for the Project, the Agency had described requirements for
the Proponent to periodically assess the feasibility of using grid electricity to
power the Project, and to implement these alternatives should they become
feasible. 21 Draft Condition 3.2 would have required the Proponent to assess
every five years if electrical power was available to power non-compression
load for the full Project, and for compression load for Phase 2 of the Project,
and to implement these alternatives if they were found to be technically and
17
18
Woodfibre LNG, Parent company authorizes Woodfibre LNG to proceed with project, 2016.
https://www.woodfibrelng.ca/parent-company-authorizes-woodfibre-lng-to-proceed-with-project/
19
R318 ID# EA06536. A copy of this document is attached to this affidavit as Exhibit C.
20
R318 ID# EA06534. A copy of this document is attached to this affidavit as Exhibit D.
21
R318 ID# EA02909. A copy of this document is attached to this affidavit as Exhibit E.
549
8
economically feasible for the Project. This draft condition does not appear in
the final set of conditions for the Project. 22
21. Below I offer my opinion on the Agencys assessment of alternative ways of

powering the project. First, I describe my opinion on the Agencys assessment
of using grid electricity for non-compression load. Second, I offer my opinion
on the Agencys assessment of using grid electricity for compression load.
B. Opinion on the Agencys assessment of the alternative of using grid electricity

for non-compression load
22. In my opinion the Agency failed to assess the economic feasibility of using grid
electricity for non-compression load. I explain my reasoning below.
23. In the CEAA Report, the Agency assessed the technical feasibility of using grid
electricity for non-compression load, and found that this alternative is
technically feasible in the timeline proposed for the Project. In the CEAA
Report, the Agency states that electrification of non-compression load would
require around 215 MW of stable power. 23,24 Furthermore, the CEAA Report
states that BC Hydro provided information to the Agency that there is sufficient
generation and transmission capacity to serve approximately 200 to 250 MW of
new industrial load in the Port of Prince Rupert lands. 25 The Agency
acknowledged that infrastructure upgrades would be required to deliver the
stable electricity needed for LNG operations. However, it also acknowledged
that the LNG Canada and Woodfibre LNG projects are pursuing such
22
CEAA, Decision Statement for the Pacific NorthWest LNG Project, 2016.
23
CEAA Report, Pacific NorthWest LNG Project Environmental Assessment Report, 2016. Pg. 42.
24
R318 ID# EA02788. A copy of this document is attached to this affidavit as Exhibit F.
25
550
9
infrastructure upgrades in order to use grid power for their projects. 26 Given
this information, the Agency concluded:
while infrastructure improvements would be required in

order to bring reliable grid power to the Project site, these
upgrades and changes are manageable within the planning
horizon for this Project. As a result, the Agency is of the
view that using grid power for non-compression needs at
the Project site would be technically feasible. 27
24. In my opinion, the Agency failed to carry out an assessment of the economic
feasibility of using grid electricity for non-compression load. The Proponent
stated that it never undertook a cost analysis for using grid electricity for non-
compression load. 28 Nor was such a cost analysis undertaken by the Agency. As
such, the Agency did not assess the economic feasibility of using grid
electricity for non-compression load in the CEAA Report. This is surprising
considering that the Agency had determined that this alternative was technically
feasible for the Project. 29
25. Even though the Agency did not undertake an assessment of the economic
feasibility of using grid electricity for non-compression load in the CEAA
Report, NRCan provided the Agency with its views on this very matter. 30 This
took the form of an economic analysis that compared using grid electricity for
non-compression load versus generating power inside the fenceline using
natural gas turbines, as proposed by the Proponent. The analysis included
assumptions designed to make the modelling relevant to the Project, including
the Proponent being responsible for paying for electric infrastructure upgrades
(estimated at $1 billion), and deploying four 55 MW backup generators. Based
on this NRCan analysis, the Agency stated in internal documents that:
26
27
CEAA Report, Pacific NorthWest LNG Project Environmental Assessment Report, 2016, Pg. 42.
28
29
See block quote in previous paragraph
30
551
10
Based on the analysis conducted by Natural Resources

Canada, the Agency is of the view that it would be
economically feasible for the Proponent to utilize grid
power for the non-compression needs of the Project once it
reaches full build out or by 2030. 31
The final CEAA Report makes no mention of NRCans analysis nor of the
Agencys apparent conclusion, based on the NRCan report that it would be
economically feasible to use grid power by 2030 or upon full build out
26. Nor does the CEAA Report include any analysis of the relevance of the two
main comparator projects described earlier (LNG Canada and Woodfibre LNG)
to an assessment of power generation alternatives. In my opinion, the Agency
could have done a direct comparison to the LNG Canada project considering
the many similarities between LNG Canada and the Project. LNG Canada is a
project of comparable size and timeline to the Project, located in nearby
Kitimat. LNG Canada is planning to use grid electricity to meet non-
compression load for both phases of its project. This will allow LNG Canada to
achieve GHG emissions intensity that is 29-32% lower compared to the
Project. 32 The Agency acknowledged that LNG Canada would have to update
electric infrastructure to implement the use of grid power for non-compression
load, similar to what the Agency acknowledged would be required by the
Proponent were it to implement this alternative. 33 Despite these similarities, the
Agency does not compare LNG Canada and the Project in regards to using grid
electricity for non-compression load. This failure is surprising, given that these
two LNG projects offer many similarities and considering that LNG Canada is
committing to use grid electricity for non-compression load, the only power
alternative the Agency assessed for the Project in the CEAA Report.
31
32
Based on LNG Canada having an emissions intensity of 0.15 t-CO2e/t-LNG produced (see note of
Table 1). In comparison, the PNW LNG Project, according to the Project conditions, must achieve an
emissions intensity of 0.22t-CO2e/t-LNG (Phase 1) and 0.21 t-CO2e/t-LNG (Phase 2).
33
552
11
C. Opinion on the Agencys assessment of the alternative of using grid electricity

for compression load
27. For the reasons that follow, it is my opinion that the Agency failed to consider
the feasibility of grid electricity for compression load for Phase 2 of the Project.
28. The Project is planned to proceed in two separate phases. The initial phase
(Phase 1) will see the installation of two LNG trains with a capacity of 6.84
mtpa of LNG each for Phase 1 (total of 13.7 mtpa). The Proponent has stated
that it plans to achieve an aggressive timeline for Phase 1 of the Project. 34,35
As a result of this aggressive timeline, the Proponent moved quickly to the
Front-End Engineering and Design (FEED) phase, in which the Proponent
instructed all prospective engineering contractors to design the Project to be
powered almost entirely with natural gas. 36,37 A subsequent phase (Phase 2)
will add one additional LNG train to the Project, to meet growing LNG demand
in the future. I am not aware of a timeline for implementing Phase 2 of the
Project. The later start-up date for Phase 2 could allow for additional
opportunities to include alternative power options into the Project due to
expected future technology improvements and electricity infrastructure
upgrades. This could include using grid electricity to power the compression
process of Phase 2 of the Project.
29. The Proponent did assess the technical feasibility of using grid electricity for
compression load. 38 Its study concluded that BC Hydro would not have enough
34
PNW LNG, Environmental Impact Statement and Environmental Assessment Certificate
Application, (2014). Page 2-34. http://www.ceaa.gc.ca/050/documents/p80032/98676E.pdf
35
36
The Front-End Engineering and Design (FEED) process requires detailed specifications of the
Project, including power supply options for non-compression load. In 2012, the Proponent instructed
all Engineering Procurement providers competing in the FEED process to use natural gas turbines for
both compression and non-compression load. The start of the 2012 FEED study was before the
Proponents 2013 Feasibility study was published that showed that sufficient electricity would not be
available in time to achieve the Proponents aggressive schedule, and therefore grid powered
alternatives were not considered viable.
37
R318 ID# EA02781. A copy of this document is attached to this affidavit as Exhibit G.
38
Application, (2014). Pg. 2-31. http://www.ceaa.gc.ca/050/documents/p80032/98676E.pdf
553
12
reliable electricity available in the timeline specified for the Project. I

understand that this timeline refers to the aggressive timeline for Phase 1 of
the Project. 39,40 It is my understanding that no study of the technical feasibility
of using grid electricity for compression load was done for Phase 2 of the
Project.
30. The Proponent acknowledged that opportunities exist to include grid electricity
into Phase 2 of the Project to lower the Projects GHG emissions, including
possibilities for using electricity to power compression load. In a letter to the
Tsimshian Nation Chiefs, the Proponent stated that it looks forward to
incorporating renewable power in future expansion plans (i.e., Phase 2) by,
amongst others, assessing the technical feasibility and viability for
compression processes to utilize electrical power from renewables should they
become available. 41
31. In the CEAA Report, the Agency does not describe or assess the alternative
power option of using grid electricity for compression load. 42 Furthermore, in
the report, the Agency does not discuss opportunities to improve the GHG
performance of Phase 2 of the Project. In particular, the report does not discuss
the technical or economic feasibility of using grid electricity for compression
load for Phase 2 of the Project. This is surprising, given that the Proponent
highlighted that the short timeline was a major obstacle to using grid power for
compression load. 43
32. However, during the Agencys review of the Project, NRCan provided the
Agency with an analysis that assessed the economic feasibility of using grid
electricity for compression load for LNG Projects. The analysis shows that
39
Application, (2014). Pg. 2-34. http://www.ceaa.gc.ca/050/documents/p80032/98676E.pdf
40
41
R318 ID# EA02781. A copy of this document is attached to this affidavit as Exhibit G.
42
43
Application, 2014. Pg. 2-34. http://www.ceaa.gc.ca/050/documents/p80032/98676E.pdf
554
13
using grid electricity for compression load has a 2% higher internal rate of
return 44 compared to generating power on site, as currently proposed by the
Proponent. 45 In the CEAA Report, the Agency does not mention this study.
33. Finally, the CEAA Report does not include in its alternatives assessment a
comparison between the Project and other relevant comparators, including the
Woodfibre LNG project. Woodfibre LNG will use grid electricity for
compression load (see paragraph 17). 46 It also went through the environmental
assessment process at the same time as the Project. Woodfibre LNG is planning
to have an earlier start up date than Phase 1 of the Project (and of course much
earlier than Phase 2 of the Project). 47 The Agencys failure to consider a
comparison to the Woodfibre LNG project is surprising given that Woodfibre
will use grid electricity for compression load within a shorter timeframe
compared to the Project, while the Projects aggressive timeline was seen as a
major obstacle to the technical feasibility of this alternative power option.
D. Conclusion
34. Above I described two alternatives to power LNG projects other than burning
natural gas: using grid electricity to power non-compression load, and using
grid electricity to power compression load. These two alternatives will be used
in other LNG projects in B.C., notably the LNG Canada project and the
Woodfibre LNG project, the two other most advanced LNG project proposals
for British Columbia. These alternatives offer material GHG emissions benefits.
Specifically, my analysis (see Table 1) shows that the alternatives could reduce
44
Internal rate of return (IRR) is a metric used in capital budgeting measuring the profitability of
potential investments. Internal rate of return is a discount rate that makes the net present value
(NPV) of all cash flows from a particular project equal to zero. Generally speaking, the higher a
project's internal rate of return, the more desirable it is to undertake the project. Investopedia, Internal
Rate of Return - IRR, http://www.investopedia.com/terms/i/irr.asp
45
R318 ID# EA06534. A copy of this document is attached to this affidavit as Exhibit D.
46
47
Woodfibre LNG, Parent company authorizes Woodfibre LNG to proceed with project, 2016.
https://www.woodfibrelng.ca/parent-company-authorizes-woodfibre-lng-to-proceed-with-project/
555
14
GHG emissions from the Project by between 6% and 44% compared to the
Project as approved. 48 Over its 30-year lifespan, using these alternatives could
reduce the Projects cumulative emissions by between 8 and 57 Mt CO2e. The
amount of GHG emissions reduction depends on which alternatives are used
and when they are implemented (see Scenario 2 to 5 in Table 1). These
potential emissions reductions are especially significant given the Projects long
operating life, B.C.s legislated long-term climate targets, and Canadas Paris
climate commitments.
35. One way to ensure that the Proponent continues to consider alternative means
of powering the project is to require the Proponent to revisit the technical and
economic feasibility of the two alternatives on a periodic basis. This
requirement was included in draft conditions prepared by the Agency. 49 Draft
condition 3.2 stated that:
The Proponent shall assess, every five years, if electrical

power is available to power components of the Designated
Project, including the compression requirements of Train
3, and shall report the results of that assessment to the
Agency. In the event that electrical power becomes
available, the Proponent shall use it to power non-
compression requirements of the Designated Project and
compression requirements of Train 3, unless the Proponent
can demonstrate to the Agency that it is not technically or
economically feasible to do so.
48
This is seen as a best-case scenario for the Proponent, as it assumes the emissions intensity of 0.21 t-
CO2e/t-LNG. If Phase 2 is substantially delayed the emissions effective average emissions intensity
over the 30 year life of the project would be higher, and the cumulative emissions benefits from
alternative technologies would also be higher.
49
R318 ID# EA02909. A copy of this document is attached to this affidavit as Exhibit E.
556
15
36. According to documents that I have seen, this draft condition was last included
in draft conditions dated August 23,2016.s0 However, this condition does not
appear in the final CEAA Decision Statement on the Project published on
September 27,2016.s1
AFFIRMED BEFORE ME )
in the City of Vancouver, )
Province of British Columbia )
onApril27,2017. )
)
)
)
)
A commissioner for ) Maximilian Kniewasser
for British Columbia )
)
Pak Yan Anthony HO )
Barrister & Solicitor )
Suite 16 Shoal Point )
21 Dallas Road )
Victoria, BC Y8Y 429 )
Tel: (778) 678-3818 )
Email : anho @pacifi ccell. ca )
50
R3l8 ID# EA02909. A copy of this document is attached to this affidavit as Exhibit ,,E',.
sr
CEAA, Decision Statementfor the PaciJic NorthWest LNG Project,2016.
http://www.ceaa. gc.calO50/documents/p 800321 ll 5669E.pdf
557
April 19,2017 Our file: 2016-09-0002

A
VIA ELECTRONIC MAIL This is Exhibit "Il[." to
1't affidavit of
affirmed before me on the
day of Anril, 2017.
Analyst
Suite 610, 55 Water Street
Vancouver, BC V6B 1Al
max im iliank@pembina.org
Dear Mr. Kniewasser:
FfEz Skeenawild Conservation Trust v. Her Majesty The Queen et al.

Expert Retainer Letter
We are counsel to SkeenaWild Conservation Trust ("SkeenaWild") in the above-referenced legal

proceeding before the Federal Court.
We confirm that you have agreed to provide an affidavit containing your expert opinion for this
proceeding. We are writing to set out the questions that we would like you to address in your
affidavit.
Material Facts
SkeenaWild is a charitable purpose trust whose goal is to make the Skeena River watershed and
nearby coastal communities a global model of sustainability. A significant portion of its work
involves the protection and conservation of salmon and salmon habitats within the Skeena River
watershed.
Pacific NorthWest LNG Limited Partnership (the "Proponent") is proposing to construct the
Pacific NorthWest LNG Project (the "Project"), which consist of a liquefied natural gas ("LNG")
liquefaction facility and export terminal on Lelu Island at the mouth of the Skeena River,
approximately 15 km southwest of the Prince Rupert, BC. The purpose of the Project is to
convert processed natural gas from Progress Energy Canada Ltd. reserves into LNG for export to
the Pacific Rim markets in Asia.
The Project was subject to an environmental assessment under the Canadian Environmental
AssessmentAct,20l2,S.C.2012, c. 19, s. 52(the "Act"). On September27,2016,the following
occurred:
l. The Canadian Environmental Assessment Agency (the "CEA Agency"), which was the
responsible authority for the environmental assessment of the Project under the Act,
released its final environmental assessment report (the "Report") pursuant to subsection
Suite 16 Shoal Point,21 Dallas Rd., Victoria BC, V8V 4Zg
(250l 721-8170 | http://www.pacificcell.ca I info@pacificcell.ca
page 1 of 4
558
25(2) of the Act. In the Report, the CEA Agency concluded that the Project is likely to
cause significant adverse environmental effects as a result of greenhouse gas (GHG)
emissions.
2. The Minister of Environment and Climate Change (the Minister) released a decision
statement pursuant to section 54 of the Act. Among other things, this decision statement
contained the following:
a. The Ministers decision under subsection 52(1) of the Act that the Project is likely to
cause significant adverse environmental effects, after considering the Report and the
implementation of mitigation measures that she had considered appropriate; and,
b. The Ministers decision under subsection 52(2) of the Act to refer to the Governor in
Council the matter of whether those significant adverse environmental effects were
justified in the circumstances.
3. The Governor in Council issued Order in Council P.C. 2016-0838 pursuant to subsection
52(4) of the Act, in which the Governor in Council determined that the significant adverse
environmental effects associated with the Project are justified in the circumstances.
On October 27, 2016, SkeenaWild filed a notice of application for judicial review at the Federal
Court challenging the Report, the Ministers decisions, and the Order in Council.
Relevant Documents
In connection with your report, we will provide copies of the following documents, which are
expected to be introduced as evidence in these proceedings:
1. The Report dated September 27, 2016; and,

2. A document by Environment and Climate Change Canada entitled Review of Related
Upstream Greenhouse Gas Emissions Estimates dated September 27, 2016, which was
submitted to the CEA Agency.
We may provide you with additional documents and other materials to review.
Code of Conduct for Expert Witnesses*****
As an expert witness, you owe a duty to the Court that overrides any duty to a party to these
proceedings. These duties are set out in the Federal Courts Code of Conduct for Expert
Witnesses. You must review this document, which is available here: http://www.laws-
lois.justice.gc.ca/eng/regulations/SOR-98-106/page-127.html#h-254.
We will ask you to accompany your report with Form 52.2, in which you will certify that you
have read the Code of Conduct for Expert Witnesses and agree to be bound by it.

Page 2 of 4
559
Questions
Based on the facts set out above, and your review of the documents we have provided and other
materials, please provide your professional opinion on the following:
1. Please offer your opinion on the CEA Agencys assessment of alternative means of
powering the Project as set out in Report.
In preparing your expert opinion, you should rely on any source that you consider to be reliable
in support of your opinion and on your own knowledge and experience.
Please feel free to offer any addition expert opinion beyond answers to the above questions that
you believe is necessary to provide a full and comprehensive expert opinion.
Form of Affidavit
In your affidavit, please:
1. State your full name and address;

2. Describe your areas of expertise and qualifications in relation to the issues addressed in
your affidavit;
3. State the facts and assumptions on which your opinions are based;
4. Provide your answers and opinions to the questions set out above, and your reasons for
those answers and opinions;
5. List any literature or other materials specifically relied on in support of your answers and
opinions;
6. If applicable, describe the methodology that you used in providing your answers and
opinions, including any examinations, tests or other investigations on which you have
relied, including (if applicable) details of the qualifications of the person who carried them
out, and whether a representative of any other party was present;
7. State any caveats or qualifications necessary to render your expert opinion complete and
accurate, including those relating to any insufficiency of data or research and an indication
of any matters that fall outside your field of expertise;
8. State, if applicable, the particulars of any aspect of your relationship with a party to the
proceeding or the subject matter of your proposed evidence that might affect your duty to
the Court;
9. Attach, as Exhibit A to your affidavit, this retainer letter;
10. Attach, as Exhibit B to your report, your most up-to-date curriculum vitae; and,
11. Attach, in successive Exhibits to your affidavit as needed, any other material you deem
relevant or necessary to render your expert opinion complete and accurate.

Page 3 of 4
560
Timeline
You must complete your report by April 28, 2017.
Privilege and Confidentiality
Please be advised that all communications between us, including this letter, are confidential and
privileged. However, if we introduce into evidence any report that you prepare, privilege is
waived. At that time, all correspondence between us, and any drafts of reports and related notes
may become available to other parties in the proceedings.
You understand that your work product may, at our discretion on the instructions of our client, be
shared with common interest parties, any other intervenors, and their legal representatives who
may ultimately be granted standing to join in these proceedings.
Yours truly,
_________________________ _________________________
Chris Tollefson Anthony Ho
Barrister & Solicitor Barrister & Solicitor
Counsel for the applicant Counsel for the applicant
Tel: (250) 888-6074 Tel: (778) 678-3818
Email: ctollefson@pacificcell.ca Email: anho@pacificcell.ca

Page 4 of 4
This is Exhibit ".Y." referred to in the 561
I't affidavit of Maximilian Kniewasse
ay of April, 2017.
Curriculum Vitae
affidavits in
British Columbia
2-2153 Timber Ridge, Whistler Bc, VON.1B2 Phone Number: (G04) 908 4164
maxrmiliank@pembina. org
I Graduate Studies
201L- 20L4 - MRM (REM), Simon Fraser University, Burnaby BC
o Master's thesis in Resource and Environmental Management under environmental economist
Mark Jaccard, studying Canada's energy and climate policies using an energy-economy model;
o Assess effectiveness and economic impact of current energy and climate policies;
o Assess economic impact of domestic and global climate policy on the fossil fuel industry;
o Completed graduate courses on energy policy/economics and energy-economy modelling;
o Completed courses in stakeholder engagement/conflict resolution.
Undergraduate Education
2005 - 2009 B.Sc (Geography and Political Science), McGill University, Montreal QC
WOfk EXpefienC - tn the energy and energy-poticy fietd
2014 - Current - Pembina lnstitute -Analyst, Voncouver, BC

o Policy and Technical Analyst at Canada's leading clean energy think tank;
o Conduct high quality research and write/publish reports under tight time constraint;
o Build Excel based analytical models to assess environmental and economic impacts of energy
development and policy;
o Main developer of the B.C. Shale Tool, a scenario analysis tool that assesses the
impacts of different shale gas scenarios. The Tool was extensively used by ECCC to
inform the environmental assessments of major LNG projects;
o Main researcher for the lnstitute's shale gas, LNG, and methane files;
o Project manager for a project with an annual budget of 5200,000;
o Strategize on how to make projects as influential as possible with limited resources;
o Work closely with, learned from and supported senior Pembina Staff;
. Exposure to the high pace and quality driven work Pembina is known for.
20L1to 20L4 - Simon Fraser University (CIEEDAC)- Research Associate, Burnoby, BC
o Research and write energy and air emissions reports for major industries and government
ministries (including Environment Canada);
o Produced professional looking reports in a short timeframe.
2013 Simon Fraser University - Teaching assistant for REM 35O Burnoby, BC
o Teaching assistant for Prof. Mark Jaccard's upper level undergraduate course on energy
economics and policy;
o Lead three tutorials per week, facilitated dialog among students on energy issues;
o Held lecture before l-50 students, showing assurance talking before large audiences.
20L0 Mawera Energy Consulting (ltd) - Writer and proof readerReyelstoke, BC

o Assistant to consultant on a feasibility study for a district heating project;
o Gained understanding on the economics of large resource projects in British Columbia.
562
Important publications
Backgrounder: Can Pacific NorthWest LNG pass Canadas climate test (2016);
Pacific NorthWest LNG B.C.s climate plan fails to limit emissions (2016);
Woodfibre LNG emissions analysis highlights methane policy gaps in B.C. -- formal submission to
CEAA (2016);
Change in emissions by 2030 for Canadas most populous provinces (2016);
California gas leak should spur Canada to get methane emissions under control (2016);
The Path to Net-Zero Energy Buildings in B.C. (2015);
BC Shale Tool Model and Technical Report (2015);
Canadas Policy Support for Clean Technology Exports Report Card (2015);
British Columbia Green Buildings Map (2015);
British Columbia Clean Energy Jobs Map (2015);
Masters thesis. Achieving Canadas climate targets and the impacts on Albertas oil sands
industry (2014).
Energy and GHG Emissions in British Columbia 1990 2010 (2012)
Conferences and self directed learning
2016 Oil & Gas Development in North America: Opportunities for Regulatory
Harmonization
Workshop in Washington D.C. with senior representatives from governments, industry and non-
profit research groups on improving the regulatory system for North Americas O&G industry.
2011-2014 Clean Energy BC Conference
Connected with various players in the B.C. energy field;
Engaged in the debate on independent power production in BC with a variety of stakeholders.
2013-2014 Carbon Management Canada Conference
Present research to funders and discussed key challenges;
Acquainted with stakeholders of the carbon management field and familiarized with key issues.
Other Work Experience
2008 to current - Freelance photographer and writer
Maintaining a positive relationship with over a dozen magazine editors internationally, providing
photography and creative well written feature length articles;
Author of seven feature articles in domestic and international outdoor sports magazines;
Published photography in over 50 magazine issues.
2006-2010 Outdoors guide and trip leader Whistler, BC
Organize logistics and provide leadership in a busy outdoor recreation company;
Comfortable communicating with large and diverse groups of people.
Special Projects:
2006-Present High-level whitewater kayaker
Expedition kayaker with numerous trips around the world.
2012 Homathko to Waddington ski and kayak expedition
Completed the first ever major kayak accessed ski-mountaineering expedition;
Organized expedition funding and logistics;
Described as "one of the all time great achievements in adventure sports".
2010 Completed the Triple Crown of kayaking (BC & The Yukon)
Trip leader of the third successful Triple Crown expedition down North Americas three hardest
whitewater wilderness rivers. Comprehensive photo documentation and publication on the trip.
563
2009 Feature athlete for a National Geographic movie on whitewater kayaking

Documentary of the last free descent of the Ashlu River, British Columbia;
Part of a public river awareness project.
Further Skills:
Fluent in German and adequate in French
Experience with CIMS energy-economy model and several statistical programs
Outgoing personality, good public speaking and excellent people skills
References available upon request
564
FW: could you review this please
From: ibit ".-L--" referred to in the

't affidavit of
"Chapman,Steve [CEAA]" <steve.chapman@ceaa-acee. gc.ca> a1'lirmed befbre me on the
ay ofAoril,20l7.
To:
"l nrvin, Ki mberly [CEAA]" <ki m berly. inrui n@ceaa-acee. gc.ca>
Date:
Tue, 10 Jan2017 21:23:18 +0000
From: Mongrain,Steve [CEAA]

Sent: September 12, 2015 3:50 PM
To: Chapman,Steve ICEAAI
Subject: RE: could you review this please
ls this for "full build out" or by 2030
From: Chapman,Steve ICEAA]

To: Mongrain,Steve ICEAAI
Subject: could you review this please
Natural Resources Canada conducted an economic analysis of using grid power for non-compression needs. Based on
available information, Natural Resources Canada estimates that the use of BC Hydro grid power for non-compression
needs over the long term could provide an economic benefit to the proponent. ln its modelling, Natural Resources
Canada assumed that the proponent would be responsible for the capital cost of upgrading existing transmission
capacity and building new transmission infrastructure to the Project site, estimated at S1 billion. ln addition, it was
assumed that the proponent would also need four 55MW generators on site for backup non-compression power needs.
Based on the analysis conducted by Natural Resources Canada, the Agency is of the view that it would be economically
feasible for the proponent to utilize grid power for the non-compression needs of the Project once it reaches full build
out or by 2030.
Steve Chapman
Director, National Programs

Canadian Environmental Assessment Agency / Government of Canada
Steve. Cha pma n @ceaa -acee. qc. ca / Tel : 61 3-957 -029 4 / Cel : 6t3-808- 47 47
Directeur, Programmes nationaux

Agence canadienne d'6valuation environnementale / Gouvernement du Canada
T6l. : 613-957-0294 / T6l. cell. : 6L3-808-4747
EAo6536
565
FW: As discussed
From: "Y.." refened to in the

affidavit of
"Chapman,Steve [CEAA]" <steve.chapman@ceaa-acee. gc. ca>
To: day of April,2017.
"l nrvin, Ki mberly [CEAA]" <ki mberly. inrui n@ceaa-acee. gc.ca>

Date:
Tue, 10 Jan2017 21:18:08 +0000
From : Cla u sen, ft ott ( N RCa n/RNCa n) [ ma lto scott.cla

i : usen @canada, ca]
To: Chapman,Steve [CEAA]; Mongrai n,Steve [CEAA]
Cc: Foran2, John (NRCan/RNCan)
Subject: As discussed
Impoftance: High
Hi Steve and Steve:
As discussed. The above email from Cristobal is the most recent - updated from the one below it.
ljust went up to seek confirmation, but those who did this analysis were not at their desks, so I cannot yet confirm that
this is NRCan's final numbers/analysis. That's a key point in terms of whether and how far to quote this, since we would
still need to vet it. so, maybe keep this tight until I can confirm their comfort.
However, the team that prepared it did say they are fairly confident the numbers are in the ball-park, with the caveat
that they have made certain assumptions noted below.
Hope this helps. Let me know if we can be of further assistance.
Scott Clausen
Deputy Director, LNG & lndigenous Affairs
Petroleum Resources Branch
Natural Resources Canada
D-7, 17th Floor, 580 Booth St
Ottawa, ON K1A 0E4
sclausen @nrcan.gc.ca
613-808-7998
From: Miller, Cristobal (NRCan/RNCan)

Sent: September 8, 20L6 L7:36
To: McCallum, Shawn (NRCan/RNCan); Clausen, Scott (NRCan/RNCan)
Cc: Schmidtke, Jeffery (NRCan/RNCan); Pr6fontaine, Simon (NRCan/RNCan); Neill, Warren (NRCan/RNCan)
Subject: RE: Required Analysis
HiShawn/Scott,
EAo6534
566
EA06534
567
EA06534
568
This is Exhibit ".h"refened to in the
1't affidavit of
day of Aoril, @.
Version: August 23, 2OtG
Potential Conditions A commissioner for

British Colurnbia
The Canadian Environmental Assessment Agency is contemplating the following potential conditions in
relation to the Pacific NorthWest LNG Project (the Designated Project) for recommendation to the
Minister of Environment and Climate Change for inclusion in a Decision Statement issued under the
Canadian Environmental Assessment Act, 20L2.lf the Designated Project is ultimately a llowed to
proceed because the Minister of Environment and Climate Change decides that the carrying out of the
Designated Project is unlikely to cause significant adverse environrnental effects as defined under
subsections 5(1) and 5(2), or if the Minister decides that the Designated Projects is likely to cause
significant adverse environmental effects and the Governor in decides such effects are justifia ble
ln the circumstances, any conditions established bythe Minister e legally binding.
Definitions
1.1. Agency means the Canadian Environmental Assessrnent Age
1.2. Airdroft means the distance from the vessel.
1.3. Designated
7.4. Cetoceon means whales,
1.5. Construction mean s t repa ration, building or

installation of any- ken bythe Proponent.
1.6. Culturally m by lndigenous Peoples as part of their

traditional use of
7.7. hou r after sunset, as ca lculated by the
1.8.
L.9. seofthe Designated Project where the Proponent has

permanen I production and has commenced removal from service of any
components o and continues until the site is restored.
1.10. Designated Project Pacific NorthWest LNG Project as described in section 2 of the
environmentalasses t report prepared by the Canadian Environrnental Assessment Agency
(Canadian Environmental Assessment Registry Reference Number 80032).
1-.1L. Environment and Climate Change Conado means the Department of the Environment as
established under subsection 2(L) of the Deportment of the Environment Act.
7.1,2. Environmental effects means "environmental effects" as described in section 5 of the Conodiqn
Environmental Assessm ent Act, 2O12.
7.73. Fish means "fish" as defined in subsection 2(1) of the Fisheries Act.
Page [APG] of [ANP]
EA02909
569
EA02909
570
EA02909
571
EA02909
572
EA02909
573
EA02909
574
EA02909
575
EA02909
576
EA02909
577
EA02909
578
EA02909
579
EA02909
580
EA02909
581
EA02909
582
EA02909
583
EA02909
584
EA02909
585
EA02909
586
EA02909
587
EA02909
588
EA02909
589
EA02909
590
EA02909
591
EA02909
592
EA02909
593
EA02909
594
EA02909
595
a
2500.666 llre(l iilnlnt r;rB ril.r.q7(1:) 'ntla+F.wt.a .r r:
"iil'#8ffi***
Sr
n,uo,t VdrKf,il!sr. llriti:il t0i.ufl !hid Fa;(i iitl4 03.1f,$61 PdriR{Norfl1We!tLN6J-om

C0tddii ViiC 3P1
C
is is Exhibit ",I..." referred to in the
September 8,2015 't affidavit of
of April,2017.
Canadian Environmental Assessment Agency (CEAA or Agency)
150 Elgin Street, 22nd Floor
Ottawa, ON K1A 0H3
A commissioner tbr tak
Attention: Heather Smith, Vice President - Operations
, British Columbia
.$i-. Subjectr Response to CEAA letter received on September 6, 2O16
Dear Ms. Smith:

a
I am writing in response to your letter received by Pacific NorthWest LNG (PNW LNG) on
September 6,2016, in which the Agency is asking for further information related to the proposed
Project's power requirements and related matters. We would like to begin by addressing the
CO2e lntensity Rate, which is raised in the first part of your letter. We will then answer each of
your specific questions in turn.
CO2e Emissions lntensity Rate
ln the correspondence to the Agency, the Ministry of Environment and Climate Change and
Tsimshian Nation Chiefs, PNW LNG has stated that the project will emit between 0.19 and 0.22
tonnes of CO2e per tonne of LNG produced.
As PNW LNG has previously stated, there are several factors, rnostly uncontrollable, that can
affect emission intensity from the LNG facility, including:
1. Gas turbine technology
Beginning in 2013, we undertook a competitive Front End Engineering & Design (FEED)
process that included three industry leading LNG Engineering Procurement Construction
and Commissioning (EPCC) Contractors working on competitive designs over a period of
nearly two years. With the use of best in class aero derivative gas turbine technology, the
CO2e per tonne intensity will be reduced by approximately twenty percent from the Pre-
FEED design.
2 Ambient Temperature
The annual temperature variance is 19o Celsius in the Prince Rupert area. This will
influence the operating efficiency of the gas turbines.
3 Natu ral gas composition

This is a significant unknown factor. lt is PNW LNG's intent to source its natural gas from
Progress Energy Canada ttd. (PECL). PECL will deliver its natural gas to an open or common
carrier transmission system in which the natural gas supply will be corningled with gas
from other shippers. The natural gas will be first delivered into the NEB regulated
TransCanada Nova Gas Transmission Ltd. (NGTL) system which has defined gas
composition ranges within its tariff. ln the NEB hearing proceedings for the NGTL North
EAO27B8
596
EA02788
597
EA02788
598
EA02788
599
"ill' PRoGREss ENERG,

: q. CANADA LTD l:il;,iXiJJllli,#,L,,,'.5t;' frZil:;:*",t ir.,,,vG i i:[:;:i,'i.ili*.*".,
July 13, 2016 HAND
This is Exhibit "I)l." referred to in the
ll affidavit of Maximilian Kniewasser
TO; day of Aoril, 2017.
Tsimshian Nation Chiefs
Dear Chiefs of the Tsimshian Nation:
RE: Pacific NorthWest LNG & Progress Energy lntegrated Project A commissioner for taking-affidavits in
British Columbia
This letter cornes as a response to requests for clarification on Pacific NorthWest LNG's (PNW
LNG) approach to Greenhouse Gas (GHG) emisslons related to our lntegrated natural Bas to LNG
project. By way of background, in 2012, because of the limited capacity and reliability of the
local electrical grid, PNW LNG dlrected all of the competing Englneerlng Procurement and
Construction Contractors to base their Front End Engineering Design (FEED) work on the most
advanced yet proven natural gas turbine technology for the proiects power requirements
(compression, non-compression and ancillary) for the first phase of our proposed facility (Trains
x&2).
For the construction of the first phase of the project, it is anticipated that due to the lack of
commercially available supplied power, construction activity on Lelu lsland will consume self-
generated power. PNW LNG is working wlth its contractor to have temporary facllities such as
the workers' camp that will be located on the main land connected to the BC Hydro grid,
The facility, through continued improvemerlt during the FEED the first phase, is expected to
have a carbo n inten sity ratio of O, 19-0.22 CO2 equava lent for every one ton of LNG produced. For
comparison purposes; LNG facalities currently in operation have a rypical carbon intensity
benchmark of 0.31C02 equivalent for every one ton of LNG (please see attached factsheet for
further inforhation), British Columbla has taken this a step further by requiring LNG facilities to
meet the lowest in the world carbon intensity ratio of 0.15 CO2,through offset prograrrrs that
will be established.
As you are aware, Pacific NorthWest LNG's vision is to add additional LNG capacity to the project
in the future through the construction and operation of incremental "trains", based on growing
LNG demand. With that in mind, we are encouraged by the potential for future renewable
energty opportunities in northwestern British Columbia to meet these power needs. PNW LNG
looks forward to incorporating reliable renewable power in our future expansion plans in the
following waysl
1. Future phases of the project will be designed for a certainty that reliable grid to be
I
provided to the facility for ancillary purposes.
2. Retrofitting at the end of the useful life of the first phase of the profect may enable the
utilization of grid power and/or electric turbines to power non-compression processes.
EA02781
600
EA02781
601
LIST OF AUTHORITIES
Case Tab
Abbott Laboratories Limited v Canada (Attorney General), 2008 FCA 354 1.
Alberta Wilderness Association v Canada (Environment), 2009 FC 710 2.
Association of Universities and Colleges of Canada v Canadian Copyright 3.

Licensing Agency (Access Copyright), 2012 FCA 22
Adam v Canada (Environment) 2014 FC 1185 4.
Bernard v Canada (Revenue Agency), 2015 FCA 263 5.
Canada (Attorney General) v Quadrini, 2010 FCA 47 6.
Canada (Board of Internal Economy) v Canada (Attorney General), 2017 FCA 43 7.
Canadian Tire Corporation v Canadian Bicycle Manufacturers Association, 2006 8.

FCA 56
Delios v Canada (Attorney General), 2015 FCA 117 9.
Shoan v Canada (Attorney General), 2017 FC 426 10.
Tsleil-Waututh Nation v Canada (Attorney General), 2017 FCA 116 11.
Wagner v Canada (Attorney General), 2016 FC 412 12.
Legislation
Federal Courts Act, R.S.C. 1985, c F-7, s 18.4 13.

Crown's Motion To Strike Zickfeld and Kniewasser Affidavits PNW LNG July 2017

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Crown's Motion To Strike Zickfeld and Kniewasser Affidavits PNW LNG July 2017

Uploaded by

Copyright:

Available Formats

Court File No.

SKEENAWILD CONSERVATION TRUST

HER MAJESTY THE QUEEN, ATTORNEY GENERAL OF CANADA,

MOTION RECORD OF THE RESPONDENTS, HER MAJESTY THE

Counsel for the Respondent, Her Counsel for the Applicant,

Mark Andrews, QC and Bridget Gilbride

TAB DOCUMENT PAGE

2. Affidavit #1 of Jane Tyler (affirmed June 12, 2017) 5

3. Written Representations 419

4. SkeenaWild Notice of Application (dated October 27, 2016) 439

5. Affidavit #1 of Kirsten Zickfeld (affirmed April 26, 2017) 455

Affidavit #1 of Maximilian Kniewasser (affirmed April 27,

7. Book of Authorities (Index) 601

SKEENAWILD CONSERVATION TRUST

HER MAJESTY THE QUEEN, ATTORNEY GENERAL OF CANADA,

MOTION RECORD OF THE RESPONDENTS, HER MAJESTY THE

Counsel for the Respondent, Her Counsel for the Applicant,

Mark Andrews, QC and Bridget Gilbride

This is the 1st affidavit of

Court File No.: T-1836-16

SKEENAWILD CONSERVATION TRUST

HER MAJESTY THE QUEEN, ATTORNEY GENERAL OF CANADA,

APPLICATION UNDER SECTIONS 18 & 18.1 OF THE FEDERAL COURTS ACT,

I, Kirsten Zickfeld, Associate Professor in the Department of Geography, Simon

1. I am a climate scientist who has reviewed the Canadian Environmental

2. I have been retained by SkeenaWild Conservation Trust as an expert in these

3. I am a climate scientist and Associate Professor in the Geography Department

4. Scientific evidence suggests that the climate effects of emissions of long-lived

5. Human activities, such as combustion and production of fossil fuels,

12. In a cumulative GHG emissions framework, cumulative Project GHG

in terms of their share of a provincial or national carbon budget. A carbon

14. There is considerable uncertainty surrounding 2C carbon budget estimates due

A commissioner for taki Kirsten Zickfeld

Pak Yan Anthony HO

April 19,2017 Our file: 2016-09-0002

VIA ELECTRONIC MAIL

Dear Dr. Zickfeld:

FfEz SkeenalYild Conservation Trust v. Her Majesty The Queen et aL

We are counsel to SkeenaWild Conservation Trust ("SkeenaWild") in the above-referenced legal

Suite 16 Shoal Point,2l Dallas Rd., Victoria BC, V8V 429

1. The Report dated September 27, 2016; and,

Code of Conduct for Expert Witnesses*****

Suite 16 Shoal Point, 21 Dallas Rd., Victoria BC, V8V 4Z9

In your affidavit, please:

1. State your full name and address;

Suite 16 Shoal Point, 21 Dallas Rd., Victoria BC, V8V 4Z9

You must complete your report by April 28, 2017.

Privilege and Confidentiality

Suite 16 Shoal Point, 21 Dallas Rd., Victoria BC, V8V 4Z9

Employment History at Academic Institutions

Kirsten Zickfeld Curriculum Vitae

Other - Education and Employment Information

Awards, Honors and Scholarships

Updated: May 11, 2016 Page: 2

Kirsten Zickfeld Curriculum Vitae

Refereed Journal Articles

Updated: May 11, 2016 Page: 3

Kirsten Zickfeld Curriculum Vitae

Updated: May 11, 2016 Page: 4

Kirsten Zickfeld Curriculum Vitae

Reviewed Book Chapters

Updated: May 11, 2016 Page: 5

Kirsten Zickfeld Curriculum Vitae