1
Projected Shifts in Köppen Climate Zones over China and their Temporal Evolution
2
in CMIP5 Multi-Model Simulations
3
Duo Chan, Qigang Wu*, Guixiang Jiang and Xianglin Dai
4
School of Atmospheric Sciences, Nanjing University
5
22 Hankou Road, Nanjing, JiangSu, China, 210093
6
7
8
9
10
*
Corresponding author: Dr. Qigang Wu (Email: [email protected])
11
Tel: 86-25-83593953; Fax: 86-25-3084
12
August 27, 2014
13
14
15
16
17
1
18
ABSTRACT
19
Because Köppen-Geiger climate classes are based on major ecosystem thresholds, analysis of
20
spatial changes in climate types can significantly improve public communication of ongoing and
21
future anthropogenic climate change impacts. Previous studies have examined the projected
22
climate types in china by 2100, which would shift toward warmer climate types from current
23
climate distribution. This study identifies the emergence time of climate shifts at a 1ºscale over
24
China from 1990-2100 and investigates the temporal evolution of climate classifications
25
computed from CMIP5 multi-model run outputs. By 2010, climate shifts are detected in
26
transition regions (7.5% of total land area), including rapid replacement of mixed forest (Dwb)
27
by deciduous forest (Dwa) over Northeast China, strong shrinkage of alpine (ET) climate type on
28
the Tibetan Plateau, weak northward expansion of subtropical winter-dry (Cwa) over eastern
29
China and contraction of oceanic climate (Cwb) in Southwest China. Under all future
30
Representative Concentration Pathway (RCP) scenarios, reduction of Dwb in Northeast China
31
and ET on the Tibetan Plateau accelerates substantially during 2010-2030, and half of the total
32
area occupied by ET in 1990 is projected to be redistributed by 2040. Under the RCP 8.5
33
scenario, sub-polar continental winter dry climate over Northeast China disappears by 2040-2050,
34
ET on the Tibetan Plateau disappears by 2070, and the climate types in 37.7% and 52.6% of
35
China’s land area change by 2050 and 2100, respectively.
36
Keywords: Köppen-Geiger climate classification, China, climate change, CMIP5, RCP
37
scenarios
2
38
1. Introduction
39
China has a population over 1.3 billion, the world’s third largest land area, varied topography,
40
and diverse ecosystems, so the impacts of projected climate changes are expected to be large (Yu
41
et al. 2006; Ni 2011). To estimate ecological changes, some studies have used climate model
42
outputs to drive a specialized vegetation model (VM) that projects changes in vegetation cover
43
based on temperature, precipitation and other factors such as surface runoff and soil feedbacks
44
(Wang and Zhao 1995; Ni et al. 2000; Zhao et al. 2002; Zhao and Wu 2013; Wang 2014). VMs
45
project the following changes in vegetation cover over China by the end of the 21st century: a
46
northward shift of all forests, disappearance of boreal forest over northeastern China, new
47
tropical forests, an eastward expansion of grasslands, and a reduction in alpine vegetation (Ni
48
2011). Other studies, using Köppen or related climate classifications to investigate possible shifts
49
in climate zones over China (Baker et al. 2010; Shi et al. 2012), generally project similar
50
decreases of boreal forest in northeastern China, evergreen forest in southeastern China, and
51
alpine tundra over the Tibetan Plateau.
52
This study extends those large-scale studies by investigating, with detailed time and space
53
resolution, the temporal evolution of observed and projected climate types over China, including
54
the emergence time and further rate of anthropogenic-related change of climate types, and the
55
relative role of temperature and precipitation in projected climate shifts. Other considerations are
56
as follows. First, China encompasses a variety of climate types, from southern tropical to
57
northern boreal, and from eastern humid to western arid and alpine climates (Ni 2011). Second,
58
projected future climate changes are not uniformly distributed among different regions in China.
59
For instance, projected warming is up to 1.5ºC greater in northern and western regions than in
60
southeastern areas in the RCP8.5 scenario (Xu and Xu 2012, also see Fig. 1). Annual
3
61
precipitation increases in all areas, with the largest increases in southwestern regions (see Fig.
62
1b), but the greatest percentage increases are in northern and western regions (Xu and Xu 2012).
63
Third, vegetation cover in different regions has varying sensitivity to climate change.
64
Ecosystems over western and northern China are found more vulnerable to changed climate than
65
those over eastern China (Ni 2011), and low-latitude mountainous regions are more likely to
66
experience a shift in climate types (Mahlstein et al. 2013).
67
The Köppen climate classification scheme was developed to explain observed biome
68
distributions, which have many sharp boundaries due to plant sensitivity to threshold values of
69
average monthly temperature and precipitation and their annual cycle (Köppen 1936). Changes
70
in climate classes are more general indicators for climatic changes than are temperature and
71
precipitation trends alone. While climate zones are not exact boundaries for all species, locations
72
with a change in climate class are expected to experience the most noticeable ecosystem stress,
73
such as widespread tree dieoffs in forests. Köppen or similar classifications have been used to
74
estimate the potential impacts of past and projected future climate on prevalent ecoregions on
75
regional and global scales (e.g., Wang and Overland 2004, Gnanadesikan and Stouffer 2006, De
76
Castro et al. 2007, Baker et al. 2010, Feng et al. 2014, Mahlstein et al. 2013). The Köppen-
77
Geiger classification with slight revisions by Peel et al. (2007) is used here to depict the
78
projected evolution of climate types over China during the 21st century. In the rest of this paper,
79
data and methods are described in section 2, results are presented in section 3, and a conclusion
80
is provided in section 4.
81
2. Data and method
4
82
We use monthly temperature and precipitation grids from the University of Delaware (Legates
83
and Willmott 1990a,b) to calculate the observed Köppen climate type in each grid box using
84
1980-1999 (the “base period") averages for each month of the year. Version 3.02 covers 1900-
85
2010 and is derived from GHCN2 (Global Historical Climate Network) and additional data
86
obtained by Legates and Willmott covering the entire terrestrial area. Simulated fields of
87
temperature and precipitation through 2100 are from the Coupled Model Intercomparsion Project
88
5 (CMIP5) simulations performed using 31 global climate models (Table 1). The model runs are
89
forced by observed atmospheric composition changes (reflecting both anthropogenic and natural
90
sources) from 1950-2005 (the historical data period), and by prescribed changes in greenhouse
91
gases (GHGs) and anthropogenic aerosols from 2006-2100 following four different IPCC
92
Representative Concentration Pathways (RCP2.6, RCP4.5, RCP6 and RCP8.5) from 2006-2100
93
(Taylor et al. 2012). Radiative forcings in 2100 are about 2.6 (after a peak value 3.0 around
94
2040), 4.5, 6.0 and 8.5 Wm-2 respectively relative to preindustrial including most known forcing
95
factors (compared to 1.75 Wm-2 in 2000), and corresponding equivalent CO2 concentrations
96
(expressing all other radiative forcings as added or subtracted CO2 amounts) in 2100 are
97
projected to be about 470 (after a peak of 490 around 2040), 650, 850 and 1370 ppm (van
98
Vuuren et al. 2011). As in Mahlstein et al. (2013), for a specific model and scenario with
99
multiple ensemble runs, only output from one ensemble (r1i1p1) is used so that all models are
100
afforded equal weight.
101
Köppen-Geiger climate classes are determined using the criteria defined in Peel et al. (2007).
102
Classification first identifies 5 major climate types: A – tropical, B – dry (arid and semiarid)
103
climate, C – subtropical, D – continental and E – polar. Type B is based on moisture availability
104
while other types are based on annual temperatures. Each major type has 2 or 3 subtypes, and
5
105
each subtype in types B to D has 2 to 4 minor subtypes, all classified according to the seasonal
106
cycle of temperature, precipitation or both. There are 30 possible climate types, and Table 2 lists
107
criteria for the 18 climate classes relevant for China. While types A, C, D and E are mutually
108
exclusive temperature categories, any arid or semiarid location (Type B) also fits the temperature
109
criteria of another climate type, so tests for Type B are performed first.
110
The following pretreatments, which were also performed by Mahlstein et al. (2013) and Feng et
111
al. (2014), are applied to both observed and simulated fields before computing the Köppen
112
classes. First, the monthly grids of data are interpolated onto a 1º× 1ºgrid to ensure the same
113
resolution. Second, anomalies relative to the simulated monthly means for 1980-1999 are
114
calculated in each model and each of the 4 scenarios, and are then added to the observed 1980-
115
1999 monthly means to reduce errors in the climate classification. This is necessary since the
116
Köppen-Geiger climate classification is highly sensitive to thresholds, and current models have
117
temperature and precipitation biases that cause difficulties in simulating the correct Köppen
118
classes even for present conditions (Gnanadesikan and Stouffer 2006; Mahlstein et al. 2013).
119
Third, to emphasize trends rather than short-term climate variations that cause erratic climate
120
zone movements, a 15-yr running smoothing is applied to all CMIP5 grids (monthly grids are the
121
averages of the same month of each year). Fifteen years is the optimal averaging interval for
122
Köppen or related classifications (Fraedrich et al. 2001), but results here are not highly sensitive
123
to the length of the averaging period. In the rest of this paper, model-derived data for a “year”
124
refers to a 15-year centered average, and the last available year of 2093 is the 2086-2100 average.
125
This study does not focus on differences between models, so grids from all model runs in each
126
scenario are averaged together. The pretreated monthly multi-model averaged temperature and
127
precipitation grids are then used to assign Köppen classes for each grid box and year from 1991
6
128
(1984-1998 average) to 2093 (2086-2100 average), and the year of change (if any) of climate
129
class from the 1990 base year is noted. To distinguish possible natural fluctuations of climate
130
zones, the naturally occurring climate shifts at grid box are identified with detrended 15-year
131
running mean temperature and precipitation for each model and scenario (natural group). In this
132
way, a change for a particular grid point in projected climate regions is counted only if the
133
appearing novel climate is also different from the naturally fluctuating climate in the same year
134
or does not return to the present type until the end of 21st century. Taking into account the
135
significant observed linear trend in temperature during the period of 1980-1999, the distribution
136
of climate zones in the natural group should resemble the observational distribution in 1990 most.
137
All changes found are statistically significant and the underlying climate changes are caused by
138
increasing anthropogenic forcing, because they are not due to natural variability. In following
139
analyses, the percentage of land area covered by each climate type is computed by area-
140
weighting all grid boxes by the cosine of their latitude.
141
3. Results
142
3.1 Projected Climate Changes
143
Figure 1 displays multi-model ensemble-mean warming and precipitation trends during 1990-
144
2100 in the RCP8.5 scenario. Trends are consistent with previous studies (Xu and Xu 2012).
145
Three centers of greater warming are located over Northeast and Northwest China, and the
146
Tibetan Plateau (Fig. 1a), with generally more warming in winter than in summer (Fig. 1c and
147
Fig. 1e). The projected annual precipitation increases over all of China (Fig. 1b). While the
148
southeastern part of China is currently the wettest region, only small increases are expected there,
149
with a band of increased precipitation extending from inland southeastern China to northeastern
7
150
China. The largest precipitation increase is expected in southwestern China including part of the
151
Tibetan region into central China, an area that is currently moderately dry. Figure 1f shows that
152
the increased precipitation in these regions occurs mostly in summer. Reduced precipitation in
153
winter is found in regions of Southwest and southeast China and the Himalayas (Fig. 1d). The
154
spatial distributions in other scenarios are similar to that in RCP8.5 except for smaller amplitudes
155
(not shown).
156
3.2 Projected Shifts of Climate Types
157
Figure 2a shows the 1990 base year Köppen classes over China, and encompasses 13 subtypes
158
from all 5 basic types. Table 2 lists Köppen classes observed or projected in China from Peel et
159
al. (2007), as well as their corresponding climate (Shi et al. 2012) and general vegetation types
160
(Hou et al. 1982). There is a small area of tropical climate (Am) in the very south of China while
161
deserts (Bw) and steppe (Bs) are dominant over northwestern China. Temperate climate (C),
162
mainly including three subtypes (humid (Cfa), subtropical winter dry (Cwa) and oceanic climate
163
(Cwb)), is found over Southwest and southeastern China. North and Northeast China are mainly
164
characterized with continental humid climate (Dwa and Dwb), while sub-polar climate (Dwc) is
165
detectable over the northernmost region. Finally, alpine climate (ET) dominates the central and
166
southwestern Tibetan Plateau while continental climate (mainly Dwb and Dwc) is found over the
167
eastern and southern Tibetan Plateau and arid climate (mainly BWk and BSk) is found over the
168
western and northern Tibetan Plateau. The coverage of major climate types A-E in 1990 is about
169
0.5%, 29.7%, 27.9%, 32.1% and 9.8% of the total land area respectively. Such distributions of
170
present-day Köppen classifications are consistent with regional climate types over China in
171
global studies by Peel et al. (2007) and Rubel and Kottek (2010), and reflect the vegetation map
8
172
of Hou et al. (1982) fairly well, which indicates that the Köppen climate types are well able to
173
capture signals of eco-climate regions in China.
174
The projected climate types for 2093 derived from multi-model averages are shown in Fig. 2b-e.
175
Under the low emissions RCP2.6 scenario, the most conspicuous changes are found in Northeast
176
China, characterized by the shift of Dwb towards Dwa and a retreat of Dwc, and also over the
177
Tibetan Plateau, characterized by shifts of ET to Dwc and Dwc to Dwb. Other changes include a
178
northwestward retreat of Cwb associated with an expansion of Cwa in Southwest China and a
179
northward expansion of Cwa and Cfa over East and Central China.
180
These changes are larger under the middle emissions RCP4.5 and RCP6 scenarios, including the
181
further reduction of Dwb and the disappearance of Dwc in Northeast China, the replacement of
182
ET with transitional boreal forest (Dwc and Dwb) in the eastern and southern parts of Tibetan
183
Plateau due to increasing summertime temperature (Fig. 1e), and a great shrinkage of Cfa in
184
Southeast China mainly caused by decreasing precipitation in winter. Dramatic changes are
185
found in the high emissions RCP8.5 scenario, characterized by the disappearance of Dwb in
186
Northeast China (replaced by Dwa), the disappearance of ET on the Tibetan Plateau (replaced by
187
Dwb and Dwc in southern and eastern parts, and BSk in the northwestern part), a vast shrinkage
188
of Cfa over southeastern China, the emergence of Bwa (hot desert climate) in Northwest, and
189
extension of Aw (tropical savanna) reaching the nearby coast in South China.
190
Higher emissions cause greater climate change impacts. Under the RCP2.6, 4.5, 6 and 8.5
191
emissions pathways, about 22.8%, 39.0%, 46.4% and 52.6% of total land area undergoes a
192
change in its 1990 climate, and the projected areas by 2093 are 28.4%, 28.4%, 29.2% and 30.0%
193
for type-B climate, 30.1%, 32.8%, 32.9% and 35.5% for type-C climate, 35.6%, 36.0%, 36.9%
9
194
and 31.1% for type-D climate, and 4.7%, 2.2%, 0.7% and 0% for ET. Under the RCP8.5 scenario,
195
changes in Fig. 2e agree with regional results over China of global studies by Rubel and Kottek
196
(2010) and Feng et al. (2014), and the subtype changes over eastern China are generally
197
consistent with the findings of Shi et al. (2012).
198
199
3.3 Temporal Evolution of Climate Shifts in China
200
Figure 3 shows time series from 1990 to 2093 of the percentage of the national land area
201
experiencing a change in Köppen class from 1990, and the percentages occupied by major
202
climate types B to E and certain C and D subtypes. In the historical period (2010 compared to
203
1990), anthropogenic influence is already detectable, with strong expansion (contraction) in
204
type-D (ET) climate and weak expansion (contraction) in type-C (B) climate, causing climate
205
type changes in about 7.5-8.0% of the land area. Under the RCP2.6, 4.5, 6 and 8.5 emissions
206
pathways, about 15.7%，25.0%, 25.4% and 37.6% (respectively) of total land area is projected
207
to undergo a change in 2050 compared to 1990.
208
The year when an anthropogenic-induced change in ecoregions compared to the 1990 base year
209
first emerges in each grid box for each scenario is shown in Fig. 4. There are considerable spatial
210
variations in emergence of those changes in Köppen climate types. First, changes of Köppen
211
climate types before 2010 are mainly seen in transition zones between present Köppen climate
212
types in Fig. 4a. Those changes are associated with the shift from Dwb to Dwa in Northeast
213
China, the northwestward retreat of Cwb in Southwest China, the northward expansion of Cwa
214
over the northern Central and East China, and the shift from ET to Dwc over the margins of the
215
Tibetan Plateau. Second, these ongoing shifts continue at a northward or westward pace from
216
transition zones. Northwest China and the Tibetan Plateau are two regions most sensitive to
10
217
projected warming in following decades. Rapid changes of Dwb to Dwa in Northeast China, and
218
strong shrinkage of ET on the Tibetan Plateau are projected before 2030-2040 under the RCP4.5
219
or higher scenarios (Figs. 4b-e). The projected area of ET is under 5.0% of the national land area
220
by 2040 (Fig. 3e). Third, the annual percent decrease of ET is projected to increase substantially
221
after 2010. The relative area of ET decreased about 1.9% during 1990-2010, but is projected to
222
decrease by about 2.7%, 2.7%, and 2.5% and 3.0% of China’s land area during 2010-2030 under
223
RCP 2.6, 4.5, 6 and 8.5 scenarios respectively (Fig. 3e). Fourth, the pace of northward or
224
westward shifts of climate zones increases under high RCP scenarios after 2030-2040. For these
225
grid boxes away from transition zones, projected climate shifts tend to emerge earlier under the
226
higher RCP scenarios than the lower scenarios. For instance, most shifts from Dwc to Dwb over
227
northern Northeast China will happen around 2040-2050 in RCP8.5, but 2060-2070 in RCP4.5.
228
Similar advances in emerging anthropogenic signals by 20-30 years can also be found in the
229
northward shift of temperate forests (Cwa and Cfa) over northern East China, the change of ET
230
to Dwb on the Tibetan Plateau, and the shrinkage of oceanic climate (Cwb) in Southwest China,
231
the shift of Cfa to Cwa in over southeastern China. Finally, the coverage of type-B climate is
232
about 28-30% through the 21st century under all scenarios (Fig. 3b), and the replacement of Cwa
233
by Af over South China occurs near the end of the 21st century under the RCP 8.5 scenario (Fig.
234
4d).
235
Figures 3f-g further investigate the temporal evolution of the percentage distribution of six sub-
236
types of major climate type-C and type-D under the RCP 8.5 scenario. The coverage of
237
temperate climate increases from the northward and westward expansion of Cwa and Cwb (Fig.
238
3e). Cwb is projected to decrease in Southwest China, but increase over the Tibetan Plateau (Fig.
239
3f). The evolutions of sub-type climate Cwa and Cfa over eastern China exhibit considerable
11
240
variability, and the projected coverage of Cwa (Cfa) shows a strong increase (decrease) during
241
2020-2060. The coverage of temperate climate increases about 4.6% and 7.6% by 2050 and 2093,
242
respectively (Fig. 3c). As climate change accelerates, Dwc is rapidly replaced by Dwb over
243
Northeast China before 2050 and primarily confined over the Tibetan Plateau after 2050. The
244
annual percent decrease of Dwb in Northeast is projected to increase substantially after 2010.
245
The coverage of Dwb is projected to be about 9.1%, 7.6%, 5.7%, 3.4% and 1.3% in 1990, 2010,
246
2030, 2050, and 2070 respectively, and decrease about 1.5%, 1.9%, 2.3%, 2.1% of China’s land
247
area during 1990-2010, 2010-2030, 2030-2050, and 2050-2070 respectively (Fig. 3g). The
248
coverage of Dwa is projected to increase to 15% (17%) in 2050 (2093) from 10% in 1990,
249
suggesting that the expansion of Dwa mostly finishes before 2050. The projected decrease of the
250
total area of type-D climate after 2070 in Fig. 3d is mainly due to the shift of Dwb to Cwb on the
251
Tibetan Plateau (Fig. 3g).
252
3.4 Relative roles of temperature and precipitation in climate shifts
253
Most changes in climate types over China seem to be temperature rather than precipitation driven.
254
Figure 2g-h show spatial distributions of projected Köppen classes for 2093 calculated with
255
changes in individual variables under the RCP8.5 scenario. Most of the climate shifts in Fig. 2e
256
are reproduced in the case in which precipitation is held constant, while only the present-day
257
distribution in Fig. 2a is broadly captured when the temperature is fixed. Similar results were
258
discussed in previous VM studies (Zhang and Zhou 2008; Wang et al. 2011).
259
Figure 2g-h also reveal that the area of arid climate increases (decreases) with growth in
260
temperature (precipitation) over North and Northwest China, suggesting that water and heat
261
factors play the opposite roles in arid climate B. The two offsetting effects may explain why the
12
262
coverage of arid climate B shows no significant changes in the 21st century even under the
263
RCP8.5 scenario (Fig. 2e). In addition, Figure 2h confirms that decreasing winter precipitation
264
causes the reduction of subtropical humid Cfa in southeastern China (Shi et al. 2012), while
265
increasing temperature results in the northwestward retreat of Cwb in Southwest China.
266
4. Conclusions
267
This study extends previous general studies of observed and projected climate shifts in China by
268
performing a detailed investigation of the temporal evolution of Köppen-Geiger climate types
269
over the national land area using observational data and simulations from multiple global climate
270
models participating in the CMIP5 project. Based on the projected changes in temperature and
271
precipitation, regions of temperate climate types are projected to expand, while regions of sub-
272
polar and alpine climate types are projected to contract significantly over the national land area
273
by 2100. These overall results agree with previous studies using Köppen or related climate
274
classifications (Baker et al. 2010; Shi et al. 2012; Feng et al. 2014), and studies with specialized
275
vegetation models driven by climate model outputs (Wang and Zhao 1995; Ni et al. 2000; Zhao
276
et al. 2002; Zhao and Wu 2013; Wang 2014). In addition, our results indicate that impacts of
277
anthropogenic climate change on regional climate zones are already evident and are expected to
278
accelerate in all future scenarios before 2050. The following specific findings are helpful for
279
promoting the public’s awareness of impacts of anthropogenic climate change on China’s
280
ecoregions:
281
1) The magnitudes of impacts of warming and the pace of shifts tend to be larger under high
282
emission scenarios. The proportion of the national land area experiencing changes in climate
283
types increases gradually from about 7.5-8.5% in 2010 to about 22.8%, 39.0%, 46.4% and
284
52.6% by the end of this century under the RCP2.6, 4.5, 6 and 8.5 emissions pathways.
13
285
2) Northeast China is one of the regions most sensitive to increasing warming in the first half
286
of this century. The pace of the shift of Dwb to Dwa, corresponding to the replacement of
287
temperate/boreal mixed forest by temperate deciduous forest (Leng et al. 2008; Ni et al.
288
20000; Ni 2011), will significantly increase in the next decades. Under RCP 8.5, Dwc
289
(boreal forest) disappears over northern Northeast China around 2040-2050.
290
3) The Tibetan Plateau is also a vulnerable region, with the disappearance of the current alpine
291
climate area by 2070 under the RCP8.5 scenario. The loss of alpine climate is concentrated
292
in the period of 2010-2030 and its projected area is under 5% by 2040 under all scenarios,
293
implying that impacts of climate changes over this region will accelerate in the near future.
294
There will be a major westward and northward shift of boreal deciduous forest/woodland
295
and alpine meadow, and a major reduction in alpine tundra/polar desert (Song et al. 2005; Ni
296
2011).
297
4) The region of temperate climate (C) is projected to expand due to the northwestward shift of
298
subtropical winter dry (Cwa) over eastern China and the increase of temperate ocean climate
299
(Cwb) over the Tibetan Plateau under all four scenarios. The coverage increases in 2093
300
from 1990 under all scenarios, but most of the increases are projected before 2050. A
301
northward and westward expansion of Cwa over Central and East China is associated with a
302
northwestward retreat of Cwb in Southwest China under all four scenarios. The shrinkage of
303
subtropical humid (Cfa) in East China after 2050 is due to the decrease in winter
304
precipitation under RCP 6 and RCP 8.5 scenarios.
305
5) There is no significant change of the total area of semi-arid and arid (B-type) climate over
306
North and Northwest China when taking both the increasing temperature and rainfall into
307
account.
14
308
Acknowledgements We acknowledge PSD group for providing UDel_AirT_Precip data. We
309
also acknowledge all participants of CMIP5 for producing and making available model outputs.
310
This work is supported by the National Key Scientific Research Plan of China (Grant
311
2012CB956002) and the National Natural Science Foundation of China (Grant 41075052).
312
References
313
314
315
Baker B, Diaz H, Hargrove W, Hoffman F (2010) Use of the Köppen–Trewartha climate
classification to evaluate climatic refugia in statistically derived ecoregions for the People’s
Republic of China. Clim Change 98:113-131
316
317
318
De Castro M, Gallardo C, Jylha K, and Tuomenvirta H. (2007) The use of a climate-type
classification for assessing climate change effects in Europe from an ensemble of nine regional
climate models. Clim Change 81: 329-341.
319
320
321
Feng S, Hu Q, Huang W, Ho CH, Li R, Tang Z (2014) Projected climate regime shift under
future global warming from multi-model, multi-scenario CMIP5 simulations. Global Planet
Change 112: 41-52
322
323
Fraedrich K, Gerstengarbe FW, Werner PC (2001) Climate shifts during the last century. Clim
Change 50: 405-417
324
325
326
Gnanadesikan A, Stouffer RJ (2006) Diagnosing atmosphere‐ocean general circulation model
errors relevant to the terrestrial biosphere using the Köppen climate classification. Geophys Res
Lett 33 doi 10.1029/2006GL028098
327
328
Hou XY, Sun SZ, Zhang JW, He MG, Wang YF, Kong DZ, Wang SQ (1982) Vegetation Map of
the People’s Republic of China. China Map Publisher Beijing (in Chinese)
329
330
331
332
Köppen W, 1936. Das geographisca systemder Klimate. In: Köppen,W., Geiger, G. (Eds.),
Handbuch der Klimatologie. 1. C. Gebr, Borntraeger, pp. 1–44.
333
Legates DR, Willmott CJ (1990a) Mean seasonal and spatial variability in gauge‐corrected,
global precipitation. Int. J. Climatol. 10: 111-127
334
335
Legates DR, Willmott CJ (1990b) Mean seasonal and spatial variability in global surface air
temperature. Theor Appl. Climatol. 41: 11-21
15
336
337
338
Leng W, He HS, Bu R, Dai L, Hu Y, Wang X (2008) Predicting the distributions of suitable
habitat for three larch species under climate warming in Northeastern China. For. Ecol. Manage.
254: 420-428
339
340
Mahlstein I, Daniel JS, Solomon S (2013) Pace of shifts in climate regions increases with global
temperature. Nat. Clim Change 3: 739-743
341
343
Ni J, Sykes MT, Prentice IC, Cramer W (2000) Modelling the vegetation of China using the
process‐based equilibrium terrestrial biosphere model BIOME3. Glob. Ecol. Biogeogr. 9: 463479
344
345
Ni J (2011) Impacts of climate change on Chinese ecosystems: key vulnerable regions and
potential thresholds. Reg. Environ. Change 11: 49-64
346
347
Peel, M.C., Finlayson, B.L., McMahon, T.A., 2007. Updated world map of the Köppen–Geiger
climate classification. Hydrol. Earth Syst. Sci. 11, 1633–1644.
348
349
Rubel F, Kottek M (2010) Observed and projected climate shifts 1901-2100 depicted by world
maps of the Koppen-Geiger climate classification. Meteorol. Z. 19: 135-141
350
351
Shi Y, Gao XJ, Wu J (2012) Projected changes in Köppen climate types in the 21st century over
China. Atmos. Oceanic Sci. Lett. 5: 495-498
352
353
Song M, Zhou C, Ouyang H (2005) Simulated distribution of vegetation types in response to
climate change on the Tibetan Plateau. J. Veg. Sci. 16: 341-350
354
355
Taylor KE, Stouffer RJ, Meehl GA (2012) An Overview of CMIP5 and the Experiment
Design. Bull. Am. Meteorol. Soc. 93: 485-498
356
357
Van Vuuren DP, Edmonds J, Kainuma M, et al (2011) The representative concentration
pathways: an overview. Clim Change 109: 5-31
358
359
Wang F, Zhao ZC (1995) Impact of climate change on natural vegetation in China and its
implication for agriculture. J. Biogeogr. 22: 657-664
360
361
Wang H, Ni J, Prentice IC (2011) Sensitivity of potential natural vegetation in China to projected
changes in temperature, precipitation and atmospheric CO2. Reg. Environ. Change 11: 715-727
362
363
Wang H (2014) A multi-model assessment of climate change impacts on the distribution and
productivity of ecosystems in China. Reg. Environ. Change 14: 133-144
364
365
Wang M., Overland JE (2004) Detecting Arctic climate change using Köppen climate
classification. Clim Change 67, 43–62.
342
16
366
367
Xu CH. Xu Y (2012) The projection of temperature and precipitation over China under RCP
scenarios using a CMIP5 multi-model ensemble. Atmos. Oceanic Sci. Lett. 6: 527-533
368
369
Yu L, Cao M, Li K (2006) Climate-induced changes in the vegetation pattern of China in the
21st century. Ecol. Res. 21: 912-919
370
371
Zhang Y, Zhou G (2008) Terrestrial transect study on driving mechanism of vegetation
changes. Science in China Series D: Earth Sciences 51: 984-991
372
373
Zhao D, Wu S (2013) Responses of vegetation distribution to climate change in China. Theor.
Appl. Climatol. doi 10.1007/s00704-013-0971-4
374
375
Zhao MS, Neilson RP, Yan X (2002) Modelling the vegetation of China under changing climate.
ACTA Geographica Sinica 57: 38-45 (in Chinese)
376
377
Figure Captions
378
Figure 1 Spatial distributions of multi-model mean linear trends [(°C)/110a for temperature or
379
(mm/mo)/110a for precipitation from 1990-2100 in the RCP8.5 scenario] in annual mean
380
temperature (a) and precipitation (b), winter (DJF) mean temperature (c) and precipitation (d),
381
and summer (JJA) mean temperature (e) and precipitation (f) under RCP8.5 during 1990-2100.
382
Summer precipitation trends (f) are divided by 2 to meet the range of the color bar.
383
Figure 2 Spatial distribution of Köppen classes over China in the 1990 base year computed from
384
observed data (a), and derived from multi-model averaged projections for 2093 in RCP 2.6 (c),
385
RCP 4.5 (d), RCP 6 (e) and RCP 8.5 (f) scenarios. (g) and (h) are the same as (f) but with trends
386
of precipitation or temperature removed, respectively.
387
Figure 3 Time series for 1990 to 2093 (all values are 15-year centered averages) of (a)
388
percentages of China land area with changes in Köppen classes (either major types or subtypes)
389
from the 1990 base year, (b-e) the total area (%) occupied by each major climate type B, C, D
17
390
and E under RCP 2.6 (green or thin solid lines), RCP 4.5 (orange or thick dashed lines), RCP 6.0
391
(yellow or thin dashed lines) and RCP 8.5 (red or thick solid lines) scenarios, and (f-g) for the
392
RCP8.5 scenario, areas (%) in three sub-types of major climate type C and D. Temperate ocean
393
climate (Cwb) over Southwest China (SW) and Tibetan Plateau (TP), and two Continental winter
394
Dry climate types (Dwb and Dwc) over Northeast China (NE) and the Tibetan Plateau are
395
plotted separately in (f) and (g).
396
Figure 4 Maps of first emerging times, the year a specific grid box undergoes a human-induced
397
shift in climate zone from the 1990 base year in RCP2.6 (a), RCP 4.5 (b), RCP 6 (c) and RCP 8.5
398
(d) scenarios. White areas have the same climate class throughout 1990-2093.
399
18
400
Tables
401
402
403
Table 1. Models and specific CMIP5 runs (one run selected from each ensemble) used in this
study. “Hist” refers to 1950-2005 historical runs driven by observed forcing, and the “RCP”
runs start from a historical run and cover 2006-2100 using prescribed forcing.
Hist
√
√
RCP2.6
ACCESS1.0
ACCESS1.3
RCP4.5
√
√
RCP6.0
BCC-CSM1.1
√
BCCCSM1.1(m)
CCSM4
RCP8.5
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
CESM1(CAM5)
√
√
√
√
√
CMCC-CESM
√
CMCC-CM
√
CNRM-CM5
√
CSIRO-Mk3.6.0
√
√
√
√
√
√
√
√
√
CanESM2
√
√
√
√
FGOALS-g2
√
√
√
√
FIO-ESM
√
√
√
√
√
GFDL-CM3
√
√
√
√
√
GFDL-ESM2G
√
√
√
√
√
GFDL-ESM2M
√
√
√
√
√
√
√
404
405
19
GISS-E2-H
GISS-E2-R
HadGEM2CC
HadGEM2ES
INM-CM4
IPSLCM5A-LR
IPSLCM5A-MR
IPSLCM5B-LR
MIROCESM
MIROCESMCHEM
MIROC5
MPI-ESMLR
MPI-ESMMR
MRICGCM3
NorESM1M
Hist
√
√
RCP2.6
√
√
√
√
RCP4.5
√
√
RCP6.0
√
√
√
√
√
√
RCP8.5
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
406
407
Table 2. Köppen classes observed or projected over China, with climate criteria and
corresponding climate/vegetation types.
Type Subtype Long Name
A
Tropical
Tropical
Am
Monsoon
Aw
Tropical Savanna
B
Arid
BWh
Hot Arid
BWk
Temperate Arid
BSh
Hot Semi-Arid
Temperate SemiBSk
Arid
C
Temperate
Subtropical
Cfa
Humid
Subtropical
Cwa
Winter Dry
Temperate
Cwb
Oceanic
D
Continental
Continental
Dfa
Humid
Continental
Dfb
Humid
Dfc
Dwa
Dwb
Dwc
Dsb
Dsc
E
ET
408
409
410
411
412
413
414
415
416
Sub-polar Humid
Criteria
Tcold≥18
Vegetation cover
Pdry<60 & Pdry ≥100–MAP/25
Tropical Seasonal
Forest
Savanna
Pdry<60 & Pdry >100–MAP/25
MAP<10×Pthreshold
MAP<5×Pthreshold & MAT≥18
MAP<5×Pthreshold & MAT<18
MAP≥5×Pthreshold & MAT≥18
MAP≥5×Pthreshold & MAT<18
Thot>10 & 0<Tcold<18
Not (Psdry<40 & Psdry< Pwwet/3) or
(Pwdry<Pswet/10) & Thot≥22
Pwdry<Pswet/10 & Thot≥22
Pwdry<Pswet/10 & Thot<22 & Tmon10≥4
Thot>10 & Tcold≤0
Not (Psdry<40 & Psdry< Pwwet/3 (Ds)) or
(Pwdry<Pswet/10 (Dw)) & Thot≥22
Not (Ds) or (Dw) & Thot<22 & Tmon10≥4
Not (Ds) or (Dw) & Thot<22 & Tmon10<4
& Tcold>-38
Continental
Winter Dry
Continental
Winter Dry
Sub-polar Winter
Dry
Continental
Summer Dry
Sub-polar
Summer Dry
Polar
Pwdry<Pswet/10 & Thot≥22
Alpine Climate
Thot> 0
Pwdry<Pswet/10 & Thot<22 & Tmon10≥4
Pwdry<Pswet/10 & Thot<22 & Tmon10<4 &
Tcold>-38
Psdry<40 & Psdry< Pwwet/3 & Thot<22 &
Tmon10≥4
Psdry<40 & Psdry< Pwwet/3 & Thot<22 &
Tmon10<4 & Tcold>-38
Thot<10
Desert
Desert
Steppe
Steppe
Temperate Broadleaved
Evergreen Forest
Temperate Broadleaved
Evergreen Forest
Temperate Broadleaved
Evergreen Forest
Temperate Deciduous
Forest
Temperate and Boreal
Mixed Forest
Boreal Conifer Forest /
Woodland
Temperate Deciduous
Forest
Temperate / Boreal
Mixed Forest
Boreal Deciduous
Forest / Woodland
--Alpine Tundra / Polar
Desert
All precipitation amounts are mm and all temperatures are °C. MAP = mean annual precipitation,
MAT = mean annual temperature, Thot = temperature of the hottest month, Tcold = temperature of
the coldest month, Tmon10 = number of months where the temperature is above 10, Pdry =
precipitation of the driest month, Psdry = precipitation of the driest month in summer, Pwdry =
precipitation of the driest month in winter, Pswet = precipitation of the wettest month in summer,
Pwwet = precipitation of the wettest month in winter, Pthreshold = varies according to the following
rules (if 70% of MAP occurs in winter then Pthreshold = 2 x MAT, if 70% of MAP occurs in
summer then Pthreshold = 2 x MAT + 28, otherwise Pthreshold = 2 x MAT + 14). For China, summer
is defined as April through September and winter is October through March. (Peel et al., 2007)
20
417
Figures
418
419
420
421
422
423
Figure 1 Spatial distributions of multi-model mean linear trends [(°C)/110a for temperature or
(mm/mo)/110a for precipitation from 1990-2100 in the RCP8.5 scenario] in annual mean
temperature (a) and precipitation (b), winter (DJF) mean temperature (c) and precipitation (d),
and summer (JJA) mean temperature (e) and precipitation (f) under RCP8.5 during 1990-2100.
Summer precipitation trends (f) are divided by 2 to meet the range of the color bar.
21
424
425
426
427
428
Figure 2 Spatial distribution of Köppen classes over China in the 1990 base year computed from
observed data (a), and derived from multi-model averaged projections for 2093 in RCP 2.6 (c),
RCP 4.5 (d), RCP 6 (e) and RCP 8.5 (f) scenarios. (g) and (h) are the same as (f) but with trends
of precipitation or temperature removed, respectively.
22
429
430
431
432
433
434
435
436
437
Figure 3 Time series for 1990 to 2093 (all values are 15-year centered averages) of (a)
percentages of China land area with changes in Köppen classes (either major types or subtypes)
from the 1990 base year, (b-e) the total area (%) occupied by each major climate type B, C, D
and E under RCP 2.6 (green or thin solid lines), RCP 4.5 (orange or thick dashed lines), RCP 6.0
(yellow or thin dashed lines) and RCP 8.5 (red or thick solid lines) scenarios, and (f-g) for the
RCP8.5 scenario, areas (%) in three sub-types of major climate type C and D. Temperate ocean
climate (Cwb) over Southwest China (SW) and Tibetan Plateau (TP), and two Continental winter
Dry climate types (Dwb and Dwc) ov
23
438
439
440
441
Figure 4 Maps of first emerging times, the year a specific grid box undergoes a human-induced
shift in climate zone from the 1990 base year in RCP2.6 (a), RCP 4.5 (b), RCP 6 (c) and RCP 8.5
(d) scenarios. White areas have the same climate class throughout 1990-2093.
24