Download Recent and future changes in the combination of annual

INTERNATIONAL JOURNAL OF CLIMATOLOGY
Int. J. Climatol. (2016)
Published online in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/joc.4742
Recent and future changes in the combination of annual
temperature and precipitation throughout China
Jie Liu,a Haibo Du,a Zhengfang Wu,a* Hong S. He,a,b Lei Wanga and Shengwei Zonga
a
School of Geographical Sciences, Northeast Normal University, Changchun, China
b
School of Natural Resources, University of Missouri-Columbia, MO, USA
ABSTRACT: Climate involves different combinations of temperature and precipitation, and each year’s combination of
factors can be assigned a climatic year type (CYT; e.g. Warm-Humid). Describing the changes in the CYT provides more
information than describing the temperature or precipitation data alone. In this study, we defined nine CYTs using the
probability density function of annual temperature and precipitation. Recent and future spatiotemporal changes in CYT were
analysed using 507-station observational data and projected data obtained from the CMIP5 multi-model ensemble under
RCP2.6, RCP4.5, and RCP8.5 scenarios. China was divided into six subregions to analyse the spatiotemporal changes. Obvious
differences in spatial patterns among the various CYTs reflect the climate regime throughout China. The warmth-associated
CYTs (Warm-Humid, Warm-Dry, and Warm-Normal) mainly occur in West China (e.g. Southwest China). The cold-associated
CYTs (Cold-Humid, Cold-Dry, and Cold-Normal) dominate at high latitudes and high altitudes (e.g. Northeast China and the
Tibetan Plateau). The climate in China changed from cold to warm in the last half-century, accompanying the transformation
of Cold-Humid, Cold-Dry, and Cold-Normal before the early 1990s to Warm-Humid, Warm-Dry, and Warm-Normal from
the early 1990s onward. In the 21st century, the projected CYTs are mainly Warm-Humid, Warm-Dry, and Warm-Normal
in China. Warm-Humid dominates in West China, North China, and Northeast China. Warm-Dry is mainly projected in the
Yellow River Valley and South China. High-frequency Warm-Normal is projected in the Yellow River Valley. Warm-Humid is
projected to increase whereas Warm-Dry and Warm-Normal are projected to decrease from 2015 to 2099. All three CYTs are
projected to exhibit larger changes in trends under stronger versus weaker RCPs (RCP8.5 > RCP4.5 > RCP2.6). Compared
with temperature or precipitation data alone, CYTs provide more complete information on climate change and more accurately
characterize regional differences in climate throughout China.
KEY WORDS
combination of temperature and precipitation; climatic year type; CMIP5; temperature and precipitation; China
Received 6 December 2015; Revised 14 March 2016; Accepted 18 March 2016
1. Introduction
Changes in temperature and precipitation constitute some
of the most important concerns regarding global climate
change and are closely tied to natural ecosystems and
human society. Changes in temperature and precipitation
have altered the distribution of vegetation (Du et al., 2011;
Gottfried et al., 2012; Ge et al., 2014; Jarlan et al., 2014),
agriculture (Piao et al., 2010; Lobell et al., 2011), the
extent of glaciers (Scherler et al., 2011; O’Gorman, 2012;
Sorg et al., 2012), and sea level (Yin et al., 2009). These
changes have also caused more extreme weather and climate events (Choi et al., 2009; Yabi and Afouda, 2012;
Du et al., 2013b). Therefore, many studies have been conducted that examine changes in temperature and precipitation at different spatiotemporal scales (Wu et al., 2006;
Thompson et al., 2008; Srivastava et al., 2009; Choi et al.,
2011; Du et al., 2013a; Mengistu et al., 2014; Basha et al.,
2015; Sun et al., 2015).
* Correspondence to: Z. Wu, School of Geographical Sciences, Northeast
Normal University, No. 5268, Renmin Street, Changchun 130024, China.
E-mail: [email protected]
© 2016 Royal Meteorological Society
The average global temperature increased by approximately 0.85 ∘ C between 1880 and 2012, and Northern
Hemisphere mid-latitude land areas have exhibited
an overall increase in precipitation (IPCC, 2013). In
China, the change in annual mean temperature was
0.078 ± 0.027 ∘ C 10a−1 between 1906 and 2005 (Tang
et al., 2010), while annual mean precipitation declined
1–2 mm between 1960 and 2009 (Li et al., 2012). Many
studies of regional changes in temperature and precipitation have been conducted in China, e.g. studies of
temperature and precipitation changes in Northeast China
(Zhao et al., 2007), Southwest China (Fan et al., 2011),
North China (Ren et al., 2008), the Loess Plateau of
China (Li et al., 2010), the Tibetan Plateau (Dong et al.,
2012), South China (Fischer et al., 2011), and the Yangtze
River Valley (Hu et al., 2015). Although a single climate
variable (temperature or precipitation) can provide significant information, some limitations exist when drawing
conclusions based on one variable (Li and Zhu, 2015). The
climatic year type (CYT; which combines temperature
and precipitation) is an important method for researching
regional climate change and provides more complete
information than temperature or precipitation data alone.
J. LIU et al.
Figure 1. Study area and the 507 weather stations used in this study. The six subregions in China are Northeast China (NEC), North China (NC),
Northwest China (NWC), Southwest China (SWC), Yellow River Valley (YRV), and South China (SC).
The study of CYT can supplement the study of single
climate variable and can further illuminate the effects
of climate change on ecosystems. Moreover, assessing
the trends in terms of changes in CYTs can provide
an important understanding of overall climate change
characteristics.
In China, some studies of regional CYTs have been
conducted. For example, Shi et al. (2007) found that the
climate has changed from warm and dry to warm and
wet in Northwest China. Li et al. (2007) noted that the
trend of Cold-Warm and dry climate in recent decades may
be the main reason for the decline in the level of Lake
Qinghai in China. Li et al. (2004) found that vegetation
and soil carbon densities were high in warm and wet
regions, such as Southeast China and Southwest China.
However, the previous studies lack a clear definition of
different CYTs. A comprehensive scientific study of the
recent and future changes in CYTs throughout China was
needed. Therefore, this study aims to investigate the recent
and future changes in CYTs in China using long-term
data from 507 meteorological stations plus the output
of model simulations developed by the Coupled Model
Intercomparison Project phase 5 (CMIP5; Taylor et al.,
2011). The purposes of this study are (1) to define CYTs by
the probability density function (PDF) of temperature and
precipitation, (2) to investigate the recent spatiotemporal
changes in CYTs in China, and (3) to explore the future
changes in CYTs throughout China.
2.
2.1.
Data and methods
Data
Recorded data for daily precipitation (rainfall amount)
and daily mean temperature (Tmean) during the
period between 1961 and 2013 were obtained for 842
weather station locations throughout China from the
© 2016 Royal Meteorological Society
China Meteorological Data Sharing Service System
(http://cdc.nmic.cn/home.do). Data quality control, especially for data homogeneity tests, is important for detecting
climate changes. Data quality control efforts were comprised of two parts: (1) checking for erroneous data
(e.g. negative precipitation, days on which the maximum
temperature was less than or equal to the minimum temperature) and (2) searching for potential outliers outside a
range based on the 99th percentile during the period from
1961 to 1990 for precipitation, and defined as lying within
four standard deviations (std) of the climatological mean
of the value for the day, that is, the mean ± 4 × std for temperature (Alexander et al., 2006). These potential outliers
were checked manually. The erroneous data and identified
outliers were treated as missing values. For the homogeneity test, we used RHtestV4 software (Wang and Feng,
2013). We chose the reference series for the temperature
data using the approaches proposed by Xu et al. (2013).
For precipitation, homogeneous stations were identified
using the corresponding monthly precipitation data. However, due to the spatial heterogeneity of precipitation,
making adjustments to inhomogeneous data is a complex
and difficult task. Thus, the precipitation datasets were not
adjusted. After the quality control and homogeneity tests,
stations with less than 1% missing values were retained,
which left 507 high-quality stations; this remaining subset
of stations was used in this study (Figure 1).
For our projections of future climate change, we used
the output databases of seven CMIP5 models under
three representative concentration pathway (RCP) scenarios. Further details on these models can be found on
the CMIP5 website (http://cmip-pcmdi.llnl.gov/cmip5/
availability.html). Table 1 shows an overview of these
models, listing the associated institutions and the specific
atmospheric model component. The selected models
include both 20th-century climate simulations and
Int. J. Climatol. (2016)
CHANGES IN THE COMBINATION OF TEMPERATURE AND PRECIPITATION OVER CHINA
Table 1. Information on the twelve global climate models used in the present analysis.
Model name
Modelling centre
Australian Community Climate and
Earth-System Simulator, version 1.3
(ACCESS1-3)
Beijing Climate Center Climate System Model
version 1.1 (BCC-CSM1-1)
Beijing Climate Center Climate System Model
version 1.1 m (BCC-CSM1-1 m)
Canadian Earth System Model version 2
(CanESM2)
The Community Climate System Model version
4 (CCSM4)
Centre National de Recherches Météorologiques
Coupled Global Climate Model, version 5
(CNRM-CM5)
Institut Pierre Simon Laplace Climate Model
5A-Medium Resolution (IPSL-CM5A-MR)
Institute of Numerical Mathematics
Coupled Model, version 4.0 (INM-CM4)
Model for Interdisciplinary Research on
Climate-Earth System, version 5 (MIROC5)
Commonwealth Scientific and Industrial Research
Organization (CSIRO) and Bureau of Meteorology
(BOM), Australia
BCC, China
192 × 145
BCC, China
320 × 160
Canadian Centre for Climate Modeling and
Analysis (CCCma), Canada
The National Center for Atmospheric Research
(NCAR), USA
Centre National de Recherches Meteorologiques
and Centre Europeen de Recherche et Formation
Avancees en Calcul Scientifique, France
IPSL, France
128 × 64
144 × 143
Institute for Numerical Mathematics, Russia
180 × 120
Atmosphere and Ocean Research Institute (AORI),
National Institute for Environmental Studies
(NIES), Japan Agency for Marine-Earth Science
and Technology, Kanagawa (JAMSTEC), Japan
MPI, Germany
256 × 128
MRI, Japan
320 × 160
Norwegian Climate Centre, Norway
144 × 96
Max Planck Institute Earth System Model-Low
Resolution (MPI-ESM-LR)
Meteorological Research Institute Coupled
General Circulation
Model version 3 (MRI-CGCM3)
The Norwegian Earth System Model version 1
with Intermediate Resolution (NorESM1-M)
21st-century climate projections under low-, medium-,
and high-emissions scenarios (the RCP2.6, RCP4.5, and
RCP8.5 scenarios). The three scenarios predict radiative forcing to peak at 2.6, 4.5, and 8.5 W m−2 by 2100,
respectively (Riahi et al., 2011; Thomson et al., 2011; Van
Vuuren et al., 2011). The CMIP5 models have different
spatial resolutions. Data from the different general circulation models were bilinearly interpolated into a common
1∘ × 1∘ grid before further processing. The projections are
relative to a base period between 1986 and 2005.
We analysed the spatial patterns of changes in temperature and precipitation by dividing China into six
subregions (Xu et al., 2011). The subregions and the corresponding number of remaining station are: Northeast
China (NEC, 82 stations), North China (NC, 108 stations),
Northwest China (NWC, 88 stations), Southwest China
(SWC, 54 stations), the Yellow River Valley (YRV, 132 stations), and South China (SC, 43 stations; Figure 1). These
subregions were determined according to societal and geographic conditions.
2.2. Method
2.2.1. Definition of CYT
In the China National Standard of warm winter grade,
China Meteorological Administration (2008) classified
winter into warm winter, normal winter, and cold winter
by evenly dividing the probability density of winter mean
© 2016 Royal Meteorological Society
Resolution (lon × lat)
128 × 64
288 × 192
256 × 128
192 × 96
temperature into three grades (trichotomy-based probability density method). The prerequisite of the method is
that winter mean temperature follows normal distribution.
Similar to winter mean temperature, the long-term annual
mean temperature and precipitation follow a normal distribution pattern (Zhang et al., 2010). Therefore, the annual
cold, warm, dry, humid, and normal types were classified
by the trichotomy-based PDF of annual temperature and
precipitation. The calculation is based on a certain 30-year
climatological period. In the period of 1961–2013, we
chose the middle three decades (1971–2000) as the base
period in this study. For each station, the PDF of annual
mean temperature and precipitation in the base period is
as follows:
[
]
(x − 𝜇)2
1
(1)
PDF (x) = √ exp −
2𝜎 2
𝜎 2𝜋
where x is the annual mean temperature or precipitation; 𝜇
and 𝜎 are the mean value and standard deviation of x in the
base climatological period, respectively.
Let
x−𝜇
(2)
X=
𝜎
Then, X obeys the standard normal distribution:
( 2)
X
1
PDF (X) = √ exp −
2
2𝜋
(3)
Int. J. Climatol. (2016)
J. LIU et al.
Figure 2. Graph shows the probability for cold/dry, normal, and warm/humid conditions, based on the PDF using temperature and precipitation as
variables for two stations. One station is Tuoli station located in northern China (the first row, 83.6∘ E, 45.93∘ N). Another station (Simao station)
locates in southern China (the second row, 100.97∘ E, 22.78∘ N).
Table 2. Definition of nine CYTs.
Temperature
Precipitation
Humid
Dry
Normal
Warm
Cold
Normal
Warm-Humid
Warm-Dry
Warm-Normal
Cold-Humid
Cold-Dry
Cold-Normal
Normal-Humid
Normal-Dry
Normal-Normal
The probability density of the standard normal distribution is evenly divided into three classes at X = ±0.43
(that is, the probability of each is 33.3%). Therefore,
x = 𝜇 ± 0.43𝜎 are the two critical values evenly dividing
the probability density of annual temperature and precipitation into three classes. The corresponding thresholds of different types mentioned above are 𝜇 − 0.43𝜎 and
𝜇 + 0.43𝜎 (Figure 2).
For temperature, each year was determined to be cold,
normal, or warm based on the calculation above. For
precipitation, each year was defined as dry, normal, or
humid. Therefore, a total of nine CYTs were developed
by combining temperature and precipitation conditions
(Table 2). We used this method to calculate the CYTs for
each station.
2.2.2. Calculation of temperature and precipitation
anomalies
for annual temperature, and
x=
© 2016 Royal Meteorological Society
(4)
(5)
for annual precipitation, where x is the annual temperature
and precipitation and 𝜇 is the mean value of x in the based
period.
2.2.3. Calculation of the occurrence ratio of CYTs
Each subregion contains different number of station or
grid. To eliminate the effects of such different number
on the calculation of trends in the frequency of CYTs
among subregions, we calculated the occurrence ratio of
each CYT for each subregion. The occurrence ratio of
each CYT was defined as the percentage of the number of
stations (grids) of occurrence CYT in the number of total
stations (grids) in each subregion:
The anomaly (x) was calculated by
x=x−𝜇
x−𝜇
𝜇
𝜆i =
ni
× 100%
N
(6)
Int. J. Climatol. (2016)
CHANGES IN THE COMBINATION OF TEMPERATURE AND PRECIPITATION OVER CHINA
where 𝜆i is the ith-year occurrence ratio of each CYT in
each subregion; ni is the number of the stations (grids) of
occurrence CYT in the ith-year, and N is the number of
total stations (grids) in each subregion.
2.2.4. Calculation of trend
We calculated the trends in variables using the linear tendency method, with the ordinary least squares technique
(Ma et al., 2015). In this study, the variables contain annual
temperature and precipitation, annual temperature and precipitation anomalies, and annual occurrence ratio of CYTs
in the six subregions and all of the China. The formula is
as follows
yi = a + bti
(7)
where yi is the time series of variables; ti is the corresponding time series (1961–2013 for the recent change
and 2015–2099 for the projected change); the regression
coefficient b represents the linear trend, and a is a constant.
For the trends in annual temperature and precipitation
(and both anomalies), we firstly calculated the regional
mean values, and then calculated the regional trends. The
units were ∘ C 10a−1 and mm 10a−1 (∘ C 10a−1 and % 10a−1 )
for temperature and precipitation (and both anomalies),
respectively. For the CYTs, we firstly calculated the
regional occurrence ratio of CYT in each year, and then
calculated the trends in occurrence ratio. The unit was
% 10a−1 . The significance level (p value) for trend calculation is 0.05 in this study.
2.2.5. Evaluation of the performance of CMIP5 models
To evaluate the performance of CMIP5 models at regional
scale, we compared the root-mean-square errors (RMSEs)
of CYTs calculated by the observations and by the models.
The observation temperature and precipitation data of the
507 stations over China were interpolated into 1∘ × 1∘
grid. If there is not any observation station in a 2∘ range
around a grid, this grid was set as missing data for both
gridded observation data and model data. The RMSE of
each CYT between the observation data and each model
was calculated as
√⟨
⟩
RMSE =
(8)
(X − Y)2
where X and Y denote the simulated and observed CYTs,
respectively; the angular bracket represents spatial averaging over China on the 1∘ × 1∘ grid. Then we defined the
relative model error (RMSE′ ) for each model as
RMSE′ =
RMSE − RMSEmean
RMSEmean
(9)
where RMSEmean is the mean of RMSEs for all individual
models.
RMSE′ provides a metric to measure a model’s performance relative to the other models, with respect to
the observation (Sillmann et al., 2013). Negative (positive) RMSE′ indicates that the corresponding model
performs better (worse) than the majority (50%) of
models. Models that performed relatively well were
© 2016 Royal Meteorological Society
Figure 3. Evaluation of the performance of CMIP5 models. The values
represent the RMSE’ among individual models. Negative (positive) value
indicates that the corresponding model performs better (worse) than the
majority (50%) of models. The mean value of the RMSE’ of all CYTs
was marked as ‘All CYTS’ for each model.
different for different CYTs (Figure 3). It’s hard to select
the models that generally perform well for all of the
CYTs. Therefore, we evaluated the performance of each
model by calculating the mean value of the RMSE’ of
all CYTs, which was marked as ‘All CYTS’ in Figure 3.
Generally, seven models performed relatively well,
including BCC-CSM1-1, CanESM2, MIROC5, CCSM4,
IPSL-CM5A-MR, MPI-ESM-LR, and NorESM1-M (see
Table 1 for information on models).
The remaining seven-model ensemble-averaged precipitation was expected to increase throughout China between
2015 and 2099 by 0.48% 10a−1 for RCP2.6, 0.92% 10a−1
for RCP4.5, and 1.66% 10a−1 for RCP8.5 (relative
to 1986–2005) (Figure 4(a)). The expected warming
trends throughout China between 2015 and 2099 were
0.05 ∘ C 10a−1 for RCP2.6, 0.24 ∘ C 10a−1 for RCP4.5, and
0.62 ∘ C 10a−1 for RCP8.5. The simulation and projections
by the seven-model ensemble in this study were similar to
the results by Xu and Xu (2012) using 11-model ensemble
and by Tian et al. (2015) using 22-model ensemble. This
indicated that the remaining seven models were enough to
project the changes in CYTs for China. We also compared
the observed historical annual mean temperature and
precipitation data with the simulation by the seven-model
ensemble (Figure 4). There were differences between the
observation and the model simulation from 1961 to 2005,
which were greater for precipitation than temperature.
However, the differences in the change trends were small
between the observation and simulation. The trends in
simulated (purple line) historical temperature and precipitation were 0.24 ∘ C 10a−1 (p < 0.05) and −0.022% 10a−1
(p > 0.05) throughout China from 1961 to 2005 (relative
to 1986–2005), respectively. In the same period, the
trends in observed (black line) historical temperature
Int. J. Climatol. (2016)
J. LIU et al.
Figure 4. Time series of the annual average (a) surface precipitation
anomalies (units: %) and (b) air temperature anomalies (units: ∘ C)
throughout China between 1961 and 2099, showing projections from
the multi-model ensemble (relative to 1986–2005). The black dotted
line represents historical observations; the purple dotted line represents
historical simulations; and the green, red, and blue lines represent the
RCP2.6, RCP4.5, and RCP8.5 scenarios, respectively.
and precipitation were 0.234 ∘ C 10a−1 (p < 0.05) and
−0.107% 10a−1 (p > 0.05) for China, respectively. The
differences were 0.085% 10a−1 for precipitation and
0.006 ∘ C 10a−1 for temperature. Such small differences
between observed and simulated data indicated, on the
other hand, that the multi-model ensemble projections
in this study performed relatively well. The projected
changes in CYTs based on the ensemble-averaged data
performed well and outperformed individual models
because some of the systematic errors in individual
models were cancelled out in the multi-model mean.
3.
3.1.
Results
Observed changes in CYTs
3.1.1. Temporal changes in observed temperature and
precipitation
The change trends in precipitation anomalies were clearly
different among the various subregions (Table 3). Three
subregions, NEC, NWC, and SC, showed positive trends
with the largest value of 9.7 mm 10a−1 recorded in SC.
Precipitation decreased in the other three subregions,
NC, SWC, and YRV, with the largest amplitude of
(−12 mm 10a−1 ) recorded in SWC. However, only two
trends were significant (p < 0.05): 7.1 mm 10a−1 for NWC
and −12 mm 10a−1 for SWC. In total, the decreasing
amplitudes were larger than the increasing amplitudes.
Therefore, the annual mean precipitation throughout
China generally showed a non-significant negative trend
(−1.5 mm 10a−1 ) during the period from 1961 to 2013.
In contrast with changes in precipitation, significant positive trends for temperature were observed for all six subregions (Table 3). The coldest region was NEC, which is
the most sensitive region to global climate change, and it
showed the largest changes (0.3 ∘ C 10a−1 ) among all of
the subregions. SC exhibited the maximum mean annual
temperature in China, and it was comparatively the most
stable region (with the trend of 0.147 ∘ C 10a−1 ). In general,
colder regions are more sensitive to global climate change,
and vice versa. From most sensitive to most stable, the
order of the regions was NEC, NWC, NC, SWC, YRV, and
SC. In other words, the sensitivity to global climate change
varies from a low level in low latitudes (and low altitudes)
to a high level in high latitudes (and high altitudes). On
average, the trend in temperature anomalies for all of China
was an increase of 0.233 ∘ C 10a−1 between 1961 and 2013.
3.1.2. Observed spatial patterns in the frequency of
CYTs
The spatial distribution of the occurrence frequency for
the nine CYTs throughout China from 1961 to 2013 is
shown in Figure 5. Warm-Humid mainly occurs in SWC,
especially in the Tibetan Plateau region, which is few
in NEC. Warm-Dry dominates in SWC, YRV, SC, and
western NEC; this CYT is rare in the Tibetan Plateau
region. More Warm-Normal occurs commonly in west
China, but is less commonly distributed in east China.
Cold-Humid is mostly distributed in the NEC and NC
regions, and that CYT occurs least often in west and southeast China. Cold-Dry mainly appears at high latitudes and
high altitudes (NEC, NWC, and the Tibetan Plateau). The
NC and YRV regions have few occurrences of Cold-Dry.
Cold-Normal is distinctly concentrated in the NEC region.
Normal-Humid mostly occurs in the NC and SC regions.
Few instances of Normal-Dry occur in the SWC region.
For Normal-Normal, non-evident spatial assembly patterns were observed.
3.1.3. Observed temporal changes in the frequency of
CYTs
Table 4 represents the trends in the average occurrence
rate (the total number of occurrences divided by the
Table 3. Trends in annual mean temperature and precipitation for the six subregions in China between 1961 and 2013.
Precipitation (mm/10a)
Temperature (∘ C/10a)
NEC
NC
NWC
SWC
YRV
SC
China
1.4
0.3*
−10
0.267*
7.1*
0.297*
−12*
0.221*
−1.5
0.154*
9.7
0.147*
−1.5
0.233*
The symbol ‘*’ indicates a significant trend (p < 0.05).
© 2016 Royal Meteorological Society
Int. J. Climatol. (2016)
CHANGES IN THE COMBINATION OF TEMPERATURE AND PRECIPITATION OVER CHINA
Figure 5. Spatial distribution of the occurrence frequency for the nine CYTs throughout China between 1961 and 2013.
Table 4. Trends in the occurrence ratio of the nine CYTs for all subregions in China between 1961 and 2013.
NEC
NC
NWC
SWC
YRV
SC
WarmHumid
WarmDry
ColdHumid
5.5*
6.1*
5.6*
4.8*
6.4*
3.7*
4.1*
4.2*
4.1*
5.4*
4.6*
6.6*
−3*
−3*
−3*
−3.8*
−3.3*
−3.7*
ColdDry
−5.8*
−5.6*
−5.8*
−5.6*
−5.6*
−5.5*
WarmNormal
4.8*
5.2*
4.8*
4.5*
5.3*
4.3*
NormalDry
ColdNormal
−1.8
−2.3*
−1.9*
−1
−2.7*
−0.5
−5.2*
−5*
−5.1*
−6.2*
−4.7*
−6.8*
NormalHumid
NormalNormal
2.3*
1.7*
2.2*
2.8*
1.2
2.6*
−0.9
−1.3
−0.9
−0.9
−1.3
−0.8
The occurrence ratio is equal to the number of occurrences divided by the number of total stations for each region. The unit is %/10a. The symbol
‘*’ indicates a significant trend (p < 0.05).
number of stations for each region) of the nine CYTs
throughout China between 1961 and 2013. For all regions,
the Warm-Humid, Warm-Dry and Warm-Normal CYTs
showed significant positive trends (p < 0.05). Instances of
Normal-Humid also increased in all subregions, but that
trend was not significant (p > 0.05) in the YRV subregion.
The fastest changes for these four CYTs occurred in
the YRV, SC, YRV, and SWC subregions, respectively,
which are mainly in southern China. The other five CYTs,
Cold-Humid, Cold-Dry, Cold-Normal, Normal-Dry, and
Normal-Normal, showed decreasing trends. The negative
trends in Cold-Humid, Cold-Dry, and Cold-Normal were
significant (p < 0.05) in all subregions. The decrease
in Normal-Dry was significant (p < 0.05) only in the
NC, NWC, and YRV subregions. The Normal-Normal
CYT showed a non-significant decreasing trend in all
subregions. The largest decreases in occurrences of
Cold-Humid, Cold-Dry, and Cold-Normal occurred in
SWC, NEC and NWC, and SC, respectively.
The annual changes in the occurrence rate of the nine
CYTs throughout China between 1961 and 2013 are
shown in Figure 6. The changes are different among
the various CYTs. The CYTs associated with cold
conditions (Cold-Humid, Cold-Dry, and Cold-Normal)
mainly dominated in the period before the early 1990s,
whereas the CYTs associated with warm conditions
(Warm-Humid, Warm-Dry, and Warm-Normal) occurred
comparatively less frequently during this period. In
© 2016 Royal Meteorological Society
contrast, Warm-Humid, Warm-Dry, and Warm-Normal
significantly increased from the early 1990s onward. The
cold-associated CYTs became rare during this period.
Normal-Dry also occurred mainly before the mid-1990s.
The frequency of occurrence for the Normal-Humid CYT
was less than that of Normal-Dry during the early period,
but this trend reversed from the late 1990s onward. The
annual variability of Normal-Humid and Normal-Normal
was relatively small, showing a non-significant negative
trend for Normal-Normal and a very weak decreasing
trend for Normal-Humid. In the context of global warming, the climate throughout China changed from cold
to warm during the last half-century; this accompanies
the transition from higher occurrences of Cold-Humid,
Cold-Dry, and Cold-Normal before the early 1990s to
higher occurrences of Warm-Humid, Warm-Dry, and
Warm-Normal, respectively, from the early 1990s onward.
An abrupt detection by the Mann-Kendall test (Mann,
1945; Kendall, 1975) showed that the annual mean temperature experienced a significant abrupt warming at the
early 1990s, whereas not any abrupt change was found
for the annual mean precipitation between 1961 and
2013 (figure is not shown for brevity). The warm- and
cold-associated CYTs experienced an abrupt change at the
early 1990s. Therefore, the early 1990s were the transition
period for these CYTs. Similar changing characteristics
were also observed for the six subregions (which are not
shown in this paper).
Int. J. Climatol. (2016)
J. LIU et al.
projects a higher frequency of Warm-Humid and a lower
frequency of Warm-Dry and Warm-Normal.
3.2.2. Projected temporal changes in the frequency of
CYTs
Figure 6. Annual changes in the average occurrence rate of the nine
CYTs throughout China between 1961 and 2013.
3.2.
3.2.1.
CYTs
Projected changes in CYTs
Projected spatial patterns in the frequency of
Figure 7 shows the spatial distribution of the projected frequency of occurrence for the Warm-Humid, Warm-Dry,
and Warm-Normal CYTs throughout China between 2015
and 2099 for each of the three RCP scenarios. According
to the simulated projections, future CYTs are expected to
be mainly Warm-Humid, Warm-Dry, and Warm-Normal.
Therefore, in this study, we only analysed the changes
in these three CYTs. The spatial patterns of the three
CYTs are similar among the three RCP scenarios. A high
frequency of occurrence of Warm-Humid is mainly projected in West China, NC, and NEC for all three RCP
scenarios, whereas a low frequency for Warm-Humid is
mostly projected in YRV and SC. In contrast, a high
frequency of Warm-Dry is projected to occur in YRV
and SC, especially in the southeast coastal areas, but a
low frequency of Warm-Dry is mostly projected in West
China, NC, and NEC, according to the three RCP scenarios. For Warm-Normal, a high frequency is projected
in YRV, while a low frequency is projected in east NWC
and SWC. The differences in projected frequency for these
three CYTs among the three RCP scenarios were readily apparent in SC and NEC. In these two subregions, the
low scenario model (RCP2.6) projects fewer occurrences
of Warm-Humid and more occurrences of Warm-Dry and
Warm-Normal; but the high scenario model (RCP8.5)
© 2016 Royal Meteorological Society
Table 5 shows the trends in the projected average occurrence rate (the total number of occurrences divided
by number of the grid for each region) of the three
CYTs for all subregions between 2015 and 2099, as
modelled by the three RCP scenarios. For all subregions under the three RCP scenarios, the frequency of
Warm-Humid is projected to increase while Warm-Dry
and Warm-Normal decrease. Most of the changes are
significant (p < 0.05). The most obvious trends for WarmHumid, Warm-Dry and Warm-Normal were: 6.6% 10a−1
in NEC and YRV for RCP8.5, −5.1% 10a−1 in YRV for
RCP8.5 and −4.9% 10a−1 in NC for RCP8.5. In contrast,
the least-obvious trends were: 1.6% 10a−1 in NWC for
RCP4.5; 0.6% 10a−1 in NWC for RCP8.5; and 0.3% 10a−1
in SC for RCP2.6.
Figure 8 shows the annual changes in the projected
occurrence rates of CYTs throughout China between
2015 and 2099, as modelled by the three RCP scenarios. Warm-Humid is expected to increase significantly
(p < 0.05) under all three RCPs. The projected increase
under the RCP8.5 scenario was approximately twice that
of RCP2.6. All three RCPs projected significant (p < 0.05)
decreasing trends in the occurrences of Warm-Dry and
Warm-Normal. For Warm-Dry, the negative trends are similar among the different RCPs. However, under the RCP8.5
scenario, occurrences of Warm-Normal are expected to
decrease more quickly than for RCP4.5 and RCP2.6.
For all three RCP scenarios, the projected change in
Warm-Humid is more obvious than that in Warm-Dry and
Warm-Normal. In the future, the Warm-Humid CYT may
dominate in China, indicating that China may become
increasingly warm and humid. A few small differences
among the six subregions are projected, but the trends for
the six subregions are generally similar to those over all of
China (which are also not shown in this paper).
3.3. Comparison of single- and dual-variable CYTs
To demonstrate the difference between single-variable
CYTs (temperature or precipitation alone) and the more
sophisticated dual-variable CYTs (combined temperature
and precipitation), we first compared spatial distribution.
Figure 9 depicts temperature and precipitation separately;
it shows the spatial distribution of temperature-associated
CYTs (a–c: Warm year, Cold year, and Normal temperature year) and precipitation-associated CYTs (d–f: Humid
year, Dry year, and Normal precipitation year) between
1961 and 2013 throughout China. We found numerous
differences in the spatial distribution of the single-variable
CYTs in Figure 9 and the dual-variable CYTs as shown
in Figure 5. Warm years most often occurred in the SWC
and YRV subregions (Figure 9(a)). In comparison, a high
frequency of Warm-Humid mainly occurs in the Tibetan
Plateau region, but Warm-Dry is rare in this region
Int. J. Climatol. (2016)
CHANGES IN THE COMBINATION OF TEMPERATURE AND PRECIPITATION OVER CHINA
Figure 7. Spatial distribution of the projected frequency of the Warm-Humid (first column), Warm-Dry (second column), and Warm-Normal (third
column) CYTs throughout China between 2015 and 2099 for the RCP2.6 (first row), RCP4.5 (second row), and RCP8.5 (third row) scenarios.
Table 5. Projected trends in the occurrence ratio for the Warm-Humid, Warm-Dry, and Warm-Normal CYTs for all subregions from
2015 to 2099 under the RCP2.6, RCP4.5, and RCP8.5 scenarios.
Warm-Humid
NEC
NC
NWC
SWC
YRV
SC
Warm-Dry
RCP2.6
RCP4.5
RCP8.5
2.9*
2
1.7*
3.3*
4.6*
1.8
5.2*
6.2*
1.6
5.6*
3.9*
3.8*
6.6*
7.5*
2*
5.9*
6.6*
4.1*
Warm-Normal
RCP2.6
RCP4.5
RCP8.5
RCP2.6
RCP4.5
RCP8.5
−1.2*
−1.3*
−0.9
−1.8*
−2.8*
−1.5
−2.3*
−3.5*
−0.4
−2.1*
−3.4*
−2.3*
−2.5*
−2.6*
−0.6*
−1.7*
−5.1*
−2*
−1.7*
−0.7
−0.8
−1.5*
−1.9*
−0.3
−2.9*
−2.7*
−1.2*
−3.6*
−0.5
−1.5*
−4.1*
−4.9*
−1.4*
−4.2*
−1.5*
−2*
The occurrence ratio is equal to the number of occurrences divided by the number of total grids for each region. The unit is %/10a. The symbol ‘*’
indicates that the change trend is significant (p < 0.05).
(Figure 5). Cold years were mostly observed in NEC
(Figure 9(b)), but Cold-Humid occurs mainly in the southeast part of NEC and rarely in northwest regions of NEC
(Figure 5). The distribution of Cold-Dry is opposite to that
of Cold-Humid. Humid years are distinctly concentrated
in the SWC and NC subregions (Figure 9(d)). However,
Warm-Humid occurs mainly in the Tibetan Plateau region
(SWC) and is relatively rare in NC. Cold-Humid is rare
in SWC but mainly occurs in NEC (Figure 5). Dry years
occurred more frequently than any other CYT in China
between 1961 and 2013, and that CYT was mainly distributed in the NEC, NWS, and YRV subregions, plus a
portion of the SWC region (Figure 9(e)). The distribution
of Warm-Dry is similar to that of the Dry year CYT, but
Cold-Dry is rare in YRV (Figure 5).
These differences between the spatial distribution of
the single-variable CYTs and the dual-variable CYTs
show that the latter more accurately characterize regional
differences in climate throughout China and provide more
complete information than temperature or precipitation
data alone.
Temporal analysis also demonstrates differences
between the single-variable CYTs and the dual-variable
CYTs. Single-variable temperature or precipitation CYTs
provide only limited information (Figure 10). Throughout
© 2016 Royal Meteorological Society
Figure 8. Annual changes in the average occurrence rate of the projected
Warm-Humid, Warm-Dry, and Warm-Normal CYTs throughout China
between 2015 and 2099 under the three RCP scenarios.
Int. J. Climatol. (2016)
J. LIU et al.
Figure 9. Spatial distribution of the occurrence frequency for single-variable CYTs, which separately analyse (a)–(c) temperature and (d)–(f)
precipitation throughout China between 1961 and 2013.
China between 1961 and 2013, the significant trends
(p < 0.05) in the average frequency of Warm year, Cold
year, and Normal temperature year are 14.8% 10a−1 ,
−11.7% 10a−1 , and −3.1% 10a−1 , respectively. The average frequency for Warm year (14.8% 10a−1 ) actually
contains 4.6% 10a−1 for Warm-Humid, 5.6% 10a−1 for
Warm-Dry, and 4.7% 10a−1 for Warm-Normal (Figure 6).
However, these more comprehensive results could not
be derived if temperature data is analysed alone. Similarly, valuable information is lost if precipitation data is
used alone. Throughout China, the average frequency of
Humid year, Dry year, and Normal precipitation year are
0.5% 10a−1 , 0.1% 10a−1 , and −0.7% 10a−1 , respectively,
none of which are significant trends (p > 0.05). When
Humid year (0.5% 10a−1 ) is examined using dual-variable
CYTs, as shown in Figure 6, significant trends emerge.
The average frequency of the Warm-Humid CYT, at
4.6% 10a−1 , is a significant trend, as is the average frequency of the Cold-Humid CYT, at −3.7% 10a−1 . But
the trend for the Normal-Humid CYT, at −0.3% 10a−1 ,
is not significant. Looking at humid-associated CYT, if
one is using only single-variable CYT, then the HumidCYT would simply be treated as an insignificant weak
trend of 0.5% 10a−1 between 1961 and 2013 for all of
China, which is misleading. The dual-variable CYTs that
combine temperature and precipitation (Warm-Humid,
Dry-Humid, and Normal-Humid) provide more in-depth
and accurate information on climate change in China.
4.
Discussion
Climate has changed from cold to warm over the vast
majority of regions in the world, and from humid to
dry in eastern United States, southeastern South America, northern Europe, and northeastern Australia and dry
to humid in southern Europe, western Africa, and eastern
Asia (IPCC, 2013). However, separate analysis of temperature or precipitation cannot characterize the coupled
relationship of these two climate factors. For the joint
analyses of temperature and precipitation, there are many
sophisticated indices, such as various moisture indices
and drought indices. These climate indices were widely
used to quantify climate variability (Grundstein, 2009;
© 2016 Royal Meteorological Society
Figure 10. Annual changes in the average occurrence rate for two separate single-variable CYTs, (a) temperature and (b) precipitation, throughout China between 1961 and 2013.
Vicente-Serrano et al., 2010), to assess drought severity
(Gunda et al., 2016; Zhang and Zhang, 2016; Zhang et al.,
2016), to do climatic regionalization (Fraser et al., 2013;
Liu and Zhai, 2014), and commonly led to a synthetic
result. However, we usually cannot obtain the temperature and precipitation situation from such a result. For
example, if the moisture index MI (MI = P/WI, where P is
the annual precipitation, and WI is the sum of the >5 ∘ C
monthly average temperature, Xu, 1985) increased in a
certain period for a certain region, we could not know
which of the cases (precipitation increased, or temperature decreased, or both increased, but the increasing rate of
precipitation was greater than that of temperature) caused
the result. By contrast, we can easily obtain the change
situation of temperature and precipitation from the CYT
results. Moreover, CYT can directly reflect the allocations
of temperature and precipitation, which is important for
climate change research (e.g. for simulating/assessing climate change, crop growth, and vegetation distribution).
Int. J. Climatol. (2016)
CHANGES IN THE COMBINATION OF TEMPERATURE AND PRECIPITATION OVER CHINA
Figure 11. The standard deviation of the total frequencies of Warm-Humid (first column), Warm-Dry (second column), and Warm-Normal (third
column) in 2015–2099 among individual models for each grid under the RCP2.6 (first row), RCP4.5 (second row), and RCP8.5 (third row) scenarios.
Several regional CYT studies have been conducted in
China. The results, such as an increase in Warm-Humid in
NWC (Shi et al., 2007) and a dominating Warm-Humid
in SWC and SC (Li et al., 2004), are in agreement with
our findings. Shi et al. (2007) found an increasing trend
in Warm-Humid based on increasing temperature and
precipitation. However, their definitions were not clear
as to how to determine when one CYT would change to
another and the number of years over which a certain CYT
occurs. A Warm-Cold or Humid-Dry climate in a certain
station/grid is relative to the time series; thus, a certain
CYT must be determined based on a specific standard.
For example, both temperature and precipitation increased
in NEC during 1961–2013 (Table 3), which does not
signify that NEC has only experienced the Warm-Humid
CYT during this period. Regarding this point, two aspects
should be considered: (1) The increase in temperature
indicates that the temperature changed relatively from
colder to warmer. Therefore, the CYT should change
from a cold-associated to a warmth-associated CYT. The
change in CYT in association with precipitation was similar to the change related to temperature. (2) Changes in
temperature and precipitation trends fluctuated, especially
for precipitation, for which the trends were generally not
significant. An increase in the annual precipitation series
may lead to many relatively dry years. There is not only
the Humid-associated CYT but also Dry-associated CYT
in such an annual precipitation series. Therefore, a scientific and clear definition for CYT is needed. The method
used to define the CYT in this study references the China
National Standard of warm winter grades (China Meteorological Administration, 2008). The trichotomy-based
PDF, using variables of annual temperature and precipitation, can clearly divide the annual temperature and
precipitation time series into cold/normal/warm and
dry/normal/humid years. The definition makes the CYTs
comparable among different spatial and temporal scales.
© 2016 Royal Meteorological Society
Hence, the spatiotemporal changes and comparisons of
CYTs over different subregions are reliable in this study.
Uncertainties in the projected results produced by the
model should be noted. The standard deviation of the
multi-model projected results was a metric to quantify the
uncertainties (Tian et al., 2015). The greater standard deviation indicates a greater difference among the multi-model
projections. That is, greater standard deviation denotes
greater uncertainty, and vice versa. We calculated the standard deviation of the total frequencies of Warm-Humid,
Warm-Dry, and Warm-Normal in 2015–2099 among individual models for each grid under the three scenarios
(Figure 11). Warm-Humid had greater uncertainties than
Warm-Dry and Warm-Normal. Warm-Normal had the
smallest uncertainties among the three CYTs. For the
subregions, the uncertainties in NEC, NC, and SC were
relatively small, especially for NEC. NWC, SWC, and
YRV had relatively high uncertainties. The uncertainties
in NWC were the largest among different subregions. The
uncertainties in the projections were larger under high
forcing scenario than those under low forcing scenario
(RCP8.5 > RCP4.5 > RCP2.6), which was similar to the
results by Tian et al. (2015) and Zhou et al. (2014). We
used the inter-model discrepancy to analyse the uncertainty of projections. However, other analysis of the uncertainties of the projections, such as the models’ resolution,
physical processes or other processes, was not discussed.
Further study should focus on the adoption of a more
rational and scientific approach to reduce the uncertainties
as well as on using the enhanced projections to analyse
changes in CYT.
5.
Conclusions
In this study, we examined historical and projected
changes in CYT based on the observed daily temperature
and precipitation data throughout China between 1961 and
Int. J. Climatol. (2016)
J. LIU et al.
2013, as well as data simulated by the CMIP5 multi-model
ensemble under the RCP2.6, RCP4.5, and RCP8.5 forcing
scenarios. Compared with earlier studies, this study is
more comprehensive and provides more detailed information on the recent changes in CYT for China and
its subregions. We also analysed projections for future
changes in CYT for China and its subregions based on
a new generation of multiple climate models. The main
findings are summarized below:
1. Spatial patterns are distinct among the various CYTs,
which clearly reflects China’s climate regime. Between
1961 and 2013, a high number of occurrences of CYTs
associated with warm conditions (Warm-Humid,
Warm-Dry, and Warm-Normal) mainly occurred
in West China (e.g. SWC), whereas a high frequency of the CYTs associated with cold conditions
(Cold-Humid, Cold-Dry, and Cold-Normal) dominated at high latitudes and high altitudes (e.g. NEC
and the Tibetan Plateau). Warm-Humid, Warm-Dry,
and Warm-Normal showed significant positive trends
(p < 0.05) for all subregions, whereas Normal-Humid
showed an insignificant positive trend. The other
five CYTs, Cold-Humid, Cold-Dry, Cold-Normal,
Normal-Dry, and Normal-Normal, showed decreasing trends. Throughout China, the climate has
generally changed from cold to warm in the last
half-century, accompanying the transition of predominantly Cold-Humid, Cold-Dry, and Cold-Normal
before the early 1990s to Warm-Humid, Warm-Dry,
and Warm-Normal, respectively, from the early
1990s onward. Therefore, the early 1990s marked the
transition period or period of mutation for these CYTs.
2. The projected CYTs are mainly Warm-Humid,
Warm-Dry, and Warm-Normal throughout China
for the years 2015 to 2099 for all three RCPs.
Warm-Humid is projected to dominate in West China,
NC, and NEC. Warm-Dry is mainly projected to occur
in YRV and SC. A high frequency of Warm-Normal
is projected in YRV. The difference in the projected
frequency of the three CYTs among the three RCP
scenarios was most apparent in SC and NEC, where
stronger radiative forcing is projected to have higher
Warm-Humid frequency and lower Warm-Dry and
Warm-Normal frequency, and vice versa. For all subregions and for all of China, Warm-Humid is projected
to increase, whereas Warm-Dry and Warm-Normal are
projected to decrease. A higher rate of occurrence is
expected for all three CYTs under a stronger RCP than
under a weaker one (RCP8.5 > RCP4.5 > RCP2.6).
3. The combination of temperature and precipitation (the
dual-variable CYTs) provides more complete information on climate change and more accurately characterizes regional differences in climate throughout
China than temperature or precipitation alone. Compared with other sophisticated indices, such as moisture indices, CYT can directly reflect the allocations of
temperature and precipitation.
© 2016 Royal Meteorological Society
Acknowledgements
We acknowledge the modelling groups for making their
simulations available for analysis, the Program for Climate Model Diagnosis and Intercomparison for collecting
and archiving the CMIP5 model output, and the China
Meteorological Data Sharing Service System for sharing the Chinese historical observed data (http://cdc.cma.
gov.cn/index.jsp). This study was financially supported
by the National Natural Science Foundation of China
(41171038, 41471085).
References
Alexander LV, Zhang X, Peterson TC, Caesar J, Gleason B, Klein Tank
AMG, Haylock M, Collins D, Trewin B, Rahimzadeh F, Tagipour A,
Rupa Kumar K, Revadekar J, Griffiths G, Vincent L, Stephenson DB,
Burn J, Aguilar E, Brunet M, Taylor M, New M, Zhai P, Rusticucci M,
Vazquez-Aguirre JL. 2006. Global observed changes in daily climate
extremes of temperature and precipitation. J. Geophys. Res. Atmos.
111(D5): D05109, doi: 10.1029/2005JD006290.
Basha G, Ouarda T, Marpu P. 2015. Long-term projections of temperature, precipitation and soil moisture using non-stationary oscillation
processes over the UAE region. Int. J. Climatol. 35(15): 4606–4618,
doi: 10.1002/joc.4310.
China Meteorological Administration. 2008. The national standard of
the People’s Republic of China: GB/T 21983–2008. Standards Press
of China: Beijing(in Chinese).
Choi G, Collins D, Ren G, Trewin B, Baldi M, Fukuda Y, Afzaal
M, Pianmana T, Gomboluudev P, Huong PTT, Lias N, Kwon W-T,
Boo K-O, Cha Y-M, Zhou Y. 2009. Changes in means and extreme
events of temperature and precipitation in the Asia-Pacific Network region, 1955–2007. Int. J. Climatol. 29(13): 1906–1925, doi:
10.1002/joc.1979.
Choi G, Lee B, Tenhunen J, Kang S. 2011. Intra-Annual Variations in
Temperature Lapse Rate Within a Mountainous Basin in Korea–A
Case Study of the Punch Bowl, Yanggu, In Proceedings of the TERRECO Science Conference, October 2–7, 2011, Karlsruhe Institute of
Technology, Garmisch-Partenkirchen, Germany, pp. 71–76.
Dong M, Jiang Y, Zheng C, Zhang D. 2012. Trends in the thermal growing season throughout the Tibetan Plateau during
1960–2009. Agric. For. Meteorol. 166–167: 201–206, doi:
10.1016/j.agrformet.2012.07.013.
Du H, Wu Z, Li M. 2011. Interdecadal changes of vegetation transition
zones and their responses to climate in Northeast China. Theor. Appl.
Climatol. 106(1–2): 179–188, doi: 10.1007/s00704-011-0432-x.
Du H, Wu Z, Li M, Jin Y, Zong S, Meng X. 2013a. Characteristics of
extreme daily minimum and maximum temperature over Northeast
China, 1961–2009. Theor. Appl. Climatol. 111(1–2): 161–171, doi:
10.1007/s00704-012-0649-3.
Du H, Wu Z, Zong S, Meng X, Wang L. 2013b. Assessing the characteristics of extreme precipitation over northeast China using the multifractal detrended fluctuation analysis. J. Geophys. Res. Atmos. 118(12):
6165–6174, doi: 10.1002/jgrd.50487.
Fan Z-X, Bräuning A, Thomas A, Li J-B, Cao K-F. 2011. Spatial
and temporal temperature trends on the Yunnan Plateau (Southwest
China) during 1961–2004. Int. J. Climatol. 31(14): 2078–2090, doi:
10.1002/joc.2214.
Fischer T, Gemmer M, Lüliu L, Buda S. 2011. Temperature and precipitation trends and dryness/wetness pattern in the Zhujiang River
Basin, South China, 1961–2007. Quat. Int. 244(2): 138–148, doi:
10.1016/j.quaint.2010.08.010.
Fraser EDG, Simelton E, Termansen M, Gosling SN, South A. 2013.
“Vulnerability hotspots”: integrating socio-economic and hydrological models to identify where cereal production may decline in the
future due to climate change induced drought. Agric. For. Meteorol.
170: 195–205, doi: 10.1016/j.agrformet.2012.04.008.
Ge Q, Zhang X, Zheng J. 2014. Simulated effects of vegetation
increase/decrease on temperature changes from 1982 to 2000
across the Eastern China. Int. J. Climatol. 34(1): 187–196, doi:
10.1002/joc.3677.
Gottfried M, Pauli H, Futschik A, Akhalkatsi M, Barančok P,
Alonso JLB, Coldea G, Dick J, Erschbamer B, Kazakis G. 2012.
Int. J. Climatol. (2016)
CHANGES IN THE COMBINATION OF TEMPERATURE AND PRECIPITATION OVER CHINA
Continent-wide response of mountain vegetation to climate change.
Nat. Clim. Change 2(2): 111–115.
Grundstein A. 2009. Evaluation of climate change over the continental United States using a moisture index. Clim. Change 93(1–2):
103–115, doi: 10.1007/s10584-008-9480-3.
Gunda T, Hornberger GM, Gilligan JM. 2016. Spatiotemporal patterns of
agricultural drought in Sri Lanka: 1881–2010. Int. J. Climatol. 36(2):
563–575, doi: 10.1002/joc.4365(Online first).
Hu C, Xu Y, Han L, Yang L, Xu G. 2015. Long-term trends in daily
precipitation over the Yangtze River Delta region during 1960–2012,
Eastern China. Theor. Appl. Climatol., doi: 10.1007/s00704-0151493-z(Online first).
IPCC. 2013. 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, UK and New York, NY, doi: 10.1017/
CBO9781107415324.
Jarlan L, Driouech F, Tourre Y, Duchemin B, Bouyssié M, Abaoui J,
Ouldbba A, Mokssit A, Chehbouni G. 2014. Spatio-temporal variability of vegetation cover over Morocco (1982–2008): linkages
with large scale climate and predictability. Int. J. Climatol. 34(4):
1245–1261, doi: 10.1002/joc.3762.
Kendall MG. 1975. Rank Correlation Measures. Charles Griffin: London.
Li Y, Zhu G. 2015. Interannual-interdecadal changes in climate zones in
China and controlling factors. Resour. Sci. 37(5): 605–618(in Chinese
with English abstract).
Li K, Wang S, Cao M. 2004. Vegetation and soil carbon storage in China.
Sci. China Ser. D Earth Sci. 47(1): 49–57, doi: 10.1360/02yd0029.
Li X, Xu H, Sun Y, Zhang D, Yang Z. 2007. Lake-level change and
water balance analysis at Lake Qinghai, West China during recent
decades. Water Resour. Manag. 21(9): 1505–1516, doi: 10.1007/
s11269-006-9096-1.
Li Z, Zheng F-L, Liu W-Z, Flanagan DC. 2010. Spatial distribution and
temporal trends of extreme temperature and precipitation events on
the Loess Plateau of China during 1961–2007. Quat. Int. 226(1–2):
92–100, doi: 10.1016/j.quaint.2010.03.003.
Li Q, Peng J, Shen Y. 2012. Development of China homogenized
monthly precipitation dataset during 1900–2009. J. Geogr. Sci. 22(4):
579–593, doi: 10.1007/s11442-012-0948-8.
Liu J, Zhai P. 2014. Changes in climate regionalization indices in
China during 1961–2010. Adv. Atmos. Sci. 31(2): 374–384, doi:
10.1007/s00376-013-3017-z.
Lobell DB, Schlenker W, Costa-Roberts J. 2011. Climate trends and
global crop production since 1980. Science 333(6042): 616–620.
Ma S, Zhou T, Dai A, Han Z. 2015. Observed changes in the distributions of daily precipitation frequency and amount over China from
1960 to 2013. J. Clim. 28(17): 6960–6978, doi: 10.1175/JCLI-D-150011.1.
Mann HB. 1945. Nonparametric tests against trend. J. Econ. Soc. 13(3):
245–259, doi: 10.2307/1907187.
Mengistu D, Bewket W, Lal R. 2014. Recent spatiotemporal temperature and rainfall variability and trends over the Upper Blue
Nile River Basin, Ethiopia. Int. J. Climatol. 34(7): 2278–2292, doi:
10.1002/joc.3837.
O’gorman PA. 2012. Sensitivity of tropical precipitation extremes to
climate change. Nat. Geosci. 5(10): 697–700.
Piao S, Ciais P, Huang Y, Shen Z, Peng S, Li J, Zhou L, Liu H, Ma Y,
Ding Y. 2010. The impacts of climate change on water resources and
agriculture in China. Nature 467(7311): 43–51.
Ren G, Zhou Y, Chu Z, Zhou J, Zhang A, Guo J, Liu X. 2008. Urbanization effects on observed surface air temperature trends in North China.
J. Clim. 21(6): 1333–1348, doi: 10.1175/2007JCLI1348.1.
Riahi K, Rao S, Krey V, Cho C, Chirkov V, Fischer G, Kindermann G,
Nakicenovic N, Rafaj P. 2011. RCP 8.5 – a scenario of comparatively
high greenhouse gas emissions. Clim. Change 109(1–2): 33–57.
Scherler D, Bookhagen B, Strecker MR. 2011. Spatially variable
response of Himalayan glaciers to climate change affected by debris
cover. Nat. Geosci. 4(3): 156–159.
Shi Y, Shen Y, Kang E, Li D, Ding Y, Zhang G, Hu R. 2007. Recent
and future climate change in northwest China. Clim. Change 80(3–4):
379–393.
Sillmann J, Kharin VV, Zhang X, Zwiers FW, Bronaugh D. 2013.
Climate extremes indices in the CMIP5 multimodel ensemble: part
1. Model evaluation in the present climate. J. Geophys. Res. Atmos.
118(4): 1716–1733, doi: 10.1002/jgrd.50203.
© 2016 Royal Meteorological Society
Sorg A, Bolch T, Stoffel M, Solomina O, Beniston M. 2012. Climate
change impacts on glaciers and runoff in Tien Shan (Central Asia).
Nat. Clim. Change 2(10): 725–731.
Srivastava AK, Rajeevan M, Kshirsagar SR. 2009. Development of a
high resolution daily gridded temperature data set (1969–2005) for the
Indian region. Atmos. Sci. Lett. 10(4): 249–254, doi: 10.1002/asl.232.
Sun Q, Miao C, Duan Q. 2015. Projected changes in temperature and
precipitation in ten river basins over China in 21st century. Int. J.
Climatol. 35(6): 1125–1141.
Tang G, Ding Y, Wang S, Ren G, Liu H, Zhang L. 2010. Comparative analysis of China surface air temperature series for the past
100 years. Adv. Clim. Chang. Res. 1(1): 11–19, doi: 10.3724/SP.J.
1248.2010.00011.
Taylor KE, Stouffer RJ, Meehl GA. 2011. An overview of CMIP5 and
the experiment design. Bull. Am. Meteorol. Soc. 93(4): 485–498, doi:
10.1175/BAMS-D-11-00094.1.
Thompson DW, Kennedy JJ, Wallace JM, Jones PD. 2008. A large
discontinuity in the mid-twentieth century in observed global-mean
surface temperature. Nature 453(7195): 646–649.
Thomson AM, Calvin KV, Smith SJ, Kyle GP, Volke A, Patel P,
Delgado-Arias S, Bond-Lamberty B, Wise MA, Clarke LE. 2011.
RCP4. 5: a pathway for stabilization of radiative forcing by 2100.
Clim. Change 109(1–2): 77–94.
Tian D, Guo Y, Dong W. 2015. Future changes and uncertainties in
temperature and precipitation over China based on CMIP5 models.
Adv. Atmos. Sci. 32(4): 487–496, doi: 10.1007/s00376-014-4102-7.
Van Vuuren DP, Stehfest E, Den Elzen MG, Kram T, Van Vliet J,
Deetman S, Isaac M, Goldewijk KK, Hof A, Beltran AM. 2011. RCP2.
6: exploring the possibility to keep global mean temperature increase
below 2 C. Clim. Change 109(1–2): 95–116.
Vicente-Serrano SM, Beguería S, López-Moreno JI. 2010. A multiscalar drought index sensitive to global warming: the standardized precipitation evapotranspiration index. J. Clim. 23(7): 1696–1718, doi:
10.1175/2009JCLI2909.1.
Wang XL, Feng Y. 2013. RHtestsV4 User Manual. Climate Research
Division, Science and Technology Branch, Environment Canada,
Toronto, Ontario, Canada. 29 pp. http://etccdi.pacificclimate.
org/software.shtml.
Wu S, Yin Y, Zheng D, Yang O. 2006. Moisture conditions and climate
trends in China during the period 1971–2000. Int. J. Climatol. 26(2):
193–206, doi: 10.1002/joc.1245.
Xu W. 1985. Kira’s temperature indices and their application in the study
of vegetation. Chinese J. Ecol. 4(3): 35–39(in Chinese with English
abstract).
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. 5(6): 527–533.
Xu X, Du Y, Tang J, Wang Y. 2011. Variations of temperature and
precipitation extremes in recent two decades over China. Atmos. Res.
101(1–2): 143–154, doi: 10.1016/j.atmosres.2011.02.003.
Xu W, Li Q, Wang XL, Yang S, Cao L, Feng Y. 2013. Homogenization
of Chinese daily surface air temperatures and analysis of trends in
the extreme temperature indices. J. Geophys. Res. Atmos. 118(17):
9708–9720, doi: 10.1002/jgrd.50791.
Yabi I, Afouda F. 2012. Extreme rainfall years in Benin (West Africa).
Quat. Int. 262: 39–43, doi: 10.1016/j.quaint.2010.12.010.
Yin J, Schlesinger ME, Stouffer RJ. 2009. Model projections of rapid
sea-level rise on the northeast coast of the United States. Nat. Geosci.
2(4): 262–266.
Zhang Q, Zhang J. 2016. Drought hazard assessment in typical corn
cultivated areas of China at present and potential climate change. Nat.
Hazards 81(2): 1323–1331, doi: 10.1007/s11069-015-2137-4(Online
first).
Zhang D, Zhang L, Yang J, Feng G. 2010. The impact of temperature
and precipitation variation on drought in China in last 50 years. Acta
Phys. Sin. 59(1): 655–663(in Chinese with English abstract).
Zhang Q, Zhang J, Wang C. 2016. Risk assessment of drought disaster in
typical area of corn cultivation in China. Theor. Appl. Climatol., doi:
10.1007/s00704-015-1723-4(Online first).
Zhao D, Zheng D, Wu S, Wu Z. 2007. Climate changes in northeastern
China during last four decades. Chin. Geogr. Sci. 17(4): 317–324, doi:
10.1007/s11769-007-0317-1.
Zhou B, Wen QH, Xu Y, Song L, Zhang X. 2014. Projected
changes in temperature and precipitation extremes in China by
the CMIP5 multimodel ensembles. J. Clim. 27(17): 6591–6611, doi:
10.1175/JCLI-D-13-00761.1.
Int. J. Climatol. (2016)