Agricultural Water Management 146 (2014) 249–261

Contents lists available at ScienceDirect

Agricultural Water Management

journal homepage: www.elsevier.com/locate/agwat

Responses of rice yield, irrigation water requirement and water use

efficiency to climate change in China: Historical simulation and future
projections
Weiguang Wang a,b,∗ , Zhongbo Yu a,b,d , Wei Zhang a , Quanxi Shao c ,
Yiwei Zhang a,b , Yufeng Luo a , Xiyun Jiao a , Junzeng Xu a
a
State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering, Hohai University, Nanjing 210098, China
b
College of Water Resources and Hydrology, Hohai University, Nanjing 210098, China
c
CSIRO Mathematics, Informatics and Statistics, Private Bag 5, Wembley, WA 6913, Australia
d
Department of Geoscience, University of Nevada Las Vegas, Las Vegas, NV 89154, USA

a r t i c l e i n f o a b s t r a c t

Article history: Rice is one of most important crops in China, accounting for approximately 18% of total cultivated area.
Received 5 March 2014 Rice productivity is significantly affected by undergoing climate change and vulnerable with water stress.
Accepted 19 August 2014 Therefore, investigating the responses of rice growth and water resources utilization to more pronounced
climate change is of great importance for water resources planning and management in terms of main-
Keywords: taining the ecosystem integrity and ensuring the food security. In this study, the changes of rice yield,
Rice
water consumption (ET), irrigation water requirement (IWR), water use efficiency (WUE) and irrigation
Irrigation water requirements
water use efficiency (IWUE) from 1961 to 2010 in three typical sites (Kunshan and Nanjing in the Yangtze
Water use efficiency
Climate change
River Basin, and Kaifeng in the Yellow River Basin) in rice plantation region of China were evaluated by
Statistical downscaling means of validated rice crop model ORYZA2000. Their responses to future climate scenarios of 21 century
were investigated by driving ORYZA2000 with downscaling climatic data from HadCM3 (Hadley Centre
Coupled Model version 3) under A2 and B2 emission scenarios with the help of a statistical downscal-
ing method (SDSM). The results exhibit a significant decline in rice yield was identified by 49.3 kg ha−1 ,
32.0 kg ha−1 and 45.8 kg ha−1 for Kunshan station, Nanjing station and Kaifeng station, respectively, in the
past 50 years due to obviously shortened rice growth duration (0.20 day a−1 , 0.15 day a−1 and 0.27 day a−1 ,
respectively). While changes of ET and IWE were different for three stations representing by significant
increase of ET and IWE in Kunshan, non-significant increase in Nanjing and significant decrease in Kaifeng.
Whereas accompanying production reduction, simulated WUE and IWUE for three stations all presented
significant deceasing trends ranging from 0.06 kg ha−1 mm−1 to 0.16 kg ha−1 mm−1 . The future projec-
tion results under IPCC SRES A2 and B2 emission scenarios indicated the generally negative effect of
climate warming to rice yield (maximum by −18.9% decline in 2090s in Kunshan) during the 21 century
due to remarkable shortened growth period, resulting in generally depressed WUE and IWUE, although
there would be the distinct response of the ET and IWR to future climate change for the three stations.
Meanwhile, the increase of CO2 concentration under future climate is beneficial to raise the rice yield,
alleviate crop water consumption and irrigation water requirements and improve the water use efficien-
cies of rice in a certain degree. Further works should be carried out to capture simulation uncertainties
in climate change impact assessment with consideration of interactions among anthropogenic activities,
environmental and biological factors.
© 2014 Elsevier B.V. All rights reserved.

1. Introduction

Global atmospheric concentrations of greenhouse gases (GHG)

∗ Corresponding author at: State Key Laboratory of Hydrology-Water Resources
have increased remarkably since 1750 and will continue to rise dur-
and Hydraulic Engineering, Hohai University, Nanjing 210098, China.
Tel.: +86 25 83786982; fax: +86 25 83786606.
ing the present century as a result of widespread human activities
E-mail addresses: wangweiguang006@126.com, wangweiguang006@gmail.com such as burning of fossil fuels, cement production, and land use
(W. Wang). change, which are stated by the most recent Intergovernmental

http://dx.doi.org/10.1016/j.agwat.2014.08.019
0378-3774/© 2014 Elsevier B.V. All rights reserved.
250 W. Wang et al. / Agricultural Water Management 146 (2014) 249–261

Panel on Climate Change (IPCC) report (IPCC, 2007). Global mean States by combining the crop model and different climate model-
temperatures have thus raised approximately 0.74 ◦ C in the last ing and indicated variability in both irrigation demands and crop
100 years and are expected to continually increase of 1.1–6.4 ◦ C yields increases. Tao et al. (2003) explored the changing trends in
in this century (IPCC, 2007). This climate change mainly char- agricultural water demands, the changing trends and variability
acterized by global warming has led to dramatic influences on in soil moisture associated with both drought and increased sur-
both social and natural systems, causing overwhelming public con- face runoff in China croplands during the second half 20th century.
cerns (Huntington, 2006; Papaioannou et al., 2011; Wang et al., Döll (2002) presented the first global analysis of impact of climate
2012a,b). Among the suffering, water and agriculture sectors are change on irrigation water requirement using a developed global
likely to be the most sensitive ones to climate change-induced irrigation model. As a function of both projected irrigated land and
impacts, especially in China (IPCC, 2001; Tao et al., 2003; Piao et al., climate change, Fischer et al. (2007) computed the future regional
2010). Under the background with CO2 concentration increase and global irrigation water requirements from 1990 to 2080. More
and climate warming, the agriculture and crop production may be recently, Thomas (2008) modeled the effects of climate change on
affected through the changes of physiological processes including irrigation requirements with high-resolution girded climate data
photosynthesis, respiration and partitioning of photosynthesis pro- sets.
duction due to the soil and air conditions alteration (Chartzoulakis Rice is the main food staple in Asia where more than half of
and Psarras, 2005; Yang and Zhang, 2006; Guo et al., 2010). Mean- the world’s population resides (Bachelet et al., 1992), particularly
while, the patterns of the irrigation water requirements (IWR) and in China, where rice is the staple food for more than 65% of Chi-
water use efficiency (WUE) should be changed since the balance nese population (State Environmental Protection Administration,
between precipitation, evapotranspiration and the resultant fluc- 2003). Although the rice cropping area represents only 29.1% of the
tuations have been broken due to the alteration of rainfall patterns, total national crop growing area, rice production contributes 43.7%
soil moisture and other relevant climate variables such as humidity, of total national grain production, representing 22.8% and 36.9% of
windiness and cloudiness under global warming (Komuscu et al., the total world cropping area and production respectively (Xiong
1998; Silva et al., 2007). In fact, from a water resources perspective, et al., 1992; Yao et al., 2007). Unfortunately, climate change and its
the alterations of precipitation, runoff, infiltration, groundwater impact on rice production, limiting the capacity of farmers to grow
flow, evapotranspiration, and soil moisture in many parts of the this crop, is a big challenges China and the rest of the world will
world during the past century suggest climate change is leading have to face (Peng et al., 2004; Tao et al., 2008; Piao et al., 2010).
to an intensification of regional hydrological cycles and could have Therefore, the vulnerability of rice production to global change has
major impacts on water resources, affecting both ground and sur- become of key concern with extensive studies being conducted in
face water supply irrigation and agricultural system (Alan et al., simulating the impact of climate change on rice production in Asia,
2003; Huntington, 2006). especially in China (e.g., Lin et al., 2005; Yao et al., 2007; Tao et al.,
With rising concerns over food security, exploring the effects of 2008; Xiong et al., 2009; Shen et al., 2011). However, past assess-
climate change on agriculture has been widely attempted and the ments have mostly focused on change in crop yield, and few studies
relative efforts mainly focused on estimating changes in crop pro- have assessed changing irrigation requirements and water use effi-
ductivity (e.g., Brumbelow and Georgakakos, 2001; Tao et al., 2006; ciency. Moreover, most assessment has employed climate scenarios
Hussain and Mudasser, 2007; Challinor and Wheeler, 2008; Mo from GCM with a coarse resolution to point-based crop model, sug-
et al., 2009; Liu et al., 2010). Three different ways are usually used gesting it is necessary to add fine-scale local climate information
to investigate the response of crop yield to climate change, which to the coarse-scale one to derive the crop model (WMO, 2002; Yao
are summarized though reviewing the related studies widespread et al., 2007).
in the world. First, exploring the evidences of crop response to cli- Therefore, the objectives of this paper are to: (1) explore how
mate change by correlative analyzing observation between crop the past climate change has impacted the rice growth processes,
yield and climate change directly (e.g., Peng et al., 2004; Tao et al., irrigation water requirements, and water use efficiencies with a
2006; Egli, 2008; Malone et al., 2009). Second, examining crop modeling approach that enables the assessment of climate change
yield change under different climate inputs with using weather impact with fixing rice varieties and management practices at three
generator to drive a crop model (e.g., Stockle et al., 1997; Zhang typical sites of the rice plantation areas in China, (2) conduct a
and Liu, 2005; Kou et al., 2007; Tao et al., 2008). The third and comprehensive analysis on how the rice yield, irrigation water
also the most prevalent method is to drive crop model with the requirements and water use efficiency response to climate change
future climate information from GCMs to simulate the response of under future climate scenarios accompanying CO2 concentration
crop yield in the future (e.g., Trnka et al., 2004; Xiong et al., 2007, enrichment. For this, a widely use rice crop model, ORYZA2000, was
2008; Silva et al., 2007; Chavas et al., 2009; Mo et al., 2009; Shen calibrated with experiments data. Meanwhile, the statistical down-
et al., 2011). Climate change is found to have significant impacts on scaling method (SDSM) are used to create daily weather series from
environments and conditions affecting crop productivity in China. the HadCM3 (Hadley Centre Coupled Model version 3) model under
Although some aspects of climate change such as longer growing A2 and B2 emissions.
seasons (Xiong et al., 2007), increasing atmospheric CO2 (Liu et al.,
2010), warmer nighttime temperatures and higher precipitation
(Thomson et al., 2006) may bring benefits in some regions, there 2. Materials and methods
will also be a range of adverse impacts, including reduced water
availability (Bouman et al., 2007), greater water need (Piao et al., 2.1. Study area
2010), and more frequent extreme high temperature (Shen et al.,
2011). These impacts may put agricultural activities at significant In this study, three typical rice ecological stations with geo-
risk (Eitzinger and Kubu, 2009; Mimi and Jamous, 2010). graphical and climatological differences, i.e., Kaifeng, Nanjing and
Compared with the studies on the response of crop yield to cli- Kunshan, were selected to simulate the effects of climate change
mate change, the assessment on changing water resources use, on rice growth and water utilization (Fig. 1). Kaifeng (34◦ 82 N,
irrigation water demands and requirements in agriculture field 114◦ 51 E; 69.0 m a.s.l.) is located in the southern bank of the Yellow
started later, and have been fewer in number (Wang et al., 2012a). River, while Nanjing (32◦ 00 N, 118◦ 28 E; 7.1 m a.s.l.) and Kunshan
Brumbelow and Georgakakos (2001) assessed changes in irrigation (31◦ 23 N, 120◦ 58 E; 17.5 m a.s.l.) belong to the middle and lower
requirements and crop yields for five traditional crops of the United reaches of the Yangtze River Basin. All the three stations distribute
 W. Wang et al. / Agricultural Water Management 146 (2014) 249–261 251

Fig. 1. The locations of the three experimental stations.

in the Middle and South China, the famous rice plantation regions initiation, and physiological maturity were recorded. During the
with warm climate and abundant precipitation to be suitable for growing season, crop samples were taken to measure the leaf area
the long growth of rice under summer monsoon circulation (Yao index (LAI) and the dry weights of green leaves, brown/yellow
et al., 2007). leaves, stems, and panicles. The dry weights reported here as “dry
biomass” were obtained after oven-drying at 70 ◦ C to constant
2.2. Field experiment and data collection weigh. Daily meteorological data, including precipitation volume,
wind speed, temperature (maximum, minimum, and average tem-
The experiment data of Kaifeng and Nanjing stations to peratures), sunshine duration, and relative humidity were recorded
run the ORYZA2000 crop model were collected from web- using an automatic weather station (ICT, Australia). Soil mois-
site http://www.knowledgebank.irri.org/oryza2000/. Therefore, ture was monitored daily in three replicates using time-domain
the field experiment was only conducted at Kunshan Irrigation reflectometry equipment (TDR, Soil moisture, USA) and 20 cm
and Drainage Experiment Station in Jiangsu Province from June to waveguides installed at 0–20 cm and 20–40 cm. The ponding water
October 2007 and from June to October 2008. The Kunshan exper- layer depth was monitored using a vertical ruler fixed in the field.
iment site has a subtropical monsoon climate, with an average Water volumes for irrigation were routinely recorded using water
annual air temperature of 15.5 ◦ C, a mean annual precipitation of meters installed on the pipes at the outlet for each plot. Drainage
1097.1 mm, and a frost-free growing season of 234 days a−1 (from was calculated by measuring water depths before and after the
March to November). Soil in the experimental site was classified drainage. Grain yield was determined for each plot from the smaller
as dark-yellow hydromorphic paddy soil, which represented the crop sample at maturity. The soil properties, crop management,
typical soil type in this region. The soil texture in the plowed layer crop and water monitoring method, and the detailed experimen-
was clay. Rice seedlings were transplanted to the paddy fields on tal data of Kaifeng and Nanjing stations were fully described in
June 22 and harvested on October 30, 2007, and on June 25 and http://www.knowledgebank.irri.org/oryza2000/, Cabangon et al.
harvested on October 25, 2008. (2003) and Feng et al. (2007).
Under the same condition of rice varieties, transplanting, den-
sity, plant protection, fertility, medication and other technical 2.3. ORYZA2000 model description
measures, three repeat treatments were set in three experimen-
tal fields with the field area about 150 m2 (20 m × 7.5 m). All ORYZA2000 is an advanced, explanatory and dynamic eco-
treatments received the fertilizer applications for local farmer’s physiological rice growth simulation model. The model is able to
fertilization practice (details see Xu et al., 2012). Weed, pests simulate the growth processes, development and water balance of
and diseases were well controlled using recommended practices. rice under the situation of potential, water limitation, and nitro-
All the experiments, dates of emergence, transplanting, panicle gen limitation (Bouman et al., 2001). It is assumed that the rice is
252 W. Wang et al. / Agricultural Water Management 146 (2014) 249–261

free from disease, pests and weeds for all modeling simulations. 2.5. Historical meteorological data and future projection
The model has been widely applied by various studies in terms
of nitrogen effect (e.g., Bouman and Van Laar, 2006), water man- Historical meteorological data from three National Meteorolog-
agement optimization (e.g., Feng et al., 2007; Yadav et al., 2011) ical Observatory (NMO) stations, i.e., Nanjing, Kaifeng and Suzhou,
and climate change impact assessment (e.g., Zhang et al., 2008b; were provided by the National Meteorological Information Cen-
Karim et al., 2012), and its performances were also found to be sat- tre of China (NMIC) of China Meteorological Administration (CMA),
isfactory and acceptable through these validations under different including precipitation (mm), daily air temperature (◦ C), relative
water regimes, irrigation practices, and various levels of nitrogen humidity (%), wind speed (m s−1 ) and sunshine duration (h) for the
fertilization with time step of 1-day (Boling et al., 2007; Feng et al., period of 1961–2010.
2007). More detailed description of the model and program model Twenty-six large-scale atmospheric variables, such as mean sea
can be found in Bouman et al. (2001) and Bouman and Van Laar level pressure, near surface relative humidity, near surface spe-
(2006). cific humidity, relative humidity at 500 hPa, air temperature at 2-m
height, etc, derived from the daily NCEP/NCAR reanalysis dataset
2.4. Projection of future climatic data based on statistical during the period of 1961–2001 were used to calibrate and validate
downscaling model (SDSM) the SDSM model. The NCEP/NCAR reanalysis data were downloaded
freely from the internet sites at a scale of 3.75◦ × 2.5◦ (transformed
Statistical downscaling model (SDSM), a hybrid of regres- from a spatial scale of 2.5◦ × 2.5◦ to match the resolution of HadCM3
sion based and stochastic weather generate downscaling method, model, http://www.cics.uvic.ca/scenarios/sdsm/select.cgi).
developed by Wilby et al. (2002) was employed to project future The Hadley Center’s coupled ocean/atmosphere climate model
climatic data deriving the crop model. Following steps were (HadCM3) was used for future climate projection. The same daily
involved to project future climatic variables. First, appropriate HadCM3 atmospheric variables as NCEP data under scenarios A2
atmospheric variables for predictors were determined by means (high greenhouse gas emission scenarios denoted as H3A2) and B2
of the screening way in SDSM version 4.2. Second, the statistical (low greenhouse gas emission scenarios denoted as H3B2) at a res-
regression models between climatic variables of three station (pre- olution of 3.75◦ × 2.5◦ for the period of 1961–2099 were used to
dictand) and large-scale atmospheric variables (predictor) obtained generate future climatic data. The climate change and the response
from the NCEP reanalysis for the climate during 1961–2000 of rice growth and water related variables during the three future
were constructed. Two sub-periods, 1961–1990 and 1991–2000, periods of the 2030s (2030–2039), the 2060s (2060–2069) and the
were used to calibrate and validate regression equations, respec- 2090s (2090–2099) were analyzed by comparing with those in the
tively. Four statistics, namely the coefficient of determination baseline period of 1990s (1990–1999).
(R2 ), the coefficient of efficiency (Ens), the ratio of simulated
and observed standard deviation (RS), and model biases (Bias), 2.6. Trend analysis method
were used to assess quantitatively the performance of SDSM.
Detailed calculation descriptions regarding these statistics indices In this study, a simple linear regression method was used to
can be found in Wang et al. (2013). Finally, the girded data for quantify the magnitude of overall trend (i.e., the change per unit
the future climate provided by HadCM3 under A2 and B2 sce- time) in rice yield, irrigation water requirements, and WUE, which
nario for the period 2011–2099 were inputted into the models is the most common method for detecting trends in climate vari-
to generate the downscaled future daily climatic series of each ables because of its simplicity for an unknown trend (da Silva,
station. 2004). Meanwhile, the rank-based Mann–Kendall (MK) statistical

Fig. 2. Simulated (solid line) and measured dry biomass of the total crop (star), leaves (cross), stems (circle dot), and panicles (short dash), and leaf area index (square point)
over time in calibration period (2007) and validation period (2008) in Kunshan.
 W. Wang et al. / Agricultural Water Management 146 (2014) 249–261 253

Table 1 Table 2
Selected predictor variables for all predictands including precipitation (P), maxi- Performance assessment for predictands in calibration and validation in Kunshan.
mum temperature (Tmax ), minimum temperature (Tmin ), radiation (R), water vapor
(Wv), and Wind speed (Ws). Items Period R2 Ens RS Bias

Solar radiation Calibration 0.985 0.985 0.997 0.003

Predictors Description Predictands
Validation 0.958 0.950 1.060 0.475
P Tmax Tmin R Wy Ws Minimum Calibration 0.996 0.996 0.999 0.003
√ √ √ √ √ √ temperature Validation 0.979 0.977 1.017 –1.062
Mslpas Mean sea level pressure Calibration 0.591 0.567 0.791 –0.002
√ Wind speed
pf Surface airflow strength Validation 0.437 0.380 0.684 0.060
√ √ √
pu Surface zonal velocity Calibration 0.586 0.302 0.928 0.252
√ Precipitation
pv Surface meridional velocity Validation 0.426 0.334 0.688 0.138
√ √
pz Surface velocity Calibration 0.996 0.996 0.998 0.000
√ √ √ Maximum
p5 f 500 hPa airflow strength Validation 0.993 0.993 1.000 –0.146
√ √ √ √ √ temperature
p5 u 500 hPa zonal velocity Water vapor Calibration 0.997 0.997 0.999 0.000
√ √ √
p5 v 500 hPa meridional velocity Validation 0.997 0.997 1.027 0.015
√ √ √ √ √
p5 z 500 hPa velocity
√ √ √ √ √ √
p500 500 hPa geopotential height
√ √
p5zhas 500 hPa divergence
√ to 0.9 in calibration and validation. The simulation of wind speed
p8 f 850 hPa airflow strength
√ √ and precipitation were less satisfactory (Ens and R2 were between
p8 u 850 hPa zonal velocity
√ √ √
p8 v 850 hPa meridional velocity 0.302 and 0.591). For illustrative purposes, selecting precipitation,
√ √ √ √ √ √
p8 z 850 hPa velocity minimum and maximum temperature as example, Figs. 3 and 4
√ √ √ √
p850 850 hPa geopotential height
√ √ √ √ show comparison of mean monthly variables statistics simulated
p8zhas 850 hPa divergence
√ √ by SDSM and observed data in calibration and validation periods,
r500 Relative humidity at 500 hPa
√ √ √
r850 Relative humidity at 850 hPa respectively, reflecting similar conclusions with the above analysis.
√ √
Rhumas Near surface relative humidity
√ √ √ √ √ √
Concurrently, the intra-annual variation characteristics of precipi-
Shumans Surface specific humidity tation can be overall captured by statistical downscaling model in
√ √ √ √ √
Tempas Mean temperature at 2 m
despite of relative poor statistical performance shown in Table 2.
Generally, the model could simulate most local variables well,
test (Mann, 1945; Kendall, 1975), a nonparametric method highly although the capability in simulating precipitation and water vapor
recommended by the World Meteorological Organization, was used was relatively limited. The statistical relationships built by SDSM
to test whether the trends are statistically significant or not at the were acceptable for practical uses to conduct the future climate
5% level of significance by identifying if the absolute values of Z condition as crop model input.
exceed 1.96. Moreover, before the MK test was applied, variable
series were tested for persistence by the serial correlation analysis 3.3. Observed trends of climatic factors in three stations during
method presented by Haan (2002), and the effect of serial cor- the period of 1961–2010
relation, if existing, was eliminated by a modified pre-whitening
method called trend free pre-whitening (TFPW), proposed by Yue Kaifeng, Nanjing and Kunshan stations all presented a signifi-
et al. (2002). cant decreasing trend in global solar radiation during the growth
seasons of rice from 1961 to 2010 by −0.033 MJ m−2 d−1 a−1 ,
3. Result –0.023 MJ m−2 d−1 a−1 and –0.025 MJ m−2 d−1 a−1 , respectively
(Table 3). Opposite to global solar radiation, the minimum temper-
3.1. Model calibration and validation
ature showed a significant increasing trend (Z > 1.96) in all three
stations. Minimum temperature increased by 0.055 ◦ C d−1 a−1 at
With the meteorological data obtained from automatic weather
Kunshan, by 0.033 ◦ C d−1 a−1 at Nanjing and by 0.045 ◦ C d−1 a−1 at
station set up in the experiment field, two calibration programs,
Kaifeng. Inconsistent with minimum temperature, over the past
DRATES and PARAM, built in the ORYZA2000 model were used to
50 years, during the growth season of rice, maximum tempera-
calibrate the rice genetic parameters. Graphical examples in simu-
ture increased significantly (Z > 1.96) by 0.033 ◦ C d−1 a−1 only at
lated and measured biomass and LAI of the whole crop and of crop
Kunshan, but not significantly at Nanjing and Kaifeng. This further
organs over time are shown in Fig. 2 for the calibration set (2007)
emphasized the phenomenon minimum temperature warming
and the validation set (2008). Generally, despite some underesti-
faster than maximum temperature highlighted in previous global
mations of the biomass of stems and of LAI at harvest (or at 60
study (e.g., Alexander et al., 2006) and regional studies in China (e.g.,
days after emergence), the crop variables were reproduced gener-
Zhang et al., 2008a; Wang et al., 2013). Although not significant,
ally satisfactorily, especially for the calibration set. Generally, the
decreasing water vapor can be found in all three stations. Contrarily,
results indicated that ORYZA2000 model can be used to simulate
all three stations were dominated by increasing trends of precipita-
the response to climate change of the local rice variety in a relative
tion, but not significant. As for the wind speed, Nanjing and Kaifeng
high accuracy.
stations were dominated by significant negative trends with the
3.2. Evaluation on the performance of SDSM slope of about –0.007 m s−1 d−1 a−1 and –0.022 m s−1 d−1 a−1 , while
weak increasing trend (non significant) was found in Kunshan
The predictors identified for different local variables (precipi- station.
tation, the maximum and minimum temperature, radiation, wind
speed and water vapor are summarized in Table 1. It can be seen 3.4. Responses of rice phenology and grain yield to climate
that different atmospheric predictors control different local vari- change in the past decades
ables. Taking Kunshan station as example, the statistic assessment
of calibration and validation results are shown in Table 2, from The length of rice growth periods in Kunshan, Nanjing and
which it can be seen that local variables were generally consis- Kaifeng during the past 50 years were simulated by the vali-
tent with observation. It is especially true for radiation, maximum dated ORYZA2000 model (Fig. 5). The results showed the lengths
and minimum temperature and water vapor. For these variables, R2 of growth duration in all three stations were shortened due to
and Ens between simulated and observed ones exceeded or equaled the warming climate without consideration of variety change.
254 W. Wang et al. / Agricultural Water Management 146 (2014) 249–261

Fig. 3. Monthly distributions for modeled and observed meteorological variables in calibration period.

Fig. 4. Monthly distributions for modeled and observed meteorological variables in validation period.
 W. Wang et al. / Agricultural Water Management 146 (2014) 249–261 255

Table 3
Mann–Kendall Z value and slope of radiation (R), minimum temperature (Tmin ), wind speed (Ws), precipitation (P), maximum temperature (Tmax ), water vapor (Wv) at three
sites during the period 1961–2010.

Station R Tmin Ws P Tmax Wv

Slope Z Slope Z Slope Z Slope Z Slope Z Slope Z

Kunshan –0.033 –4.433 0.055 5.638 0.001 0.284 2.283 0.954 0.033 3.731 –0.001 –0.368
Nanjing –0.023 –3.965 0.033 3.814 –0.007 –2.292 1.492 0.669 0.013 1.656 –0.001 –0.201
Kaifeng –0.025 –3.279 0.045 4.500 –0.022 –4.333 0.721 0.184 –0.002 –0.167 –0.001 –0.268

The reduction of rice growth period was statistically signifi- water requirement (IWR) for the period of 1961–2010 are pre-
cant (Z < −1.96) at all three stations. The decreasing rates of rice sented in Fig. 6. Meanwhile, the temporal change patterns of water
growth duration for Kunshan, Nanjing and Kaifeng were around use efficiency (hereafter WUE) and irrigation water use efficiency
0.20 day a−1 , 0.15 day a−1 and 0.27 day a−1 , respectively. Partly due (hereafter IWUE), reflecting water productivity, defined respec-
to the obvious decline in growth duration, the simulated grain tively by the ratio of simulated yield to ET and the ratio of simulated
yield of rice under fully irrigated conditions decreased signifi- yield to IWR, were also investigated (Fig. 6). IWR was calculated
cantly (Z < −1.96) in the past 50 years in all three stations, although under continuous flooding scenarios, and 75 mm irrigation was
there maybe some beneficial influence on crop yield due to applied when the standing water depth was lower than 10 mm. The
increasing night-temperature. Simulated rice yield was respec- changing patterns of simulated ET without consideration of vari-
tively decreased by 49.3 kg ha−1 , 32.0 kg ha−1 and 45.8 kg ha−1 for ety change for the three stations were totally different. Significant
Kunshan, Nanjing and Kaifeng. decreasing trend in ET with the slope value of about –1.12 mm a−1
was found in Kaifeng station. While significant increasing trend in
3.5. Response of evapotranspiration, irrigation water ET with the slope value of about 1.70 mm a−1 was found in Kunshan
requirement and water utilization efficiency station. The simulated ET in Nanjing station tended to increase as
well, but not significant. As for the IWR, similar changing patterns
The overall trends of simulated crop water demand (evapo- compared with ET were found in the three stations. From 1961 to
transpiration under full irrigation, hereafter, ET) and irrigation 2010, simulated IWR was significantly increased by 1.62 mm a−1

Fig. 5. Trend of rice yield and growth duration in 1961–2010.

256 W. Wang et al. / Agricultural Water Management 146 (2014) 249–261

Fig. 6. Trend of water consumption (ET) and irrigation water requirement (IWR) in 1961–2010.

and significantly decreased by 1.04 mm a−1 at Kunshan and Nan- in all schemes would be mitigated, especially in 2030s in Nan-
jing, respectively. While Nanjing station was dominated by a jing under B2 scenarios, in 2060s in Kaifeng under A2 scenarios,
non significant increasing trend in IWR with the slope value of in 2060s and 2090s in Kaifeng under B2 scenarios, when the rice
0.05 mm a−1 . yields would increase compared with that under baseline, indi-
Different from ET and IWR presenting obviously distinct chang- cating that CO2 fertilization has a certain positive effect of rice
ing patterns for the three stations, but agreed with growth yield.
duration and yield, WUE and IWUE showed consistently signifi- Compared with rice yield, more apparent spatial differences can
cant deceasing trends (Z < −1.96) in the three stations, although be found the response of ET and IWR to climate condition in the
the detailed slopes value were different (Fig. 7). During the past future (Fig. 9). The cumulative ET in Kunshan would gently decrease
50 years, the WUE was decreased by around 0.163 kg ha−1 mm−1 , for both A2 and B2 scenarios with maximum value of 7.04% occur-
0.075 kg ha−1 mm−1 and 0.062 kg ha−1 mm−1 at Kunshan, Nanjing ring in 2090s under B2 scenarios without consideration of CO2
and Kaifeng, respectively. While the IWUE was decreased by around fertilization. While in Nanjing station, the cumulative ET amounts
0.152 kg ha−1 mm−1 , 0.067 kg ha−1 mm−1 and 0.060 kg ha−1 mm−1 gradually increase with air warming for both A2 and B2 scenarios
at Kunshan, Nanjing and Kaifeng, respectively. In general, the from 2030s to 2090s with maximum value of 29.45% occurring in
decreasing trend was larger at the southern sites simultane- 2090s under A2 scenarios. As for Kaifeng, ET would slightly decrease
ously accompanying with more humid climate than northern in all schemes except in 2090s under A2 scenarios. Compared with
sites. ET, IWR present similar changing patterns except several schemes
including at Kunshan in 2060s and 2090s under A2 scenarios and
3.6. Change of yield, evapotranspiration (ET) and irrigation water at Kaifeng in 2060s under A2 scenarios and in 2090s under B2 sce-
requirements (IWR) under future scenarios narios. The maximum change of IWR can be found in Nanjing in
2090s under A2 scenarios by 35.4% without consideration of CO2
By driving the calibrated ORAZA2000 with downscaling climatic fertilization. For both ET and IWR, although unlike that of yield,
variables from HadCM3 climate model, the rice growth duration, the differences of change between with/without CO2 fertilization
the crop yield, ET, and IWR under SRES A2 and B2 scenarios were were extremely small, the elevated CO2 fertilization would present
projected. The average rice growth duration would be remark- negative effect on ET and IWR in all schemes.
ably shortened for all the three stations for three future periods
(Fig. 8) under the obvious climate warming (Table 4). Although a 3.7. Comparison of water utilization efficiency with/without CO2
consistent patterns of change of growth duration under both sce- fertilization
narios for all the stations presenting a increasing shorten days over
time with an exception of 2030s under A2 scenarios in Kunshan Fig. 10 shows the relative change of WUE and IWUE compared to
station, there were obviously spatial differences for the num- that under the baseline period under A2 and B2 scenarios with and
bers of shorten days (Fig. 8). The maximum of 16 days and the without CO2 fertilization. For Kunshan and Nanjing, in despite of the
minimum of 5 days were found in Kaifeng under B2 scenarios differences among future stages and scenarios, decreasing trends
in 2090s and in Nanjing under A2 scenarios in 2030s. Although in WUE and IWUE were projected regardless of whether or not
the growth durations presented consistent reduction for all three considering the CO2 fertilization, and the percentage of WUE and
stations in different stages of the 21st century, the changing behav- IWUE change gradually increase with air warming from 2030s to
ior of rice yield differ considerably, especially with consideration 2090s. It should be noted that although the ET and IWR in Kunshan
CO2 fertilization. The simulated rice yield would experience a for most schemes decrease (Fig. 9), the larger percentage of rice
clear drop in all schemes except in 2030s in Kaifeng under B2 yield decline (Fig. 8) would result in the WUE and IWUE decrease.
scenarios without consideration CO2 fertilization. The maximum The greatest decrease for WUE and IWUE were both found in Nan-
and minimum decrease in rice yield relative to that under base- jing with under A2 scenarios in 2090s with the value of –37.3% and
line can both be found under B2 scenarios, at Kunshan in 2090s –40.1%, respectively. While for Kaifeng, increasing WUE and IWUE
by –18.9% and at Kaifeng in 2090s by –0.3%, respectively. While were generally found under B2 scenarios, which was particularly
with CO2 fertilization, the magnitudes of decrease in rice yield true with consideration of CO2 fertilization. The greatest increase
 W. Wang et al. / Agricultural Water Management 146 (2014) 249–261 257

Fig. 7. Trend of water use efficiency (WUE) and irrigation water use efficiency (IWUE) and in 1961–2010.

Fig. 8. Changes of rice yield and growth period under A2 and B2 scenarios relative to the baseline.

for WUE and IWUE were both found in 2030s under B2 scenarios 4. Discussion
with CO2 fertilization by 10.7% and 11.1%, respectively. Generally,
relative to the WUE and IWUE change without considering CO2 fer- Regional response of irrigation water requirement and utiliza-
tilization, less decline and more enhancements for corresponding tion to climate change is important and one of major challenges for
decreasing and increasing cases can be found with considering CO2 regional agricultural production as well as water resources man-
fertilization, indicating that elevated CO2 can improve the water agement in broad regions of the world (Tao and Zhang, 2013).
use efficiencies of rice in a certain degree. Employing a physical-based crop model to evaluate climate change

Table 4
Changes of minimum temperature (Tmin ), precipitation (P) and maximum temperature (Tmax ) in the future scenarios relative to the baseline.

Scenario Station Baseline 2030s 2060s 2090s

◦ ◦ ◦ ◦ ◦ ◦
P (mm) Tmax ( C) Tmin ( C) P (%) Tmax ( C) Tmin ( C) P (%) Tmax ( C) Tmin ( C) P (%) Tmax (◦ C) Tmin (◦ C)

A2 Kunshan 875.49 27.57 20.50 –1.38 1.40 1.49 –25.53 2.71 2.38 –17.57 3.98 3.72
Nanjing 516.51 28.31 18.30 –15.17 1.57 1.49 –14.93 3.55 2.63 –16.87 5.47 4.54
Kaifeng 849.52 28.00 19.81 –7.39 0.88 1.47 –23.56 2.00 3.59 –28.06 3.33 6.09
B2 Kunshan 877.57 27.58 20.51 –5.85 1.39 1.55 –16.58 2.35 2.29 –6.21 2.98 3.33
Nanjing 509.74 28.30 18.30 –0.95 1.49 1.62 –2.04 2.62 2.49 –4.79 3.55 3.39
Kaifeng 837.92 28.00 19.82 1.09 0.96 1.95 –1.93 1.56 3.04 –15.45 2.45 4.37
258 W. Wang et al. / Agricultural Water Management 146 (2014) 249–261

Fig. 9. Changes of water consumption (ET) and irrigation water requirement (IWR) relative to the baseline without/with CO2 fertilization under A2 and B2 scenarios.

impact on crop growth, water consumption and utilization effi- and increase in minimum temperature and precipitation, the incon-
ciency and explore its related mechanism is an effective way to sistent change in wind speed and maximum temperature for the
provide some helpful decision-making suggestions for food and three stations can be found during the past decades due to their
water resources safety in the future. In this study, driving crop obviously distinct climate type. This would potentially lead to dif-
model ORYZA2000 with observed meteorological data and down- ferent results of response of rice yield for three stations, especially
scaled GCM data, a comprehensive understanding of responding for water consumption due to complex sensitivities of evapotrans-
behaviors of rice production and water utilization to climate change piration to climatic variables (Gong et al., 2006; McVicar et al., 2012;
are gained for three typical rice plantation sites (i.e., Kunshan, Wang et al., 2012b).
Nanjing and Kaifeng) in China. Trend analysis of climate factors Despite that increasing minimum temperature would have a
from 1961 to 2010 indicated that three were the significant warm- beneficial influence on crop yield (Liu et al., 2006), rice yield in
ing trends in three stations, which corresponds well to the global all three stations was simulated to decline without exception in
warming trend reported by IPCC (2007). Especially, changes in daily the past decades. The negative impacts were mainly resulted from
maximum temperature were less marked compared with that in reduced radiation together with shorted rice grow period due to
daily minimum temperature, revealing that these rice plantation general increase in temperature. This has also been drawn by the
sites have become less cold rather than hotter, which are in agree- previous study on drought crop in north China by Chen et al. (2010).
ment with previous studies global (e.g., Alexander et al., 2006) Changes of rice water demand and irrigation water requirement
and other regions of China (e.g., Wang et al., 2013). In despite of for the three stations were inconsistent during the past decades.
being located to different climate regions, all three sites showed A gradual pattern can be roughly revealed from south to north
declining trends in solar radiation during the growth seasons of with significant increase of rice water demand and irrigation water
rice, which are considered as global phenomenon (Wild, 2009) requirement in Kunshan, non-significant increase in Nanjing and
and also reported in many regions studies in China (e.g., Wang significant decrease in Kaifeng (Fig. 6), indicating that less water
et al., 2012a,b; Qian et al., 2007). There is consensus among many and more water would have been applied for irrigating rice com-
researches that this kind of decline in solar radiation should be pared with that without climate change occurring for Kaifeng
attributed to increasing aerosol loading (e.g., Kaiser and Qian, 2002; and Kunshan, respectively. While for IWUE and WUE, significant
Forster et al., 2007; Ramanathan et al., 2007). However, the recent decreasing trends can be found in all three stations, although ET
study by Chiacchio and Wild (2010) suggested that surface solar and IWR also decease in Kaifeng, which should be attributed to
radiation in fact presented decadal variations with climate shift noticeable decline in rice yields in all stations.
and related with the Pacific Decadal Oscillation (PDO). It should Although there is some spatial variability, the generally nega-
be noted that although most climatic variables for the three sta- tive effect of future climate warming to rice yield under both A2
tions demonstrated the same change directions (some of them are and B2 scenario was found at three stations during the 21 cen-
insignificant), such as decrease in solar radiation and water vapor tury due to remarkable shortened growth period (Fig. 8), especially
 W. Wang et al. / Agricultural Water Management 146 (2014) 249–261 259

Fig. 10. Changes of water use efficiencies (WUE) and irrigation water use efficiencies (IWUE) of rice relative to the baseline without/with CO2 fertilization under A2 and B2
scenarios.

for the two stations in the Yangtze basin. This is consistent with and the uncertainty in model structure and parameters (Yao et al.,
other similar studies in this region (e.g., Shen et al., 2011). Mean- 2011). While the issue of how to adequately address the uncertainty
while, our simulation results of comparison of rice yields with and associated with assessing the impact of climate change still remains
without CO2 fertilization suggest that increase of CO2 concentration a critical and challenging one. Besides, the current simulated results
under future climate is conducive to raise the rice yield, although only reflect the feedback of rice growth and water utilization to
the CO2 fertilization effects can not completely compensate for the climate factor and thus should be interpreted cautiously. The sim-
assimilation reduction due to temperature rising. This finding was ulation did not consider rice varieties changes, biotic stresses like
reported in many similar simulation researches (e.g., Guo et al., pests and diseases, irrigated and fertilized conditions. Therefore,
2010; Mo et al., 2009) and special chamber experiments (e.g., Kim the future comprehensive exploration should be conducted under
et al., 2007; Bannayan et al., 2009). Although the response of the ET the support of process-based models with consideration of inter-
and IWR to future climate change is distinct for the three stations, actions of anthropogenic activities, environmental and biological
the decrease for WUE and IWUE can be found in most schemes (with factors.
different emission scenarios and periods). It is particular true for
the two stations in the Yangtze River Basin. Our simulation results 5. Conclusion
also revealed that elevated CO2 have negative effect on ET and IWR,
but improve the water use efficiencies of rice in a certain degree. The response of rice yield, water consumption, irrigation water
Generally, adaptation countermeasures should be developed and requirements and water use efficiency to climate change in three
put forward to mitigate the adverse effect of climate warming, typical rice plantation sites (i.e., Kunshan, Nanjing and Kaifeng)
but to take the advantages of CO2 fertilization for the sustainable were simulated by driving ORYZA2000 model with downscal-
development of water resources and agriculture. ing HadCM3 outputs by a statistical downscaling model (SDSM).
With downscaled climate data driving the crop model we got the Simulated and measured crop information matched well for the
future picture of how rice yield, ET, IWE, WUE and IWE response calibration and validation period suggesting that the model was
to the climate change. The picture demonstrated the informa- applicable to simulate the yield and water consumption. SDSM
tion of growth and water utilization at a relatively high temporal model performed well in reproducing the present climatic vari-
resolution, which can provide a more precise view for decision ables and thus was adaptable to project future climatic variables
makers than the simulation results under raw GCMs data. However, as an effective tool. Significant decline in rice yield was identified
inevitable uncertainties associated with assessing the response of by 49.3 kg ha−1 , 32.0 kg ha−1 and 45.8 kg ha−1 for Kunshan, Nanjing
rice growth and water consumption to climate change should exist and Kaifeng, respectively, in the past 50 years. Significant increase
from the uncertainty of future climate change scenario (Tao et al., of rice water demand and irrigation water requirement in Kun-
2008), the choice of downscaling methods (Wang et al., 2014), shan, non-significant increase in Nanjing and significant decrease
260 W. Wang et al. / Agricultural Water Management 146 (2014) 249–261

in Kaifeng were found. While accompanying production reduction, Chartzoulakis, K., Psarras, G., 2005. Global change effects on crop photosynthesis and
WUE and IWUE showed consistently significant deceasing trends in production in Mediterranean: the case of Crete, Greece. Agric. Ecosyst. Environ.
106, 147–157.
the three stations. Under IPCC SRES A2 and B2 emission scenarios, Chavas, D.R., Izaurralde, R.C., Thomson, A.M., Gao, X., 2009. Long-term climate
the generally negative effect of climate warming to rice yield was change impacts on agricultural productivity in eastern China. Agric. For. Mete-
found at three stations during the 21 century due to remarkably orol. 149, 1118–1128.
Chen, C., Wang, E., Yu, Q., Zhang, Y., 2010. Quantifying the effects of climate trends in
shortened growth period, resulting in generally depressed WUE the past 43 years (1961–2003) on crop growth and water demand in the North
and IWUE in despite of the distinct response of the ET and IWR China Plain. Clim. Change 100, 559–578.
to future climate change for the three stations. Moreover, com- Chiacchio, M., Wild, M., 2010. Influence of NAO and clouds on longterm seasonal
variations of surface solar radiation in Europe. J. Geophys. Res. 115, D00D22,
parisons of simulation results with and without CO2 fertilization
http://dx.doi.org/10.1029/2009JD012182.
suggested that increase of CO2 concentration under future climate da Silva, V.P.R., 2004. On climate variability in Northeast of Brazil. J. Arid Environ.
is conducive to raise the rice yield, alleviate crop water consump- 58, 575–596.
Döll, P., 2002. Impact of climate change and variability on irrigation requirement: a
tion and irrigation water requirements and improve the water use
global perspective. Clim. Change 54, 269–293.
efficiencies of rice in a certain degree. Egli, D.B., 2008. Soybean yield trends from 1972 to 2003 in mid-western USA. Field
Crops Res. 106, 53–59.
Eitzinger, J., Kubu, G., 2009. Impact of Climate Change and Adaptation in Agricul-
Acknowledgements ture. Extended Abstracts of the International Symposium. June 22–23 2009,
University of Natural Resources and Applied Life Sciences (BOKU), Vienna,
1994-4179 (Print), 1994-4187 (Online), http://www.boku.ac.at/met/report
This study was jointly supported by the National Natural Sci- BOKU-Met Report 17.
ence Foundation of China (51379057, 51009046), the National Feng, L., Bouman, B.A.M., Tuong, T.P., Cabangon, R.J., Li, Y., Lu, G., Feng, Y., 2007.
key Technologies Research and Development Program of China Exploring options to grow rice using less water in northern China using a mod-
elling approach I. Field experiments and model evaluation. Agric. Water Manage.
during the 12th Five-year Plan Period (2013BAC10B01), Qing Lan 88, 1–13.
Project, the Fund of Advanced Science and Technology Innovation Fischer, G., Tubiello, F.N., Velthuizen, H.V., Wiberg, D.A., 2007. Climate change
in Colleges and Universities in Jiangsu Province, the Program for impacts on irrigation water requirement: effects of mitigation, 1990–2080.
Technol Forecast. Soc Change 74, 1083–1107.
Distinguished Talents in Hohai University, a Project Funded by the Forster, P.V., et al., 2007. Changes in atmospheric constituents and in radiative forc-
Priority Academic Program Development of Jiangsu Higher Educa- ing. In: Solomon, S., et al. (Eds.), Climate Change 2007: The Physical Science Basis,
tion Institutions (PAPD), and CSIRO Computational Sciences and Contribution of Working Group I to the Fourth Assessment Report in the Inter-
governmental Panel on Climate Change. Cambridge University Press, Cambridge,
Simulation Capability Platform. The authors gratefully acknowl- pp. 153–179.
edge the IRRI for sharing the validated ORYZA model for Kaifeng Gong, L., Xu, C.Y., Chen, D., Halldin, S., 2006. Sensitivity of the Penman—Monteith ref-
and Nanjing on the Internet. Thanks to the National Meteoro- erence evapotranspiration to key climatic variables in the Changjiang (Yangtze
River) basin. J. Hydrol. 329, 620–629.
logical Information Center, China Meteorological Administration
Guo, R., Lin, Z., Mo, X., Yang, C., 2010. Responses of crop yield and water use efficiency
(http://cdc.cma.gov.cn/) for offering the meteorological data and to climate change in the North China Plain. Agric. Water Manage. 97, 1185–1194.
also thanks to the Canadian Climate Impacts and Scenarios project Haan, C.T., 2002. Statistical Methods in Hydrology (2nd Version). Blackwell, Ames,
(http://co-public.lboro.ac.uk/cocwd/SDSM/) for providing the sta- IA.
Huntington, T.G., 2006. Evidence for intensification of the global water cycle: review
tistical downscaling model (SDSM) and data. and synthesis. J. Hydrol. 319, 83–95.
Hussain, S.S., Mudasser, M., 2007. Prospects for wheat production under changing
climate in mountain areas of Pakistan—an econometric analysis. Agric. Syst. 94,
References 494–501.
IPCC, 2001. Climate Change 2001: The Scientific Basis. Cambridge University Press,
Alan, D., Justin, Z., Edwin, S.P.M., Bart, N., Eric, F.W., Dennis, P.L., 2003. Detection Cambridge, pp. 785.
of intensification in global and continental-scale hydrological cycles: temporal IPCC, 2007. Summary for policymakers. In: Solomon, S., Qin, D., Manning, M.,
scale of evaluation. J. Clim. 16, 535–547, http://dx.doi.org/10.1175/1520-0442. Chen, Z., Marquis, M., Avery, K.B., Tignor, M., Miller, H.L. (Eds.), Climate
Alexander, L.V., Zhang, X., Perterson, T.C., Caesar, J., Gleason, B., Klein Tank, Change 2007: The Physical Science Basis. Contribution of Working Group I
A.M.G., Haylock, M., Collins, D., Trewin, B., Rahimzadeh, F., Tagipour, A., to the Fourth Assessment Report of the Intergovernmental Panel on Climate
Rupa Kumar, K., Revadekar, J., Griffiths, G., Vincent, L., Stephenson, D.B., Change. Cambridge University Press, Cambridge, United Kingdom, New York,
Burn, J., Aguilar, E., Brunet, M., Taylor, M., New, M., Zhai, P., Rusticucci, NY, lang;http:/www.ipcc.ch/SPM2feb07.pdf.
M., Vazquez-Aguirre, J.L., 2006. Global observed changes in daily climate Kaiser, D.P., Qian, Y., 2002. Decreasing trends in sunshine duration over China for
extremes of temperature and precipitation. J. Geophys. Res. 111, D05109, 1954–1998: indication of increased haze pollution? Geophys. Res. Lett. 29 (21),
http://dx.doi.org/10.1029/2005JD0 06290. 2042, http://dx.doi.org/10.1029/2002GL016057.
Bachelet, D., Brown, D., Bohm, M., Russell, P., 1992. Climate change in Thailand and Karim, M.R., Ishikawa, M., Ikeda, M., Islam, M.T., 2012. Climate change model predicts
its potential impact on rice yield. Clim. Change 21, 347–366. 33% rice yield decrease in 2100 in Bangladesh. Agron. Sustain. Dev. 32, 821–830.
Bannayan, M., Tojo Soler, C.M., Garcia, A., Garcia, Y., Guerra, L.C., Hoogenboom, G., Kendall, M.G., 1975. Rank Correlation Methods. Griffin, London.
2009. Interactive effects of elevated CO2 and temperature on growth and devel- Kim, S., Gitz, D.C., Sicher, R.C., Baker, J.T., Timlin, D.J., Reddy, V.R., 2007. Tempera-
opment of a short- and long-season peanut cultivar. Clim. Change 93 (3–4), ture dependence of growth, development, and photosynthesis in maize under
389–406. elevated CO2 . Environ. Exp. Bot. 61, 224–236.
Boling, A., Bouman, B.A.M., Tuong, T.P., Murty, M.V.R., Jatmiko, S.Y., 2007. Modeling Komuscu, A.U., Erkan, A., Oz, S., 1998. Possible impacts of climate change on soil
the effect of groundwater depth on yield-increasing interventions in rainfed moisture availability in the southeast Anatolia Development Project Region: an
lowland rice in Central Java, Indonesia. Agric. Syst. 92, 115–139. analysis from an agricultural drought perspective. Clim. Change 40, 519–545.
Bouman, B.A.M., Kropff, M.J., Tuong, T.P., Wopereis, M.C.S., Ten Berge, H.F.M., Van Kou, X., Ge, J., Wang, Y., Zhang, C., 2007. Validation of the weather generator CLI-
Laar, H.H., 2001. Modelling lowland rice. In: ORYZA2000. International Rice GEN with daily precipitation data from the Loess Plateau, China. J. Hydrol. 347,
Research Institute, Wageningen University and Research Centre, Los Baños, 347–357.
Philippines Wageningen, The Netherlands, p. 235. Lin, E., Xiong, W., Ju, H., Xu, Y., Yue, Li, Bai, L., Xie, L., 2005. Climate change impacts
Bouman, B.A.M., Lampayan, R.M., Tuong, T.P., 2007. Water Management in Irri- on crop yield and quality with CO2 fertilization in China. Philos. Trans. R. Soc., B
gated Rice: Coping with Water Scarcity. International Rice Research Institute, 360, 2149–2154.
Los Baños. Liu, X., Yin, Z., Shao, X., Qin, N., 2006. Temporal trends and variability of daily max-
Bouman, B.A.M., Van Laar, H.H., 2006. Description and evaluation of the rice growth imum and minimum, extreme temperature events, and growing season length
model ORYZA2000 under nitrogen-limited conditions. Agric. Syst. 87, 249–273. over the eastern and central Tibetan Plateau during 1961–2003. J. Geophys. Res.
Brumbelow, K., Georgakakos, A., 2001. An assessment of irrigation needs and crop 111, D19109, http://dx.doi.org/10.1029/2005JD006915.
yield for the United States under potential climate changes. J. Geophys. Res. 106, Liu, S., Mo, X., Lin, Z., Xu, Y., Ji, J., Wen, G., Richey, J., 2010. Crop yield responses to
27383–27405. climate change in the Huang–Huai–Hai Plain of China. Agric. Water Manage. 97,
Cabangon, R., Lu, G., Tuong, T.P., Bouman, B.A.M., Feng, Y., Zhang, Z., 2003. Irrigation 1195–1209.
management effects on yield and water productivity of inbred and aerobic rice Malone, R.W., Meek, D.W., Hatfield, J.L., Mann, M.E., Jaquis, R.J., Ma, L., 2009. Quasi-
varieties in Kaefeng. Proceedings of the First International Yellow River Forum biennial corn yield cycles in Iowa. Agric. For. Meteorol. 149, 1087–1094.
on River Basin Management, vol. 2, The Yellow River Conservancy Publishing Mann, H.B., 1945. Nonparametric tests against trend. Econometrica 13, 245–259.
House, Zhengzhou, Henan, China, pp. 65–76. McVicar, T.R., et al., 2012. Global review and synthesis of trends in observed terres-
Challinor, A.J., Wheeler, T.R., 2008. Crop yield reduction in the tropics under climate trial near-surface wind speeds: implications for evaporation. J. Hydrol. 416–417,
change: processes and uncertainties. Agric. For. Meteorol. 148, 343–356. 182–205.
 W. Wang et al. / Agricultural Water Management 146 (2014) 249–261 261

Mimi, Z.A., Jamous, S.A., 2010. Climate change and agricultural water demand: under changing climate. Adv. Water Sci. 23, 656–664 (in Chinese with English
impacts and adaptations. Afr. J. Environ. Sci. Technol. 44, 183–191. abstract).
Mo, X., Liu, S., Lin, Z., Guo, R., 2009. Regional crop yield, water consumption and Wang, W., Shao, Q., Peng, S., Xing, W., Yang, T., Luo, Y., Yong, B., Xu, J., 2012b. Ref-
water use efficiency and their responses to climate change in the North China erence evapotranspiration change and the causes across the Yellow River Basin
Plain. Agric. Ecosyst. Environ. 134, 67–78. during 1957–2008 and their spatial and seasonal differences. Water Resour. Res.
Papaioannou, G., Kitsara, G., Athanasatos, A., 2011. Impact of global dimming and 48, W05530, http://dx.doi.org/10.1029/2011WR010724.
brightening on reference evapotranspiration in Greece. J. Geophys. Res. 116, Wang, W., Shao, Q., Xing, W., Yu, Z., Peng, S., 2014. “Closure to estimating the effects
D09107, http://dx.doi.org/10.1029/2010JD015525. of climatic variability and human activities on streamflow in the Hutuo River
Peng, S.B., Huang, J.L., John, E.S., 2004. Rice yields decline with higher night temper- Basin, China” by Shizhang Peng, Wanxin Liu, Weiguang Wang, Quanxi Shao,
ature from global warming. PNAS 101 (27), 9971–9975. Xiyun Jiao, Zhongbo Yu, Wanqiu Xing, Junzeng Xu, Zengxin Zhang, and Yufeng
Piao, S., Ciais, P., Huang, Y., Shen, Z., Peng, S., Li, J., Zhou, L., Liu, H., Ma, Y., Ding, Y., Luo. Hydrol. Eng. 19, 836–839.
Friedlingstein, P., Liu, C., Tan, K., Yu, Y., Zhang, T., Fang, J., 2010. The impacts Wang, W., Shao, Q., Yang, T., Peng, S., Yu, Z., Taylor, J., Xing, W., Zhao, C., Sun, F.,
of climate change on water resources and agriculture in China. Nature 467, 2013. Changes in daily temperature and precipitation extremes in the Yellow
43–51. River Basin, China. Stochastic Environ. Res. Risk Assess. 27, 401–421.
Qian, Y., Wang, W., Leung, L.R., Kaiser, D.P., 2007. Variability of solar radiation under Wilby, R.L., Dawson, C.W., Barrow, E.M., 2002. SDSM-a decision support tool for
cloud-free skies in China: the role of aerosols. Geophys. Res. Lett. 34, L12804, the assessment of regional climate change impacts. Environ. Modell. Softw. 17,
http://dx.doi.org/10.1029/2006GL028800. 147–159.
Ramanathan, V., Ramana, M.V., Roberts, G., Kim, D., Corrigan, C., Chung, C., Winker, Wild, M., 2009. Global dimming and brightening: a review. J. Geophys. Res. 114,
D., 2007. Warming trends in Asia amplified by brown cloud solar absorption. D00D16, http://dx.doi.org/10.1029/2008JD011470.
Nature 448, 575–578. WMO, 2002. Atmospheric RCMs: a multi-purpose tool? In: Report of the joint
Shen, S., Yang, S., Zhao, Y., Xu, Y., Zhao, X., Wang, Z., Liu, J., Zhang, W., 2011. Simulating WGNE/WGCM Ad Hoc Panel on Regional Climate Modeling. WMO.
the rice yield change in the middle and lower reaches of the Yangtze River under Xiong, W., Holman, I., Conway, D., Lin, E., Li, Y., 2008. A crop model cross calibration
SRES B2 scenario. Acta Ecol. Sin. 31, 40–48. for use in regional climate impacts studies. Ecol. Modell. 213, 365–380.
Silva, C.S.D., Weatherhead, E.K., Knox, J.W., Rodriguez-Diaz, J.A., 2007. Predicting the Xiong, W., Matthews, R., Holman, I., Lin, E., Xu, Y., 2007. Modelling China’s poten-
impacts of climate change—a case study of paddy irrigation water requirements tial maize production at regional scale under climate change. Clim. Change 85,
in Sri Lanka. Agric. Water Manage. 93, 19–29. 433–451.
State Environmental Protection Administration, 2003. Progress Report of Xiong, Z.M., Cai, H.F., Min, S.K., Li, B.C., 1992. Rice in China. Chinese Publishing Press of
Trade Liberalization in the Agriculture Sector and the Environment, with Agricultural Science and Technology, Beijing (in Chinese with English abstract).
Specific Focus on the Rice Sector in China. State Environmental Protec- Xiong, W., Conway, D., Lin, E.D., Xu, Y., Ju, H., Jiang, J., Holman, I.P., Li, Y., 2009. Future
tion Administration, Available at http://www.unep.ch/etb/events/Events2003/ cereal production in China: modelling the interaction of climate change, water
pdf/FinalDraftofChinaStudy.pdf. availability and socio-economic scenarios. Global Environ. Change 19 (1), 34–44.
Stockle, C.O., Campbell, G.S., Nelson, R., 1997. ClimGen for Windows: A Weather Xu, J., Peng, S., Yang, S., Wang, W., 2012. Ammonia volatilization loss from a rice
Generator Program. Biological Systems Engineering Dept., Washington State paddy with different irrigation and nitrogen managements. Agric. Water Man-
University, Pullman, WA, http://www.bsyse.wsu.edu/climgen. age. 104, 184–192.
Tao, F., Hayashi, Y., Zhang, Z., Sakamoto, T., Yokozawa, M., 2008. Global warming, Yadav, S., Li, T., Humphreys, E., Gill, G., Kukal, S.S., 2011. Evaluation and application
rice production, and water use in China: developing a probabilistic assessment. of ORYZA2000 for irrigation scheduling of puddled transplanted rice in north
Agric. For. Meteorol. 148, 94–110. west India. Field Crops Res. 122, 104–117.
Tao, F., Yokozawa, M., Hayashi, Y., Lin, E., 2003. Future climate change, the agricul- Yang, J.C., Zhang, J.H., 2006. Grain filling of cereals under soil drying. New Phytol.
tural water cycle, and agricultural production in China. Agric. Ecosyst. Environ. 169, 223–236.
95, 203–215. Yao, F., Qin, P., Zhang, J., Lin, E., Boken, V., 2011. Uncertainties in assessing the
Tao, F., Yokozawa, M., Xu, Y., Hayashi, Y., Zhang, Z., 2006. Climate changes and trends effect of climate change on agriculture using model simulation and uncertainty
in phenology and yields of field crops in China, 1981–2000. Agric. For. Meteorol. processing methods. Chin. Sci. Bull. 56 (8), 547–555.
138, 82–92. Yao, F., Xu, Y., Lin, E., Yokozawa, M., Zhang, J., 2007. Assessing the impacts of climate
Tao, F., Zhang, Z., 2013. Climate change, wheat productivity and water use in the change on rice yields in the main rice areas of China. Clim. Change 80, 395–409.
North China Plain: a new super-ensemble-based probabilistic projection. Agric. Yue, S., Pilon, P., Phinney, B., Cavadias, G., 2002. The influence of autocorrela-
For. Meteorol. 170, 146–165. tion on the ability to detect trend in hydrological series. Hydrol. Processes 16,
Thomas, A., 2008. Agricultural irrigation demand under present and future climate 1807–1829.
scenarios in China. Global Planet. Change 60, 306–326. Zhang, Q., Xu, C.Y., Zhang, Z., Ren, G., Chen, Y.D., 2008a. Climate change or variability?
Thomson, A.M., Izaurralde, R.C., Rosenberg, N.J., He, X., 2006. Climate change impacts The case of Yellow River as indicated by extreme maximum and minimum air
on agriculture and soil carbon sequestration potential in the Huang–Hai Plain temperature during 1960–2004. Theor. Appl. Climatol. 93 (1–2), 35–43.
of China. Agric. Ecosyst. Environ. 114, 195–209. Zhang, T., Zhu, J., Yang, X., Zhang, X., 2008b. Correlation changes between rice yields
Trnka, M., Dubrovsky, M., Semeradova, D., Zalud, Z., 2004. Projections of uncertain- in north and northwest China and ENSO from 1960 to 2004. Agric. For. Meteorol.
ties in climate change scenarios into expected winter wheat yields. Theor. Appl. 148, 1021–1033.
Climatol. 77, 229–249. Zhang, X.C., Liu, W.Z., 2005. Simulating potential response of hydrology, soil erosion,
Wang, W., Peng, S., Sun, F., Xing, W., Luo, F., Xu, J., 2012a. Spatialtemporal variations and crop productivity to climate change in Changwu tableland region on the
of rice irrigation water requirements in the mid-lower reaches of Yangtze River Loess Plateau of China. Agric. For. Meteorol. 131, 127–142.