Development of a gridded dataset of annual irrigation water withdrawal in China

Xiufang Zhu, Peijun Shi, Yaozhong Pan

State Key Laboratory of Earth Processes and Resource Ecology, College of Resources Science & Technology, Beijing Normal University, 100875, Beijing, China

Abstract—Irrigated agriculture in China is of national and global significance. Official statistical data on agricultural water withdrawal might be more relevant to the actual water consumption in an agriculture sector than the irrigation requirements estimated from modeling studies. However, agricultural water withdrawal is always reported on the basis of geopolitical units or river basins, without spatial distribution information. Here, we developed a time series gridded datasets of annual irrigation water withdrawal in China from 1978 to 2008 by disaggregating the agricultural water withdrawal data from the geopolitical units to individual 1-km grids. This new dataset is the first to portray the spatially explicit distribution of actual water withdrawal in China. With this new dataset, we could better estimate and evaluate the crop yield, water stress, ground water decline, drought resistance, water use efficiency, and irrigation impact on the near surface climate in China.

Keywords- irrigation water withdrawal; disaggregation; time series; gridded dataset; China

I. INTRODUCTION Water is not only the bloodstream of the biosphere but also

a key component of society and its life support system. At present, water scarcity is becoming a huge challenge in that it can slow down the socio-economic development; increase poverty; endanger environmental security, water security, and food security; and further contribute to social instability.

The irrigation agriculture sector is the largest consumer of water on a global scale [1], with China having the second largest irrigated area in the whole world [2]. On one hand, the irrigation agriculture of China stabilizes the food prices, increases the farmers’ income, and helps to feed over 1.3 billion of the Chinese population, which is approximately 20% of the global population. On the other hand, since 1949, the total irrigated area in China has almost increased monotonically from 15 million ha to approximately 56 million ha, and the agricultural water withdrawal has increased from 100.1 billion m3 to 366.4 billion m3. This has caused many environmental problems such as drying up of the rivers [3, 4], deficit in groundwater [5-7], and salinization of soil [8-10]. Irrigated agriculture in China is therefore of national and global significance when examining the issues of food, water, and environmental security.

Information on the volume and distribution of agricultural water withdrawal in China is needed for many studies, such as the optimization of agricultural water management, projection of future water and food security, and influence of irrigation systems on earth’s hydrosphere, atmosphere, and biosphere. There are two types of data sources for irrigation water

withdrawal in China. One source is the estimates obtained from modeling studies [11-13]; the other source is the official statistical data collected by the Ministry of Water Resources. Irrigation modeling can simulate the irrigation requirement for optimal crop growth; however, the estimated irrigation amount may drift away from the actual amount owing to two reasons. First, water crises are more serious in the current era than at any other time because of the rapid population increase, continuous economic growth, and extended drought disasters. Farmers are more likely to use supplemental rather than full irrigation in order to increase grain yield under great water stress. In other words, the irrigation requirement is more likely to be higher than the real irrigation water quota. Second, agricultural water is consumed not only by field crops but is also wasted in the process of delivery from water source to the fields. In many countries, especially in poor developing countries, the inadequately maintained irrigation infrastructures, backward water management systems, and deficient agriculture investments contribute to lower the water use efficiency. As a result, more water is consumed than the actual water required for growing healthy crops. Conversely, official statistical data of agricultural water withdrawal might be more relevant to the actual water consumption in the agriculture sector. However, agricultural water withdrawal is always reported on the basis of geopolitical units or river basins, without detailed distribution information.

In this study, we developed a time series gridded dataset of annual irrigation water withdrawal in China from 1978 to 2008 by using historical agricultural water withdrawal data and an irrigation map of China in 2000.

II. DATA AND METHODOLOGY Fig.1 shows our general method for developing the gridded

dataset of annual irrigation water withdrawal in China from 1978 to 2008. Table I summarizes the data used in our study and their purposes. More details about the data are presented below.

A. Data 1) Effective irrigated area

In China, irrigated areas are reported on the basis of census data. Conventional local-level Bureau of Statistics reports include irrigation statistics at the village level, which is then aggregated to higher levels such as county, city, province, and national levels. The National Bureau of Statistics regularly releases data on provincial irrigated areas in its Statistics Yearbook. In this study, we collected provincial effective

Figure 1. Technoloy flowchart

TABLE I. THE DATA USED IN OUR STUDY AND THEIR PURPOSES

Data Utility National effective irrigation area

To help fill the missing national agricultural water withdrawal data by linear interpolation

National agricultural water withdrawal

1)To help fill the missing national agricultural water withdrawal data by linear interpolation 2)To control the total irrigation water withdrawal in China

Provincial agricultural water withdrawal from 2003 to 2008

To calculate the gross irrigation quota (GIQ) for each province with the help of provincial effective irrigated area from 2003 to 2008

Provincial effective irrigation area

1)To calculate the gross irrigation quota (GIQ) for each province with the help of provincial agricultural water withdrawal from 2003 to 2008 2)To help to allocate the national agricultural water withdrawal to the provincial irrigation quotas from 1978 to 2002

The ratio of irrigation water consumed to agricultural water withdrawal for each province

To calculate the irrigation water withdrawal from agricultural water withdrawal for each province

Net irrigation requirements of main crops for each province

To help to allocate irrigation water withdrawal from province to drayland and paddy fields in this province

Irrigation map of China for 2000

To help to distribute the irrigation water withdrawal of province to each pixel in this province

irrigation area (EIA) data for 31 provinces from 1978 to 2008, with no data available for Taiwan, Hong Kong, and Macao.

2) National agricultural water withdrawal Similar to EIA, agricultural water withdrawal (AWW) data

in China are reported in the National Statistical Yearbook; however, in most years, only the national sums are reported. In this study, national agricultural water withdrawal data from 1997 to 2008 were taken from the China Statistical Yearbook of 1998 to 2009 and those of 1978, 1980, 1985, 1990, 1993, and 1995 were from reference (Wang et al. 2010).

3) Gross irrigation quota The Gross Irrigation Quota (GIQ) is defined as the ratio of

the agricultural irrigation water quantity to the effective irrigation area [15]. The annual provincial agricultural water withdrawal information since 2002 has been reported in the China Statistical Yearbook. The China Statistical Yearbook also lists the effective irrigation area for each province. Therefore, we can calculate the gross irrigation quota for each province from 2002 to 2008; this calculated information is shown in Table II.

TABLE II. THE MEAN GIQ OF EACH PROVINCE DURING 2002 TO 2008

Province Mean GIQ (mm) Province Mean GIQ (mm)

Beijing 625 Hubei 652

Tianjin 351 Hunan 753

Hebei 335 Guangdong 1686

Shanxi 294 Guangxi 1402

Inner Mongolia 544 Hainan 1965

Liaoning 578 Chongqing 317

Jilin 442 Sichuan 479

Heilongjiang 776 Guizhou 696

Shanghai 687 Yunnan 730

Jiangsu 703 Tibet 1726

Zhejiang 755 Shaanxi 412

Anhui 363 Gansu 919

Fujian 1089 Qinghai 1117

Jiangxi 724 Ningxia 1615

Shandong 338 Xinjiang 1436

Henan 257

4) The ratio of irrigation water withdrawal to agricultural water withdrawal of each province

Agricultural water withdrawal refers to the “Water Use by Agriculture”; it “includes uses of water by irrigation of farming fields and by forestry, animal husbandry and fishing. Water use by forestry, animal husbandry and fishery includes irrigation of forestry and orchards, irrigation of grassland and replenishment of fishing farms.” Therefore, to obtain the irrigation water withdrawn by fields used for farming, we need know the ratio of irrigation water withdrawal to agricultural water withdrawal. Table III shows the ratios we collected from different sources for each province.

Step 4

Step 3

Step 2Step 1National agricultural water

withdrawal in 1978, 1980, 1985, 1990, 1993, 1995, and 1997 to

2008 and national effective irrigated area during 1978-20008

Time series agricultural water withdrawal from 1978 to 2008

Provincial agricultural water withdrawal from

2002 to 2008

Irrigation quota ratio among provinces

Provincial irrigation quota from 1978 to 2008

Mean irrigation water requirements of dryland and paddy fields in each Province

Irrigation water withdrawal per square kilometer of dryland and paddy fields in province i and year j

Irrigation water withdrawal at pixel k in year j

Effective irrigated area of each province

from 1978 to 2001

The area of irrigated dryland and paddy fields

Equation 2

Equation 4 & 5

Equation 6

Equation1

Filling data gap

Irrigation map of China 2000

Step 1: fill in missing national agricultural water withdrawals dataStep 2: allocate the historical records of national agricultural water withdrawal to provincial agricultural water withdrawal before 2002Step3: calculate the provincial irrigation water withdrawal during 1978-2008

The ratio of irrigation water withdrawal to agricultural water withdrawal of each province Equation 3

Irrigation water withdrawal of each province

Step4: downscale irrigation water withdrawal from province to pixels in a given year based on the irrigation map

TABLE III. THE RATIO OF IRRIGATION WATER WITHDRAWAL TO AGRICULTURAL WATER WITHDRAWAL OF EACH PROVINCE

Province Ratio Source

Beijing 0.79 (Song et al. 2010)

Tianjin 0.99 Tianjin Water Resources Bulletin 2001

Hebei 0.96 Thematic Database for Human-Earth System

Shanxi 0.96 Thematic Database for Human-Earth System

Inner Mongolia 0.95 Thematic Database for Human-Earth System

Liaoning 0.98 Thematic Database for Human-Earth System

Jilin 0.95 Thematic Database for Human-Earth System

Heilongjiang 0.92 Thematic Database for Human-Earth System

Shanghai 0.92 (Wang et al. 2010)

Jiangsu 0.94 Thematic Database for Human-Earth System

Zhejiang 0.92 (Wang et al. 2010)

Anhui 0.95 Thematic Database for Human-Earth System /

Fujian 0.92 (Wang et al. 2010)

Jiangxi 0.92 (Wang et al. 2010)

Shandong 0.95 Thematic Database for Human-Earth System

Henan 0.93 Thematic Database for Human-Earth System

Hubei 0.92 (Wang et al. 2010)

Hunan 0.97 Hunan Water Resources Bulletin 2007

Guangdong 0.83 China Water Resources

Guangxi 0.94 Guangxi Water Resources

Hainan 0.77 Hainan Water Resources Bulletin 2008

Chongqing 0.92 Chongqin Water Resources Bulletin 2001

Sichuan 0.96 (He and Lu 2007)

Guizhou 0.95 Guizhou Water Resources Bulletin 2001

Yunnan 0.97 Yunnan Water Resources Bulletin 2000

Tibet 0.61 (Wa 2009)

Shaanxi 0.92 Thematic Database for Human-Earth System

Gansu 0.95 Thematic Database for Human-Earth System

Qinghai 0.83 Qinghai Water Resources Bulletin 2007

Ningxia 0.92 Thematic Database for Human-Earth System

Xinjiang 0.77 Xinjiang Water Resources Bulletin 2001

Note: Thematic Database for Human-Earth System is available at http://www.data.ac.cn/

5) Net Irrigation requirements The estimation of the irrigation requirements for the main

crops in different areas of China is taken directly from the literature [16] in which the authors considered the regional differences in climate and soil conditions. They divided the map of China into 216 subregions based on differences in climate conditions and water sources, chose a typical meteorological station for each subregion, and collected the daily meteorological observations including precipitation, mean air temperature, maximum air temperature, minimum air temperature, wind speed, relative humidity, and sunshine hours from 1970 to 2000. They calculated the net irrigation

requirement (NIR) by subtracting effective precipitation from the irrigation requirement during crop growing season in each subregion. The crop irrigation requirement was calculated using the Penman-Monteith method. They reported the average irrigation requirements and average net irrigation requirements of the main crops (including corn, wheat, rice etc.) in different areas of China during 1970–2000.

To simplify our method, we used only the mean value of the maximum and minimum net irrigation requirements reported by the authors for each province to represent the mean net irrigation requirements of the entire area of the province. Because wheat and corn are planted in dryland and rice is planted is paddy field, in this study, we summed up the NIR of wheat (spring wheat, winter wheat) and corn (spring corn, summer corn) estimated by Liu et al. (2009) to get the NIR of dryland and summed up the NIR of rice (including early rice, middle rice and late rice) estimated by Liu et al. (2009) to get the NIR of paddy field for each province(Table IV).

TABLE IV. MEAN NET IRRIGATION REQUIREMENTS FOR DRYLAND AND PADDY FIELD IN DIFFERENT PROVINCES OF CHINA FROM 1970-2000

Province NIR_ Dryland

NIR_ Paddy Province NIR_

Dryland NIR_

Paddy Beijing 375 475 Hubei 100 460 Tianjin 375 475 Hunan 100 460 Hebei 375 475 Guangdong 100 460 Shanxi 730 550 Guangxi 100 460

Inner Mongolia 755 675 Hainan 100 460 Liaoning 115 265 Chongqing 225 150

Jilin 115 265 Sichuan 285 150 Heilongjiang 115 265 Guizhou 190 300

Shanghai 100 285 Yunnan 190 300 Jiangsu 250 285 Tibet 350 -

Zhejiang 100 750 Shaanxi 445 550 Anhui 250 750 Gansu 730 550 Fujian 100 630 Qinghai 350 - Jiangxi 100 460 Ningxia 755 675

Shandong 375 475 Xinjiang 785 750 Henan 375 475

6) Irrigation maps of China for 2000 We used the irrigation map of China for 2000 from our

previous study [17] to indicate the extent of irrigated areas and rainfed areas. In that study, we developed three irrigation probability indices by using the time series Normalized Difference Vegetation Index (NDVI) and precipitation data. Using these indices, we proposed a spatial allocation model for allocating the census data on irrigation from geopolitical units to individual pixels.

B. Methodology Generally, our method is a serial of downscaling processes:

downscaling national agricultural water withdrawal (AWW) to provincial AWW, downscaling provincial AWW to water consumption of irrigated dryland and paddy field in a given province, and calculating the AWW per square kilometer (km2) of irrigated dryland and irrigated paddy field in this province. Finally, the AWW of a given pixel equals to AWW of irrigated dryland within this pixel plus AWW of irrigated paddy field within this pixel. Besides the previous process, a conversion from agricultural water withdrawal to irrigation water withdrawal is also needed. More details are described below.

First, the missing national agricultural water withdrawal data were filled by applying linear interpolation with the following equation:

( )p oj j o o

p o

EIA EIAEIA EIA

AWW AWWAWW AWW

⎛ ⎞−×⎜ ⎟⎜ ⎟−⎝

− +⎠

= (1)

where AWWj is the calculated agricultural water withdrawal (m3) of China in year j; EIAj is the effective irrigated area (Kha) of China in year j; AWWp and AWWo are the reported agricultural water withdrawal of China in years p and o, respectively; and EIAp and EIAo are the effective irrigated area of China in years p and o, respectively. p and o are the nearest year with available data before and after year j, respectively.

Second, we allocated the national agricultural water withdrawal before 2002 to the provincial agricultural water withdrawal via the following formulas:

,2002 2008 ,,

,2002 2008 ,1

( )

( ( ) )

i i ji j j i j n

i i ji

mean GIQ EIAAWW AWW R AWW

mean GIQ EIA

−

−=

×= × = ×

×∑

(2)

Where AWWi,j, and EIAi,j are agricultural water withdrawal (m3) and effective irrigated area (kha) of province i in year j, respectively. AWWj is the agricultural water withdrawal of China (m3) in year j (data are taken from the National Statistical Yearbook for most years or calculated from the first step in the missing years). Ri is the ratio of the AWW of province i to the national AWW; it is decided by two factors: the Gross Irrigation Quota (GIQ) and EIA. mean(GIQi,2002-2008) is the mean GIQ of each province i during 2002 to 2008 (shown in Table II). Variable i goes from 1 to n, where n is the number of provinces, excluding Hong Kong, Macao, and Taiwan due to lack of available EIA for them in the China Statistics Yearbook. It should be noted that n changed to 31 after 1997 when Chongqing separated from the Sichuan province, and it was 30 during 1987–1996 and 29 during 1978–1986 because no data for that period was available from Hainan province. j covers the period from 1978 to 2008.

Third, we calculated the total irrigation water withdrawal of each province via equation 3.

, ,i j i i jr Arr WWI = × (3)

Where Irri,j is the irrigation water withdrawal (m3) of province i in year j. ri is the ratio of irrigation water withdrawal of farmland to agricultural water withdrawal in province i (show in table III).

Fourth, we calculated the irrigation water withdrawal per square kilometer (km2) of irrigated dryland (Irr_drylandi,j, unit: m) and irrigated paddy field (Irr_paddyi,j, unit: m) in a given province and a given year via equations 4 and 5.

,

, ,,

_

__ __ _i

i j

ij

i i j i i j

Irr paddy

NIR paddyIrrNIR paddy addy NIR drylandS p S dryland× + ×

×

= (4)

,

, ,, _ _

_

__ _

i j

i

i i j i i ji j

Irr dryland

NIR drylandIrrNIR paddy addy NIR drylanS p S drylandd× + ×

×

= (5)

Where S_paddyi,j and S_drylandi,j are the area (km2) of irrigated paddy field and irrigated dryland of province i in year j, respectively. Since it is impossible to obtain the time series values of S_paddyi,j and S_drylandi,j, the values of S_paddyi,j and S_drylandi,j in 2000 are used in each year. The values of S_paddyi,j and S_drylandi,j in 2000 can be derived from the irrigation map of China for 2000. NIR_paddyi,j and NIR_drylandi,j are the estimated net irrigation requirements (mm) of paddy field and dryland in province i, which are derived by referencing to reference [16] and shown in table IV.

Finally, the irrigation water withdrawal of each irrigated pixel can be calculated via the following equation.

, , ,

, ,6

( _ _

_ ) 0_ 1k j i j k j

i j k j

Irr Irr paddy Perc paddy

Irr dryland Perc dryland

= × +

× × (6)

Prec_paddyk,j and Perc_drylandk,j are the irrigation percentages of the paddy fields and dryland at pixel k in year j. The pixel size is 1 km2. Again, because it is impossible to obtain the time series values of Prec_paddyk,j and Perc_drylandk,j the values of Prec_paddyk,j and Perc_drylandk,j in 2000 derived from the irrigation map of China for 2000 are used in each year. Irrk,j is the total irrigation withdrawal(m3) at pixel k in year j.

III. RESULTS TableV shows the national water withdrawal of China. Fig.

2 shows the mean, standard deviation, and change trend of irrigation water withdrawal during 1978 to 2008. From Fig.2, we can see that in most irrigated areas, the mean irrigation water withdrawal did not vary significantly, and the standard deviation of mean irrigation water withdrawal was lower than 0.002 km3 per year. From 1978 to 2008, the irrigation water withdrawal increased in 14 provinces, namely, Xinjiang Uyghur Autonomous Region, Gansu, Qinghai, Inner Mongolia Autonomous Region, Jilin, Liaoning, Heilongjiang, Ningxia Hui Autonomous Region, Tibet Autonomous Region, Guizhou, Yunnan, Hebei, Henan, and Anhui; and decreased in Beijing, Tianjin, Shandong, Jiangsu, Shanghai, Zhejiang, Fujian, Jiangxi, Shanxi, Shaanxi, Sichuan, Chongqing, Hubei, Hunan, Guangdong, Guangxi Zhuang Autonomous Region,

and Hainan. The largest increase was measured in the Hetao plain around Ningxia and Inner Mongolia, in the Northeast China Plain; the largest decreases happened in Guangdong province and Beijing city.

a. Mean irrigation water withdrawal (1978-2008)

b. Standard deviation of irrigation water withdrawal (1978-2008)

c. Change trend of irrigation water withdrawal (1978-2008)

Figure 2. The mean, standard deviation, and changing trend of irrigation water consumption from 1978 to 2008

TABLE V. ANNUAL NATIONAL WATER WITHDRAWAL

Year AWW (108m3) Year AWW

(108m3) Year AWW (108 m3)

1978 3912.00 1989 3602.31 1999 3869.16

1979 4009.07 1990 3763.70 2000 3783.50

1980 3715.90 1991 3776.66 2001 3825.70

1981 3653.09 1992 3800.41 2002 3736.20

1982 3573.76 1993 3804.67 2003 3432.80

1983 3668.41 1994 3805.26 2004 3585.70

1984 3628.84 1995 3821.80 2005 3580.00

1985 3545.60 1996 3876.84 2006 3664.40

1986 3557.91 1997 3919.72 2007 3599.50

1987 3569.38 1998 3766.26 2008 3663.50

1988 3567.82 Note: The bold numbers are for the missing years.

We compared the provincial irrigation water withdrawal calculated by our method with the actual provincial irrigation water withdrawal data reported in the China Water Resources Bulletin of 2000 and result is shown in Fig.3. Cal_irr is mean irrigation water withdrawals calculated by our method in a given province. The provincial irrigation water withdrawal reported in China Water Resources Bulletin 2000 is called Ref_irr. Cal_irr is strongly related to Ref_irr; the correlation coefficient between them is 0.904. The biggest difference between the calculated and reported irrigation water withdrawal occurred in Tibet. In Tibet, the reported irrigation water withdrawal was about half of that calculated by our method.

Figure 3. Linear Regression Cal_irr and Ref_irr

IV. DISCUSSION AND CONCLUSIONS In this study, we developed a time series gridded datasets of

annual irrigation water withdrawal in China from 1978 to 2008 by using the historical effective irrigated area and agricultural water withdrawal data. Several factors contribute to the uncertainties in our products.

First, agricultural water is withdrawn not only for irrigation purposes but also for other agricultural sectors such as fisheries water use. Irrigation is not only applicable to cropland management, but also to forestry and pasture management. In our method, we collected the ratio of irrigation water withdrawal to agricultural water withdrawal around the year 2000, and used this data over the whole study period. However,

the ratio of irrigation water withdrawal for farmland is estimated to decrease with time, as more and more agricultural water is consumed by forestry, pasture, and fishery sectors because of changes in the agricultural structure. As a result, the irrigation water withdrawal before 2000 calculated using our method might be underestimated.

Second, in our method, we used the irrigation map of China for 2000 to represent the irrigated area from 1978 to 2008 and ignored the changes in the irrigated area over the past 31 years because a lack of archived information makes it impossible to obtain a time-series irrigation map in China. The extent of the irrigated area in 2000 is clearly larger than that in 1978; consequently, the irrigation water for irrigated pixels in 1978 is expected to be overestimated. Conversely, the irrigation water of irrigated pixels in 2008 is expected to be underestimated.

Third, the irrigated area in the irrigation maps shows areas that were equipped for irrigation (it is also called as effective irrigation area (EIA), not the real irrigated area. The real irrigated area might be larger or smaller than the EIA. Accordingly, the real irrigation water consumption will be slightly different from our estimation.

Fourth, the irrigation percentages of the paddy fields and dryland in each irrigated pixel in 2000 derived from the new irrigation map of China are used as an estimate of the same ratio over the whole study period (1978-2008). However, the area of irrigated paddy fields and dryland changed during those years.

Fifth, in our method, we used the irrigation requirements for the main crops reported by Liu et al. (2009) to estimate the irrigation water consumption for each main crop in a given province. The irrigation requirements reported by Liu et al. are provincial average estimates from 1970 to 2000. However, the real irrigation requirements vary year–by-year and pixel-by-pixel, because of the different climate and soil conditions.

Finally, because no available data for a robust validation, in this study, we just compared the irrigation water withdrawal map of China in 2000 with the actual provincial irrigation water withdrawal data reported in the China Water Resources Bulletin of 2000 at provincial level. More validations definitely need to be done in the future at pixel scale.

However, these maps are the first gridded datasets that portray a spatially explicit distribution of the actual water withdrawal for both irrigated paddy fields and drylands in China, with a higher resolution and covering a longer time period. We validated the irrigation water withdrawal map of China in 2000 based on the actual provincial irrigation water withdrawal reported in China Water Resources Bulletin 2000. Results show that the correlation coefficient of irrigation water withdrawal calculated from our new map and the actual provincial irrigation water withdrawal is 0.904. The time series gridded datasets of annual irrigation water withdrawal in China

from 1978 to 2008 can help to accurately simulate the irrigation impact on the near surface climate in China. Furthermore, the new dataset can improve our ability to estimate the crop yield, water stress, drought resistance, and water use efficiency in China.

REFERENCES [1] I. A. Shiklomanov, "Appraisal and Assessment of World Water

Resources," Water International, vol. 25, pp. 11-32, 2000. [2] S. Siebert, P. Doll, J. Hoogeveen, J. M. Faures, K. Frenken, and S.

Feick, "Development and validation of the global map of irrigation areas," Hydrology and Earth System Sciences, vol. 9, pp. 535-547, 2005.

[3] X. B. Ji, E. S. Kang, R. S. Chen, W. Z. Zhao, Z. H. Zhang, and B. W. Jin, "The impact of the development of water resources on environment in arid inland river basins of Hexi region, Northwestern China," Environmental Geology, vol. 50, pp. 793-801, Aug 2006.

[4] Z. T. Cong, D. W. Yang, B. Gao, H. B. Yang, and H. P. Hu, "Hydrological trend analysis in the Yellow River basin using a distributed hydrological model," Water Resources Research, vol. 45, p. 13, Feb 2009.

[5] L. Zhen and J. K. Routray, "Groundwater resource use practices and implications for sustainable agricultural development in the North China Plain: A case study in Ningjin County of Shandong Province, PR China," International Journal of Water Resources Development, vol. 18, pp. 581-593, Dec 2002.

[6] X. Y. Zhang, D. Pei, and C. S. Hu, "Conserving groundwater for irrigation in the North China Plain," Irrigation Science, vol. 21, pp. 159-166, 2003.

[7] Y. Q. Xu, X. G. Mo, Y. L. Cai, and X. B. Li, "Analysis on groundwater table drawdown by land use and the quest for sustainable water use in the Hebei Plain in China," Agricultural Water Management, vol. 75, pp. 38-53, Jul 2005.

[8] G. Wang and G. Cheng, "The characteristics of water resources and the changes of the hydrological process and environment in the arid zone of northwest China," Environmental Geology, vol. 39, pp. 783-790, May 2000.

[9] S. Z. Kang, X. L. Su, L. Tong, P. Z. Shi, X. Y. Yang, Y. K. Abe, T. S. Du, Q. L. Shen, and J. H. Zhang, "The impacts of human activities on the water-land environment of the Shiyang River basin, an arid region in northwest China," Hydrological Sciences Journal-Journal Des Sciences Hydrologiques, vol. 49, pp. 413-427, Jun 2004.

[10] L. S. Pereira, J. M. Goncalves, B. Dong, Z. Mao, and S. X. Fang, "Assessing basin irrigation and scheduling strategies for saving irrigation water and controlling salinity in the upper Yellow River Basin, China," Agricultural Water Management, vol. 93, pp. 109-122, Nov 2007.

[11] P. Doll and S. Siebert, "Global modeling of irrigation water requirements," Water Resources Research, vol. 38, Apr 2002.

[12] A. Thomas, "Agricultural irrigation demand under present and future climate scenarios in China," Global and Planetary Change, vol. 60, pp. 306-326, Feb 2008.

[13] Y. M. Yang, Y. H. Yang, J. P. Moiwo, and Y. K. Hu, "Estimation of irrigation requirement for sustainable water resources reallocation in North China," Agricultural Water Management, vol. 97, pp. 1711-1721, Nov 2010.

[14] Y. B. Wang, P. T. Wu, X. N. Zhao, and J. L. Li, "Development tendency of agricultural water structure in China," Chinese Journal of Eco-Agriculture, vol. 18, pp. 399-404, 2010. (In Chinese)

[15] P. T. Wu, J. M. Jin, and X. N. Zhao, "Impact of climate change and irrigation technology advancement on agricultural water use in China," Climatic Change, vol. 100, pp. 797-805, Jun 2010.

[16] Y. Liu, L. Wang, G. H. Ni, and Z. T. Cong, "Spatial distribution characteristics of irrigation water requirement for main crops in China.," Transactions of the CSAE, vol. 25, pp. 6-12, 2009. (In Chinese)

[17] X. Zhu, S. Liang, and Y. Pan, "Mapping Irrigated Areas in China from Remote Sensing and Statistical Data," Transactions on Geoscience and Remote Sensing, 2012.