Determination of virtual water content of rice and spatial ......L. J. Zhang et al.: Determination of virtual water content of rice 2105 economic system (Leontief, 1941), and is adopted

Hydrol. Earth Syst. Sci., 18, 2103–2111, 2014www.hydrol-earth-syst-sci.net/18/2103/2014/doi:10.5194/hess-18-2103-2014© Author(s) 2014. CC Attribution 3.0 License.

Determination of virtual water content of rice and spatialcharacteristics analysis in ChinaL. J. Zhang, X. A. Yin, Y. Zhi, and Z. F. Yang

State Key Laboratory of Water Environmental Simulation, School of Environment, Beijing Normal University, Beijing, China

Correspondence to:X. A. Yin ([email protected])

Received: 23 December 2013 – Published in Hydrol. Earth Syst. Sci. Discuss.: 22 January 2014Revised: – Accepted: 2 May 2014 – Published: 6 June 2014

Abstract. China is a water-stressed country, and agricultureconsumes the bulk of its water resources. Assessing the vir-tual water content (VWC) of crops is one important wayto develop efficient water management measures to allevi-ate water resource conflicts among different sectors. In thisresearch, the VWC of rice, a major crop in China, is takenas the research object. China covers a vast land area, and theVWC of rice varies widely between different regions. TheVWC of rice in China is assessed and the spatial character-istics are also analysed. The total VWC is the total volumeof freshwater both consumed and affected by pollution dur-ing the crop production process, including both direct andindirect water use. Prior calculation frameworks of the VWCof crops did not contain all of the virtual water content ofcrops. In addition to the calculation of green, blue and greywater – the direct water in VWC – the indirect water use ofrice was also calculated, using an input–output model. Thepercentages of direct green, blue, grey and indirect water inthe total VWC of rice in China were found to be 43.8, 28.2,27.6, and 0.4 %. The total VWC of rice generally showed aroughly three-tiered distribution, and decreased from south-east to northwest. The higher values of direct green waterusage were mainly concentrated in Southeast and SouthwestChina, while the values were relatively low in NorthwestChina and Inner Mongolia. The higher direct blue water val-ues were mainly concentrated in the eastern and southerncoastal regions and Northwest China, and low values weremainly concentrated in Southwest China. Grey water val-ues were relatively high in Shanxi and Guangxi provincesand low in Northeast and Northwest China. The regions withhigh values for indirect water were randomly distributed butthe regions with low values were mainly concentrated inNorthwest and Southwest China. For the regions with rela-

tively high total VWC the high values of blue water madethe largest contribution, although for the country as a wholethe direct green water is the most important contributor.

1 Introduction

The termvirtual water was first proposed by Allan (1994)and defined as the water embodied in traded products. Laterthe concept of virtual water was modified by Hoekstra andChapagain (2007) to indicate the water input required to gen-erate a product or service. The virtual water content of aproduct is the freshwater embodied in the product, not in thereal sense, but in the virtual sense. It refers to the volumeof the freshwater both consumed and affected by pollutionin producing the product, measured over its full productionchain (Hoekstra et al., 2011). In water-stressed regions, lim-ited water resources should be used efficiently by not allo-cating the majority of resources to the production of water-intensive products, but being made available for other eco-nomic purposes that can contribute more to the regional valueadded by consuming less water (Allan, 2002; Chapagain andHoekstra, 2008). Assessing the virtual water content (VWC)of products is the basis for developing such water resourcemanagement practices.

Hoekstra and Hung (2002) estimated the VWC of cropsfor many countries of the world. In their research, thecrop VWC was determined by estimating the accumulatedcrop evapotranspiration over the growing period, and theVWC was not divided into subtypes. To better understandthe VWC of crops, many scientists now divide the VWCinto subtypes and calculate them separately. The calculationof green (effective precipitation) and blue water (irrigation

Published by Copernicus Publications on behalf of the European Geosciences Union.

2104 L. J. Zhang et al.: Determination of virtual water content of rice

water withdrawn from ground or surface water) was ini-tially proposed. Research has been performed at global, na-tional, provincial, and river basin scales. For example, Rostet al. (2008) made a global estimate of agricultural green andblue water consumption. Siebert and Döll (2010) computedthe green and blue VWC of crops at a global scale, and foundthat the global average VWC of cereal crops was 1109 m3 t−1

of green water and 291 m3 t−1 of blue water. Scientists havealso added grey water to the VWC, defined as freshwater thatis required to assimilate the pollutant load based on naturalbackground concentrations and existing ambient water qual-ity standards. Chapagain et al. (2006) first calculated the greywater in the VWC of crops, finding that the global VWC ofrice was an average of 1325 m3 t−1 and, further, that greywater occupied about 8 % of the total VWC (Chapagain andHoekstra, 2011). Mekonnen and Hoekstra (2011a) quantifiedgrey VWC of global crop productions for the period 1996–2005, and found that green, blue and grey water accountedfor 78, 12 and 10 % in the total VWC of crops.

China is one of the world’s 13 most water-poor countries(Yu et al., 2006). In 2012, per capita use of water resourcesin China was only 2100 m3, less than 30 % of the world percapita consumption. Agriculture is the largest water user inChina, accounting for nearly 70 % of total water withdrawals(Ministry of Water Resources, 2012). Studies on the VWC ofcrops in China are relatively limited. Liu et al. (2007), fromthe perspective of crop water productivity, estimated the vir-tual water use of winter wheat. Sun et al. (2013a) used thecrop water requirement to calculate the China average VWCof wheat, maize and rice, and found the proportions of greenand blue water to be 50.98 and 49.02 %, 76.27 and 23.73 %,and 61.90 and 38.10 %, respectively. The VWC of rice wasrelatively low in the eastern part of Northeast China, middle–lower reaches of the Yangtze River and the eastern part ofSouthwest China. In contrast, the high values of VWC forrice were located in the west of Inner Mongolia and south ofthe Xinjiang Uygur autonomous region. Sun et al. (2013b)estimated the VWC of crops as 3.91 m3 kg−1 in the Hetaoirrigation district of China. The percentage of blue waterwas relatively high (90.91 %), while the share of green waterwas small (9.09 %). However, these studies all ignored greywater.

Rice as a cereal grain is the most widely consumed sta-ple food for a large part of the population, especially in Asia.According to data from FAOSTAT, rice is the grain with thesecond-highest worldwide production, after maize. China isthe biggest rice-producing country in the world. In 2007, theplanting area of rice was the second largest in China (about29 million hectares), comprising 34 % of the total plantingarea of grain crops. Rice production is the largest grain pro-duction in China (around 186 million tonnes), accounting for41 % of the total grain (Ministry of Agriculture of the Peo-ple’s Republic of China, 2008). There is no research on thetotal VWC of rice and spatial distribution characteristics inChina at the provincial scale by actual total water use. China

covers a vast land area, and the VWC of rice varies widelybetween different regions. The VWC of rice in China is as-sessed and the spatial characteristics are also analysed.

Prior research on crop VWC determination provides agood framework for this work. However, prior calculationframeworks of VWC of crops have some defects. Some usethe water requirement of crops instead of actual water use,and others ignore the freshwater affected by pollution duringcrop production. Besides, in all previous calculation frame-works the indirect water use of crops was also ignored. Onthe basis of the previous frameworks, in addition to directwater use (blue, green, and grey water), we also considerindirect water (the sum of the virtual water of all productsconsumed in the process of rice planting). In this paper, wecalculate the total VWC of rice for 29 Chinese provinces,autonomous regions, and municipalities in 2007, includingboth direct and indirect water use. Because of the lack ofdata, Tibet is not taken into account; and no rice is plantedin Qinghai, so Qinghai is not taken into account either. Thespatial distribution characteristics of the VWC of rice are alsoanalysed.

2 Methodology and data

2.1 Methodology

To reflect water consumption during crop production, directand indirect water were taken into account. Due to lack ofdata, we cannot divide the VWCindirect into VWCindirect, green,VWCindirect, blue and VWCindirect, grey:

VWCtotal = VWCindirect+VWCdirect

= VWCindirect+VWCdirect, green

+ VWCdirect, blue+ VWCdirect, grey (1)

where VWCtotal is the total volume of freshwater both con-sumed and affected by pollution during the crop produc-tion process (m3 kg−1); VWCindirect is the freshwater bothconsumed and affected by pollution that can be associatedwith the production of the goods and services or the in-puts used during the crop production process (m3 kg−1);VWCdirect is the freshwater both consumed and affectedby pollution that is associated with direct water use duringthe crop production process (m3 kg−1); VWCdirect, green isthe precipitation consumed in the crop production process(m3 kg−1); VWCdirect, blue is the surface water or ground-water consumed in the crop production process (m3 kg−1);and VWCdirect, grey is the freshwater required to assimilatethe load of pollutants during the crop production process(m3 kg−1).

2.1.1 Indirect water of crops

The input–output model represents the monetary tradeof products and services among different sectors of an

Hydrol. Earth Syst. Sci., 18, 2103–2111, 2014 www.hydrol-earth-syst-sci.net/18/2103/2014/

L. J. Zhang et al.: Determination of virtual water content of rice 2105

economic system (Leontief, 1941), and is adopted to calcu-late the indirect virtual water of crops supplied by each eco-nomic sector. The calculations are as follows (Chen, 2000;Kanada, 2001; Zhao et al., 2009; Zhang et al., 2011).

1. Direct consumption coefficient matrix. The input–output table is used because it reflects the contact ofthe materials and technologies. This contact is reflectedthrough the direct consumption coefficient:

A =∣∣aij

∣∣ =∣∣xij/xi

∣∣ , (2)

whereA is the direct consumption coefficient matrix(n × n dimensional matrix) in the IO table;aij is thedirect consumption coefficient, which means the mon-etary volume of products of sectori directly consumedby sectorj when producing one unit product;xij is themonetary volume of products from sectorj consumedby sectori in its production process (RMB); andxi isthe output of sectori (RMB).

2. Complete consumption coefficient matrix. Comparedwith the direct consumption coefficient, the completeconsumption coefficient can more accurately measurethe direct and indirect costs (the sum of the direct andindirect costs is completely consumed) of products orservices from other sectors:

B =∣∣bij

∣∣ = (I − A)−1− I , (3)

whereB is the complete consumption coefficient ma-trix (n × n dimensional matrix) in the IO table;bij isthe complete consumption coefficient, which means themonetary volume of products of sectori directly and in-directly consumed by sectorj when producing one unitproduct; andI is a unit diagonal matrix.

3. Water use coefficient. To account for indirect water, itis necessary to compute the water use coefficient of dif-ferent sectors, which is the water needed to produce onemonetary unit (RMB):

DWCi = wi/xi, (4)

where DWCi is the direct water coefficient of sectori

(m3 RMB−1) andwi is the direct water consumption ofsectori (m3). DWCi is the amount of direct water intaketo produce one monetary unit of production.

4. Indirect water of agriculture. Indirect water consump-tion of agriculture is the amount of total water inputfrom other sectors.

VWaindirect =

n∑i=1

(DWCi × bia) × Ca, (5)

where VWaindirect is the indirect water consumption of

agriculture (m3); bia is the complete consumption co-efficient of sectori for agriculture; andCa is the totalconsumption of agriculture (RMB).

5. Indirect water consumption of a crop. VWCiindirect

(m3 kg−1) is calculated according to the proportion ofindirect water use of cropi in the total indirect waterconsumption of agriculture:

VWCiindirect =

VWaindirectαi

SAi × Y, (6)

where VWCiindirect is the indirect water consumption of

crop i (m3 kg−1); αi is the proportion of indirect wateruse of cropi in the total indirect water consumption.Because of lack of data, we assume that the plantingcost is proportional to the indirect water use. SAi is thesown area of cropi (ha); andY is the crop yield per unitarea (kg ha−1). Thusαi can be calculated as follows:

αi =PCi × SAi

n∑i=1

(PCi × SAi)

, (7)

where PCi is the planting cost of cropi per unit area(RMB ha−1).

2.1.2 Direct green water of crops

Direct green water use is the lesser of potential crop evapo-transpiration and effective precipitation. Effective precipita-tion is defined as the amount of precipitation that enters thesoil and will be available in the soil for crop growth (Sun etal., 2013b):

VWCdirect, green=10min(ETc,Pe)

Y, (8)

where ETc is the crop evapotranspiration during the grow-ing period (mm) andPe is the effective precipitation overthe crop growing period (mm), calculated by the CROPWATmodel using monthly climatic data (mm) (Clarke, 1998;FAO, 2003).

Crop evaporation during the growing period is calculatedas follows (Allen et al., 1998):

ETc = ETo × kc, (9)

wherekc is the crop coefficient, reflecting the differences inphysical and physiological factors between the actual andreference crops and ETo is the soil evaporation of the ref-erence underlying surface (mm d−1), calculated by the FAOPenman–Monteith formula (Allen et al., 1998).

2.1.3 Direct blue water of crops

The direct blue water of a crop is calculated using the irri-gation water consumption,Ic. The irrigation water consump-tion is the net artificial application of water use by crops,which does not include the irrigation water losses during thetransport process from the water sources to cropland and the

www.hydrol-earth-syst-sci.net/18/2103/2014/ Hydrol. Earth Syst. Sci., 18, 2103–2111, 2014


return flows of irrigation water. The direct blue water is cal-culated according to the proportion of irrigation water con-sumption of cropi in the total irrigation water consumptionof the irrigation district (Sun, 2013b):

VWCidirect, blue = I i

c/Y =WAβi

SAi × Y, (10)

whereI ic is the irrigation water consumption of cropi per

unit area (m3 ha−1), WA is the irrigation water consumptionof the irrigation district (m3) andβi is the proportion of ir-rigation water use of cropi in the total irrigation water con-sumption of the irrigation district.βi can be calculated asfollows:

βi =

(ETi

c − P ie

)× SAi

n∑i=1

[(ETi

c − P ie

)× SAi

] , (11)

where ETic is the crop evapotranspiration of cropi during thegrowing period (mm),P i

e is the effective precipitation overthe cropi growing period (mm) and SAi is the sown area ofcropi (ha).

2.1.4 Direct grey water of crops

In this study we quantify direct grey water related to nitro-gen use only. The direct grey water is calculated by multi-plying the fraction of nitrogen that leaches or runs off by thenitrogen application rate and dividing this by the differencebetween the maximum acceptable concentrations of nitrogenand the natural concentration of nitrogen in the receiving wa-ter body and by the actual crop yield (Mekonnen and Hoek-stra, 2011a). Because of lack of data, the natural nitrogenconcentrations were assumed to be 0. On average, 10 % ofthe applied nitrogen fertilizer is lost through leaching (Cha-pagain et al., 2006). The maximum value of nitrate in surfaceand ground water recommended by the United States Envi-ronmental Protection Agency is 10 mg L−1 (Chapagain et al.,2006):

VWCdirect, grey =Nc × 10%

10− Nn/Y, (12)

whereNc is the amount of nitrogen fertilizer consumptionper hectare (g ha−1) andNn is the natural concentration ofnitrogen in the receiving water body (mg L−1).

2.2 Data

The 2007 climate data for 29 regions, including monthlyaverage maximum temperature, monthly average minimumtemperature, relative humidity, wind speed, precipitation,and sunshine hours, are taken from the National ClimaticCentre (NCC) of the China Meteorological Administration(CMA). The agricultural data, including crop yield and sownarea, are taken from the China Agricultural Yearbook (Min-istry of Agriculture of the People’s Republic of China, 2008).

Figure 1. Indirect water of rice (m3 kg−1).

The average amount of nitrogen fertilizer of rice per unit areais taken from Li et al. (2010) and Zhang et al. (2008, 2009).The irrigation water consumption for the 29 regions is takenfrom the Water Resources Bulletins (2007) of the 29 regions.Water data for various sectors in the regions are from the Sta-tistical Yearbooks (2008) of the 29 regions. The IO data forthe 29 regions come from the official IO tables (2007) of the29 regions.

3 Results

3.1 Indirect water of rice

The VWCindirect of rice varied between 0.001 and0.010 m3 kg−1. The average VWCindirect was 0.004 m3 kg−1.The regions with the same line had the same distributioncharacteristic of VWCindirect of rice in Fig. 1. The input toagriculture in relatively underdeveloped regions was rela-tively small; relatively developed regions invest more moneyin agriculture. Northwest and Southwest China are relativelyunderdeveloped with relatively low VWCindirect of rice. Bei-jing, Guangdong, and Shanghai are the three most devel-oped regions and had the highest VWCindirect of rice in 2007(Fig. 1). We found that VWCindirect of rice is directly relatedto the degree of regional economic development.

Of all sectors, the sector of forestry, animal husbandry, andfishery contributed the most VWCindirect of rice, accountingfor 52.2 %. The electricity and heat sector contributed 21.6 %and the chemical industry sector contributed 9.6 %. The restof the sectors contributed less.



Figure 2. Direct green water of rice (m3 kg−1).

3.2 Direct green water of rice

The regional differences in VWCdirect, green for rice weresignificant, owing to differences in climatic conditions andcrop yields. The VWCdirect, green of rice for the 29 regionsin 2007 ranged from 0.10 to 0.90 m3 kg−1. The averageVWCdirect, greenof rice was 0.59 m3 kg−1. VWCdirect, greenofrice increased gradually from northern to southern regions(Fig. 2). The regional variability of VWCdirect, green of ricewas in accordance with the distribution of precipitation inChina. The regions with abundant precipitation usually havehigh VWCdirect, green of rice. Precipitation in southern re-gions of China is far greater than that in northern regions.Consequently, the VWCdirect, greenof rice in southern regionswould be higher than that in northern regions.

The regions with higher VWCdirect, greenvalues were con-centrated in Southeast China and Southwest China (Fig. 2).The high VWCdirect, green in these regions is a result of thehigh ratio between effective precipitation and rice yield. Theregions in Southeast China and Southwest China had rela-tively low rice yields, much lower than the national aver-age, and relatively high effective precipitation, more than400 mm. For example, the VWCdirect, greenof rice was morethan 0.80 m3 kg−1 in Hainan (Southeast China), Guangxi(Southwest China) and Yunnan (Southwest China). TheVWCdirect, green of rice in Guizhou (Southwest China) wasalso relatively high, at more than 0.80 m3 kg−1. This ismainly because the effective precipitation over the growingperiod of rice in this region was more than 590 mm, althoughrice yield in this region was higher than the national average.

The VWCdirect, greenof rice in Northwest China and InnerMongolia were relatively low (Fig. 2), because of high yieldsand low effective precipitation. Xinjiang (Northwest China)had the lowest VWCdirect, green (0.10 m3 kg−1) because the

Figure 3. Direct blue water of rice (m3 kg−1).

rice yield in Xinjiang was 1.32 times higher than the nationalaverage and the effective precipitation over the growing pe-riod was small, less than 90 mm.

3.3 Direct blue water of rice

The differences in blue water requirements, actual irriga-tion water consumption and rice yields resulted in signifi-cantly different VWCdirect, blue values between regions. TheVWCdirect, blue values ranged from 0.07 to 1.65 m3 kg−1,and the average was 0.42 m3 kg−1. The regions with higherVWCdirect, blue of rice were mainly concentrated in the east-ern and southern coastal regions and in Northwest China(Fig. 3). The VWCdirect, blue of rice was high in the mu-nicipalities of Beijing, Tianjin and Shanghai, at more than0.83 m3 kg−1. These three municipalities have developedeconomies with relatively more developed agricultural irriga-tion systems, so irrigation water consumption was relativelylarger than that of other regions. Regions in Northwest Chinahad higher VWCdirect, blue values, perhaps because the effec-tive precipitation in these regions was limited, making it nec-essary to increase the irrigation water supply for crops.

The VWCdirect, blue values for rice in Southwest Chinawere relatively low (Fig. 3). The VWCdirect, blue values inChongqing, Guizhou, Sichuan and Yunnan were less than0.16 m3 kg−1 because in these four regions the effective pre-cipitation can almost meet the water requirements of rice,so the actual irrigation water consumption was limited. Jilin,Shandong and Henan were the other three regions with rel-atively low VWCdirect, blue of rice. Limited irrigation wa-ter consumption in the three regions might only meet lessthan 25 % of the irrigation requirement of rice. Therefore,the VWCdirect, blue of rice in Jilin, Shandong and Henan wasrelatively low.



Figure 4. Direct grey water of rice (m3 kg−1).

3.4 Direct grey water of rice

The VWCdirect, grey values of rice ranged from 0.21 to0.64 m3 kg−1 with an average 0.37 m3 kg−1. Regional dif-ferences in VWCdirect, grey for rice were insignificant(Fig. 4). Because nitrogen use is similar between regions,VWCdirect, grey mainly depends on rice yield. The rice yieldof regions in Northeast China and Northwest China were rel-atively high, making the VWCdirect, grey of rice in NortheastChina and Northwest China relatively low. The rice yield ofShandong, Henan and Chongqing was much higher than thenational average. That made the VWCdirect, grey of rice in thethree regions also relatively low. The highest VWCdirect, greyvalues for rice were in Shanxi and Guangxi, because the riceyields of Shanxi and Guangxi in 2007 were the two lowest ofall the regions studied.

3.5 Total VWC of rice

The VWCtotal values of rice for 29 regions are calcu-lated and shown in Fig. 5. There were large differencesin VWCtotal between regions, with values ranging from0.80 to 2.59 m3 kg−1. The average VWCtotal of rice was1.39 m3 kg−1. The VWCtotal values showed a roughly three-tiered distribution, gradually decreasing from the southeastto the northwest of China. The VWCtotal values in easterncoastal China, Southeast China, Beijing and Tianjin wererelatively high. The regions with lower VWCtotal valueswere mainly concentrated in Northeast China and NorthwestChina.

Our result is very different from the result of Sun etal. (2013a). Their calculation of the VWC of rice was rel-atively low in the eastern part of Northeast China, middle–lower reaches of the Yangtze River and the eastern part ofSouthwest China. In contrast, the high values of VWC for

Figure 5. Total VWC of rice (m3 kg−1).

rice were located in the west of Inner Mongolia and the southof the Xinjiang Uygur autonomous region. Their calcula-tion framework only considered the crop water requirement.However, our calculation framework considers the effectiveprecipitation, crop evapotranspiration, irrigation water con-sumption, freshwater that is required to assimilate the load ofpollutants, and the indirect water use. The difference of theVWCdirect, blue and the added VWCdirect, grey of rice causedthe large difference of spatial distribution characteristic be-tween actual VWCtotal of rice and the rice water requirement.Our result better describes the spatial distribution character-istic of actual water use of rice in China.

4 Discussion

4.1 Comparison of calculations of total VWC of rice bydifferent frameworks

Here, we compare four frameworks for total VWC determi-nations (Table 1). (1) In the crop water requirement (CWR)calculation framework, the total virtual water content ofcrops is divided into green and blue water. Green water useis the lesser of the potential crop evapotranspiration and theeffective precipitation. Blue water use is the irrigation waterrequirement, which is the potential crop evapotranspirationminus green water use (Siebert and Döll, 2010; Sun et al.,2013a). (2) In the green, blue and grey water (GBG) calcula-tion framework, the total virtual water content of crops is di-vided into green, blue and grey water. The calculation meth-ods for green and blue water are the same as in the CWRframework. Grey water is calculated by multiplying the frac-tion of nitrogen that leaches or runs off by the nitrogen ap-plication rate and dividing this by the difference betweenthe maximum acceptable concentration of nitrogen and the



Table 1.Comparison of VWC calculation frameworks.

VWC Method VWCdirect, green VWCdirect, blue VWCdirect, grey VWCindirect

Our method 10 min(ETC,Pe)/Y VWCdirect,blue= Ic/Y =WAαi

Ai/Y VWCdirect,grey=

Nc×10%10−Nn

/Y VWCindirect=VWa

blue,indirectβi

Ai/Y

Sun’s method 10 min(ETC,Pe)/Y VWCdirect,blue= Ic/Y =WAαi

Ai/Y — —

GBG method 10 min(ETC,Pe)/Y 10 max(0,ETc − Pe)/Y VWCdirect,grey=Nc×10%10−Nn

/Y —CWR method 10 min(ETC,Pe)/Y 10 max(0,ETc − Pe)/Y — —

Table 2.The volume (m3 kg−1) and proportion (%) of VWC of rice in Heilongjiang Province by four different frameworks.

VWC Method VWCdirect, green VWCdirect, blue VWCdirect, grey VWCindirect VWCtotal

Our method 0.45 (46.0) 0.29 (30.2) 0.22 (22.9) 0.01 (0.9) 0.97Sun’s method 0.45 (60.5) 0.29 (39.5) — — 0.74GBG method 0.45 (26.6) 1.01 (60.1) 0.22 (13.3) — 1.68CWR method 0.45 (30.7) 1.01 (69.3) — — 1.45

natural concentration of nitrogen in the receiving water body(Chapagain et al., 2006; Bulsink et al., 2010; Mekonnen andHoekstra, 2011b). (3) In Sun’s framework, the total virtualwater content of crops is also divided into green and bluewater. Blue water is calculated according to actual irrigationwater consumption (Sun et al., 2013b). (4) In our framework,the total virtual water content of crops is divided into directgreen, direct blue, direct grey and indirect water. The cal-culations for green and grey water are the same as in theGBG framework and the calculation method for blue wa-ter is the same as in Sun’s framework. As described pre-viously, we add indirect water to the total VWC of crops.However, in our calculation framework of the VWC of crops,we made some assumptions and simplifications in the cal-culation of VWCindirect, VWCdirect, blue and VWCdirect, grey,which causes some uncertainty in the results. The uncertaintyof the results cannot be completely eliminated. We couldonly make better assumptions and simplifications and usemore accurate data to make our results more accurate.

The VWC of rice of Heilongjiang Province in 2007 isused as an example to compare results among the four dif-ferent frameworks. As shown in Table 2, VWCtotal calcu-lated by our framework is 0.97 m3 kg−1, which accounts for66.9, 57.7 and 133.8 % of the VWCtotal of rice under theCWR framework, GBG framework, and Sun’s framework,respectively. The VWCtotal calculated by our framework islower than the VWCtotal calculated by CWR framework andGBG framework, because the VWCdirect, blue is calculated bythe crop irrigation water requirement under CWR frameworkand GBG framework and calculated by the actual irrigationconsumption under our framework. According to the calcula-tion in our method, the actual irrigation consumption cannotmeet the irrigation water requirement of rice in Heilongjiang.The VWCdirect, blue value calculated from the actual irriga-tion consumption under our framework is much lower thanthe VWCdirect, blue values calculated under the CWR and

GBG frameworks. Adding the VWCdirect, grey value resultedin the VWCtotal calculated by our framework being largerthan the VWCtotal calculated by Sun’s framework. The con-tribution of VWCdirect, grey is very important, accounting for22.9 % of the VWCtotal of rice under our framework, and can-not be ignored. The contribution of VWCindirect as calculatedunder our framework is limited and has little effect on theVWCtotal of rice in Heilongjiang.

4.2 Analysis of VWC structures of rice in China

Regional differences in VWCtotal values for rice were sig-nificant, as were regional differences in the structure of theVWCtotal of rice. The average VWCtotal of rice in Chinawas 1.39 m3 kg−1. The percentages of VWCdirect, green,VWCdirect, blue, VWCdirect, grey and VWCindirect in theVWCtotal for rice in China were 43.8, 28.2, 27.6 and 0.4 %,respectively.

Generally speaking, the regions in China mostlyhad a higher VWCdirect, green of rice and a lowerVWCdirect, blue of rice. For example, in Chongqing thepercentage of VWCdirect, green was 67.6 % and the per-centage of VWCdirect, blue was 6.6 %; in Sichuan thepercentage of VWCdirect, green was 53.7 % and the per-centage of VWCdirect, blue was 12.4 %. Six regions wereexceptions, having lower VWCdirect, green of rice andhigher VWCdirect, blue of rice. In Beijing the percent-age of VWCdirect, green was 19.8 % and the percent-age of VWCdirect,blue was 63.2 %; in Tianjin the per-centage of VWCdirect, green was 14.7 % and the percent-age of VWCdirect, blue was 69.1 %; in Fujian the per-centage of VWCdirect, green was 29.9 % and the percent-age of VWCdirect, blue was 49.7 %; in Gansu the percent-age of VWCdirect, green was 33.5 % and the percentageof VWCdirect, blue was 37.9 %; in Ningxia the percent-age of VWCdirect, green was 24.9 % and the percentage of



VWCdirect, blue was 48.1 %; and in Xinjiang the percent-age of VWCdirect, green was 12.1 % and the percentage ofVWCdirect, blue was 61.6 %. There are two reasons for the dif-ferent situation of the six regions. Excessive consumption ofirrigation water caused a lower percentage of VWCdirect, greenof rice and a higher percentage of VWCdirect, blue of ricein Beijing, Tianjin and Fujian. Limited precipitation causeda lower percentage of VWCdirect, green of rice and a higherpercentage of VWCdirect, blue of rice in Xinjiang, Ningxiaand Gansu. For the country as a whole, the proportionof VWCdirect, green of rice (43.8 %) was larger than theVWCdirect, blue of rice (28.2 %). Rice growth mainly dependson the VWCdirect, greenin China.

Regional differences in VWCdirect, grey values were in-significant. This shows that the grey water in all regions playsan important role in VWCtotal. Agricultural pollution is animportant issue in every region that cannot be ignored. Be-cause the direct grey water estimated only considers chem-ical fertilizer pollution, and not the effect of pesticides andherbicides on water quality, the result of this estimation is aconservative estimate.

The values of VWCindirect in the VWCtotal of rice forthe 29 regions were comparatively low, ranging from 0.001to 0.010 m3 kg−1. The average VWCindirect of rice was0.004 m3 kg−1. VWCindirect of rice is related to the degree ofregional economic development. The region with the high-est VWCindirect value was Beijing and the region with thelowest value was Ningxia. Because of the small contributionof VWCindirect, VWCindirect may not be considered in futureresearch on the VWC of rice. However, the VWCindirect isexpected to be higher in some agricultural products, such aspotatoes, cotton and fruits, so VWCindirect should be includedin the VWCtotal. For example, the proportion of VWCindirectin the VWCtotal of strawberry in 27 regions of China in2007 ranged from 0.8 to 38.0 %, with an average of 10.8 %.The proportion of VWCindirect of strawberry is higher in theVWCtotal. Therefore we cannot ignore the VWCindirect in thecalculation of the VWCtotal of some crops.

Assessing the VWC of crops is important to provide thebasis for agricultural water resources management, and helpto improve the efficiency of agricultural water use. Overallpressure on water resources might be relieved by locatingwater-intensive production processes in regions where wateris abundant and where it requires less VWC of product.

5 Conclusions

Faced with increasingly severe pressure on water resources,virtual water theory provides a feasible solution to improveglobal water use efficiency. Research on the VWC of cropscan provide the basis for agricultural water resources man-agement and help to improve the efficiency of agriculturalwater use. Rice is the most important food crop in China andalso one of the largest water consumers, so it is important

to study the total VWC of rice in China. Previous researchconstructed calculation frameworks using direct green water,direct blue water and direct grey water of crops. Building onthat previous research, we also considered the indirect waterof VWC. In this paper, we calculated the virtual water con-tent of rice for 29 regions of China in 2007. The followingconclusions were reached.

1. Analysis showed that the total VWC of rice in Chinadecreased gradually from southeast to northwest.

2. The regions with high indirect water were randomly dis-tributed. The indirect VWC of rice in Northwest Chinaand Southwest China was relatively low.

3. The regions with higher direct green water were mainlyconcentrated in Southeast China and Southwest China.The direct green water of rice for Northwest China andInner Mongolia was relatively low. The regions withhigher direct blue water of rice were mainly concen-trated in the eastern and southern coastal regions ofChina and in Northwest China. The direct blue water ofrice of Southwest China was relatively low. In the coun-try as a whole, the percentage of direct green water ofrice was far above that of direct blue water. Therefore,rice growth is mainly dependent on direct green waterin China.

4. The direct grey water of rice in Northeast China andNorthwest China was relatively low. But in all regions,grey water occupies a very important position in totalVWC.

Acknowledgements.We thank the International Science & Tech-nology Cooperation Programme of China (No. 2011DFA72420),the National Science Foundation for Innovative Research Group(No. 51121003), the National Basic Research Programme of China(No. 2010CB951104), and the Fundamental Research Funds forthe Central Universities for their financial support.

Edited by: Y. Cai

References

Allan, J. A.: Overall perspectives on countries and regions, in: Waterin Arab World: Perspectives and Progress, edited by: Rogers, P.and Lydon, P., Harvard University Press, Cambridge, UK, 65–100, 1994.

Allan, J. A.: The Middle East Water Question: Hydropolitics andthe Global Economy, Tauris Publishers, London, UK, 2002.

Allen, R. G., Pereira, L. S., Raes, D., and Smith, M.: Crop evap-otranspiration – guidelines for computing crop water require-ments, FAO Irrigationand Drainage Paper 56, FAO, Rome, 1998.

Bulsink, F., Hoekstra, A. Y., and Booij, M. J.: The water footprintof Indonesian provinces related to the consumption of crop prod-ucts, Hydrol. Earth. Syst. Sci., 14, 119–128, doi:10.5194/hess-14-119-2010, 2010.


http://dx.doi.org/10.5194/hess-14-119-2010



Chapagain, A. K. and Hoekstra, A. Y.: The global component offresh water demand and supply: an assessment of virtual waterflows between nations as a result of trade in agricultural and in-dustrial products, Water. Int., 33, 19–32, 2008.

Chapagain, A. K. and Hoekstra, A. Y.: The blue, green and grey wa-ter footprint of rice from production and consumption perspec-tives, Ecol. Econ., 70, 749–758, 2011.

Chapagain, A. K., Hoekstra, A. Y., Savenije, H. H. G., and Gautam,R.: The water footprint of cotton consumption: an assessment ofthe impact of worldwide consumption of cotton products on thewater resources in the cotton producing countries, Ecol. Econ.,60, 186–203, 2006.

Chen, X. K.: Shanxi water resource input-occupancy-output tableand its application in Shanxi Province of China, in: The 13thInternational Conference on Input-output Techniques, Macerata,Italy, 21–25 August 2000, 2-11, 2000.

China Meteorological Administration: China Meteorological DataSharing Service System, Beijing, China, available at:http://cdc.cma.gov.cn(last access: 6 September 2011), 2011.

Clarke, D.: CROPWAT for Windows: User Guide, University ofSouthampton, UK, 1998.

FAO (Food and Agriculture Organization): Land and Water De-velopment Division, CROPWAT model, Food and Agricul-ture Organization, Rome, Italy, available at:http://www.fao.org/nr/water/infores_databases_cropwat.html(last access: 9 June2011), 2003.

Hoekstra, A. Y. and Chapagain, A. K.: The water footprints of Mo-rocco and the Netherlands: global water use as a result of do-mestic consumption of agricultural commodities, Ecol. Econ.,64, 143–151, 2007.

Hoekstra, A. Y. and Hung, P. Q.: Virtual water trade: A quantifica-tion of virtual water flows between nations in relation to interna-tional crop trade. Value of Water Research Report Series No. 11,UNESCO-IHE, Delft, The Netherlands, 2002.

Hoekstra, A. Y., Chapagain, A. K., Aldaya, M. M., and Mekon-nen, M. M.: The Water Footprint Assessment Manual: Settingthe Global Standard, Earthscan, London, UK, 2011.

Kanada, N.: Land resources and international trade, Taga Shuppan,Tokyo, 2001.

Leontief, W.: The Structure of the American Economy, Oxford Uni-versity Press, Oxford, UK, 1941.

Li, H. L., Zhang, W. F., Zhang, F. S., Du, F., and Li, L. K.: Chemicalfertilizer use and efficiency change of main grain crops in China,J. Plant Nutrit. Fertiliz., 16, 1136–1143, 2010 (in Chinese).

Liu, J. G., Wiberg, D., Zehnder, A. J. B., and Yang, H.: Modellingthe role of irrigation in winter wheat yield, crop water productiv-ity, and production in China, Irrig. Sci., 26, 21–33, 2007

Mekonnen, M. M. and Hoekstra, A. Y.: The green, blue and greywater footprint of crops and derived crop products, Hydrol.Earth. Syst. Sci., 15, 1577–1600, doi:10.5194/hess-15-1577-2011, 2011a.

Mekonnen, M. M. and Hoekstra, A. Y.: National water footprintaccounts: The green, blue and grey water footprint of productionand consumption, Value of Water Research Report Series No. 50,UNESCO-IHE, Delft, Netherlands, 2011b.

Ministry of Agriculture of the People’s Republic of China: Chineseagricultural statistical data, Ministry of Agriculture of the Peo-ple’s Republic of China, Beijing, 2008.

Ministry of Water Resources: China Water Resources Bulletin2011, China Water & Power Press, Beijing, 2012.

Rost, S., Gerten, D., Bondeau, A., Lucht, W., Rohwer, J., andSchaphoff, S.: Agricultural green and blue water consumptionand its influence on the global water system, Water. Resour. Res.,44, W09405, doi:10.1029/2007WR006331, 2008.

Siebert, S. and Döll, P.: Quantifying blue and green virtual water-contents in global crop production as well as potential productionlosses without irrigation, J. Hydrol., 384, 198–207, 2010.

Sun, S. K., Wu, P. T., Wang, Y. B., and Zhao, X. N.: The virtual wa-ter content of major grain crops and virtual water flows betweenregions in China, J. Sci. Food. Agric., 93, 1427–1437, 2013a.

Sun, S. K., Wu, P. T., Wang, Y. B., Zhao, X. N., Liu, J., and Zhang,X. H.: The impacts of interannual climate variability and agricul-tural inputs on water footprint of crop production in an irrigationdistrict of China, Sci. Total Environ., 444, 498–507, 2013b.

Yu, P. P., Zhang, J. Z., and Lin, C. G.: Agricultural developmentstrategy of China in the early 21st century, Resour. Environ. Dev.,7, 21–27, 2006 (in Chinese).

Zhang, S. D., Zhang, W. F., Wang, J. Q.: Character of Fertilizer Con-sumption and Supply- Demand and Strategy for Management inMiddle and Lower Reaches of Yangtz River of China, Res. Agr.Modern., 29, 100–103, 2008 (in Chinese).

Zhang, S. D., Zhang, W. F., and Ma, L.: Study on the change offertilizer consumption structure of main grain crop in Hebei, Jilinand Sichuan of China, Phosphate Compound Fertilizer, 24, 89–91, 2009 (in Chinese).

Zhang, Z. Y., Yang, H., and Shi, M. J.: Analyses of water foot-print of Beijing in an interregional input-output framework, Ecol.Econ., 70, 2494–2502, 2011.

Zhao, X., Chen, B., and Yang, Z. F.: National water footprint in anIO framework: A case study of China 2002, Ecol. Model., 220,245–253, 2009.


http://cdc.cma.gov.cn

http://cdc.cma.gov.cn

http://www.fao.org/nr/water/infores_databases_cropwat.html

http://www.fao.org/nr/water/infores_databases_cropwat.html



http://dx.doi.org/10.1029/2007WR006331

Determination of virtual water content of rice and spatial ......L. J. Zhang et al.: Determination of virtual water content of rice 2105 economic system (Leontief, 1941), and is adopted

Documents