ABSTRACT The need for both global and regional assessment of water consumption has been increasing. This paper aimed to understand industrial water consumption both globally and regionally, and create the 1º*1º global map. From the analysis of data in Japan and China, it was determined that industrial water consumption correlated well with urban area which can obtain from GIS data. Based on this knowledge, industrial water consumption was distributed to the 1º*1º global map, and this calculation was named urban area model
Journal of Water and Environment Technology, Vol. 6, No.2, 2008 - 85 - Attempt to Establish an Industrial Water Consumption Distribution Model Yurina OTAK I * , Masahiro OTAKI ** and Tomoko YAMADA ** * Center for Research and Development Higher Education, University of Tokyo, 7-3-1, Hongo, Bunkyo-ku, Tokyo, 113-0033, Japan yurina00@nifty.com ** Graduate School of Humanities and Science, Ochanomizu University, 2-2-1 Otsuka, Bunkyo-ku, Tokyo 112-8610, Japan ABSTRACT The need for both global and regional assessment of water consumption has been increasing. This paper aimed to understand industrial water consumption both globally and regionally, and create the 1º*1º global map. From the analysis of data in Japan and China, it was determined that industrial water consumption correlated well with urban area which can obtain from GIS data. Based on this knowledge, industrial water consumption was distributed to the 1º*1º global map, and this calculation was named urban area model. Keywords: industrial water consumption, global map, water consumption distribution model INTRODUCTION In the year 2000, water resources were consumed worldwide for agricultural use (69%), for industrial use (21%), and for domestic use (10%) (AQUASTAT). Water in industry has been used for boilers, raw materials, washing, cooling, thermal modulation, etc. In addition to these traditional uses, new demands of high quality water for leading edge industries, such as IC (Integrated Circuit) and fine chemicals, are currently developing. Since the 1980s, industrial water consumption in many developed countries has not increased and has even decreased (Shiklomanov and Rodda, 2003). Conversely, industrialization in developing countries has been currently progressing. Therefore, industrial water consumption in the world will continue to increase hereafter. Although total industrial water consumption in a country may be understood, its distribution within the country is less well known. Water-rich and water-poor areas may exist in one country. To develop water supply plans suited for water resources situations, water consumption should be understood at a more detailed level than at the scale of the entire country. The goal of this paper is to analyze the factors which affect industrial water consumption, and to develop an industrial water consumption distribution model which can be applied globally. Currently, the importance of estimating future water stress on a global scale has been recognized and global analysis of water scarcity has been performed on a country basis, river basin basis, and for 0.5°*0.5° or 1°*1° grids for the entire globe (Alcamo et al., 2003). In case of the grid model, industrial water consumption is distributed to each grid cell in proportion to population. However, it is not considered reasonable that industrial activity is proportional to population. Thus, a more effective indicator for distribution is required, and our new distribution model will be effective for this purpose. Address correspondence to Yurina OTAKI, Center for Research and Development Higher Education, The University of Tokyo, Email: yurina00@nifty.com Received July 1, 2008, Accepted November 27, 2008. Journal of Water and Environment Technology, Vol. 6, No.2, 2008 - 86 - The nature of industrial water consumption depends, to a large extent, on the type of water supply scheme being used (Shiklomanov and Rodda, 2003). The scheme, based on how much water can be circulated and reused and how much sea water can be used, depends on conditions of location, levels of technology, etc. In this paper, we focused on the total amount of industrial water consumption, and considered that the scheme should be determined according to the location of the site. ANALYSIS OF THE FACTORS WHICH AFFECT INDUSTRIAL WATER CONSUMPTION The analysis was initiated using detailed data from 2002 obtained from prefectures in Japan (Ministry of Economics, Trade and Industry, 2002). Factors which were believed to affect industrial water consumption, including industrial area, number of business establishments, number of employees, and shipment values, in each of 47 prefectures in Japan in 2002, were examined. According to stepwise regression analysis, the factor strongly associated with industrial water consumption was industrial area (r=0.879) (Table 1, Fig. 1). To examine whether the determinant of industrial water consumption differed among industrial sectors, water consumption by top five industries, chemical, iron and steel, petroleum and coal products, pulp and paper, and transport machinery, was also analyzed using stepwise regression. As a result, industrial area was found to be the determinant in all kinds of industries tested. Thus, it was determined that industrial area is the primary determinant of industrial water consumption. Every sector demonstrated strong correlation with industrial water consumption. The regression coefficient differed by sector (Table 2). Therefore, it is indicated that industrial water consumption can be calculated from industrial area. ATTEMPT TO ESTABLISH AN INDUSTRIAL WATER CONSUMPTION DISTRIBUTION MODEL FOR JAPAN Based on the analysis of factors which affect industrial water consumption using Japanese data, establishment of a universal industrial water consumption distribution model was examined. Although it was determined that industrial water consumption was related to industrial area in the Japanese study, data sufficient to determine the industrial area does not exist for all countries. Thus, it was considered that a measure of urban area might be an appropriate alternative for the industrial area, because data for urban areas are more readily available in every country by GIS datasets. To examine the interchangeability of data for urban areas for data for industrial areas, we checked for correlation in cases of mesh sizes of 1km, 10km, 80km, 160km, and 320km using Japanese mesh data (digital national land information) (Ministry of Land, Infrastructure and Transport). Urban area could be used accurately instead of industrial area above a 100km mesh size (Fig. 2). For confirmation, the relationship between urban area and industrial water consumption by mesh size was plotted in Fig. 3. Thus, the correlation coefficient for the 100km square mesh Journal of Water and Environment Technology, Vol. 6, No.2, 2008 - 87 - size was determined to be approximately 0.8. Therefore, urban area was strongly associated with industrial water consumption in the 100km mesh size. Table 1 Coefficient of correlation with industrial water consumption Industrial Area 0.879 Number of business establioshments 0.550 Number of employees 0.645 Shipment Values 0.730 Figure 1 - Industrial area and industrial water consumption Table 2 - Correlation coefficient and regression coefficient of industrial water consumption and industrial area R a Chemical Iron and Steel Petroleum and Coal Products Pulp and Paper Transport Machinery 0.866 0.924 0.930 0.958 0.940 0.619 0.361 0.369 0.403 0.111 Total 0.827 0.185 R: Correlation coefficient a: Regression coefficient Industrial water consumption (1,000 m 3 /day) Industrial area (1,000 m 2) ) 0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 16,000 0 30,000 60,000 90,000 120,000 Journal of Water and Environment Technology, Vol. 6, No.2, 2008 - 88 - 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1 10 100 1000 Correlation coefficient Mesh size (km sq.) Figure 2 - Correlation between industrial area and urban area by mesh size 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1 10 100 1000 Correlation coefficient Mesh size (km sq.) Figure 3 - Correlation coefficient between the urban area and the industrial water consumption and mesh size ESTABLISHMENT OF WORLD WIDE APPLICATION FOR THE INDUSTRIAL WATER CONSUMPTION DISTRIBUTION MODEL Applicability of the estimation of industrial water consumption using urban area to countries other than Japan was examined. Data for China, which is now the largest user of industrial water in the world, and will continue to increase its use with increased economic growth, was used for validation. Industrial water consumption in China accounts for 21% of total world consumption of water for industrial consumptions (World Resources Institute) —a percentage that is likely to increase. Data for industrial water consumption and urban area were collected by 31 administrative districts (22 provinces, 5 autonomous territories, and 4 government-ruled municipalities; Beijing, Tianjin, Shanghai, and Chongqing) in 2001 (National Bureau of Statistics of China). It was uncertain that the definition of urban area in Statistical Yearbook of China was same with that in Japanese GIS data, because there was no detailed information about Chinese data. However, it was supposed that there was not so much difference. The area per administrative district in China ranged from 250 to 800 square km. This scale is larger Journal of Water and Environment Technology, Vol. 6, No.2, 2008 - 89 - than the standard of 100km square mesh, which was established for the correlation between industrial water consumption and urban area using Japanese data. Fig. 4 illustrates the strong correlation between industrial water consumption and urban area (r=0.8281). Thus, it was confirmed that current industrial water consumption is related to urban area in China, also. Industrial water consumption (m 3 /day) Urban area (km 2 ) 0 1,000,000 2,000,000 3,000,000 4,000,000 5,000,000 6,000,000 0 500 1,000 1,500 2,000 Fig. 4 Industrial water consumption and urban area (China) CREATION OF GLOBAL MAP FOR INDUSTRIAL WATER CONSUMPTION DISTRIBUTION The data globally available is Q t (total industrial water consumption of each country), U t (total urban area of each country), and U i (urban area of mesh i). And it was examined that Q i was proportional to U i in grids greater than 100km square mesh size: Q i = Q t * (U i /U t ) Q t = ∑Q i In accordance with this calculation, industrial water consumption was distributed on the global grid map using Lambert projection (Fig. 5). Urban area data were supplied from the1º*1º ratio of urban coverage in the year 2000 (Hall et al., 2006: Loveland and Belward, 1997). The 1º*1º grid is nearly equivalent to a 100km square mesh size. This calculation model was termed the urban area base model. Meanwhile, a second global grid map, in which the total world industrial water consumption in 1995 (World Resource Institute) was distributed to each grid proportional to population (Oki et al., 2001) (Fig. 6). This model was termed the population base model. The difference between the urban area model and the population model was investigated. Fig. 7 illustrates the difference between the value of the urban area model and the value of the population model. Only 4,000 grid cells among total 129,600 cells included urban area. The grid, which is believed to have population, but not urban area, such as cultivated area or mountain-ringed area, is over-estimated in the case of the population model. Journal of Water and Environment Technology, Vol. 6, No.2, 2008 - 90 - Generally, population data at the 1º*1º mesh level does not exist because population data is collected by the country and the municipality. Conversely, as the 1º*1º grid data of the urban area was obtained from universal satellite observation, it was effective for datasets composing the gridded map. In addition, Vassolo and Doll studied that the model efficiency using population was low (Vassolo and Doll, 2005). Thus, our urban area model was distinct because correlation between urban area and industrial water consumption was verified. Fig. 5 Global grid map of industrial water consumption (urban area model) Fig. 6 Global grid map of industrial water consumption (population model) Fig. 7 Difference between urban model and population model （Mm 3 /year） （Mm 3 /year） （Mm 3 /year） Journal of Water and Environment Technology, Vol. 6, No.2, 2008 - 91 - CONCLUSION From analysis of data from Japan, it was determined that industrial water consumption and industrial area were correlated. Substitution of the urban area, for which global grid data exist, for industrial area, was determined to be successful by the analysis of data from Japan and China. Using these results, the 1º*1º global map for industrial water consumption distribution was created. The new model was established based on the validation of the correlation between industrial water consumption and urban area. This study established an important basis for applying data to future water supply plans tailored to the amount of water resources. In this study, the effectiveness of using data representing urban area was tested for data from Japan and China. Examination of data from other geographical areas is an issue for future research. ACKNOWLEDGEMENTS The authors are grateful to Naota Hansasaki of National Institute for Environmental Studies for his cooperation. This study was supported by CREST (Core Research for Evolutional Science and Technology) of the Japan Science and Technology Corporation. REFERENCES Alcamo, J., Doll, P., Henrichs, T., Kaspar, F., Lehner, B., Rosch, T. & Siebert, S. (2003), Hydrolog. Sci. J., 48(3), 317-337. AQUASTAT, http://www.fao.org/nr/water/aquastat/main/index.stm Hall, F. G., Brown de Colstoun, E., Collatz, G. J., Landis, D., Dirmeyer, P., Betts, A., Huffman, G. J., Bounoua, L. and Meeson, B. (2006), ISLSCP Initiative II global data sets: Surface boundary conditions and atmospheric forcings for land-atmosphere studies, J. Geophys. Res., 111, D22S01. Loveland, T. R., and Belward, A. S. (1997), The IGBP-DIS global 1 km land cover data set, DISCover: first results, Int. J. Remote Sens., 18(15), 3289 -3295 Ministry of Economy, Trade and Industry (2002), Census of Manufactures, Industrial Site and Water. Ministry of Land, Infrastructure, and Transport, Digital National Land Information, http://w3land.mlit.go.jp/WebGIS/ National Bureau of Statistics of China, Statistical Yearbook of China 2001. Oki, T., Agata, Y., Kanae, S., Saruhashi, T., Yang, D. and Mushiake, K. (2001), Global assessment of current water resources using total runoff integrating pathways, Hydrolog. Sci. J., 46(6), 983-996. Shikmanov, I.A. and Rodda, J.C (2003), World Water Resources at the beginning of the 21st Century, Cambridge University Press. Vassolo, S. and Doll, P. (2005), Global-scale gridded estimates of thermoelectric power and manufacturing water use, Water Resour. Res., 41, W04010. World Resources Institute, http://www.wri.org/ . paper is to analyze the factors which affect industrial water consumption, and to develop an industrial water consumption distribution model which can be. INDUSTRIAL WATER CONSUMPTION DISTRIBUTION MODEL FOR JAPAN Based on the analysis of factors which affect industrial water consumption using Japanese data, establishment