中文摘要：

基于我国31个省市自治区1997年～2007年水资源与社会经济的面板数据，利用改进的数据包络分析法（DEA）计算出各地区在不同时期的水资源利用相对效率，运用探索性空间数据分析法（ESDA）对中国水资源利用相对效率的时空差异变化特征与规律进行了初步探索。研究结果表明：①中国水资源利用相对效率在时间上呈上升趋势，发达地区用水效率最高，欠发达地区用水效率最低；②中国水资源利用相对效率在空间上存在显著的差异，11年内经历了先缩小后变大的过程，到2007年差异总体上在扩大，水资源利用水平相似地区之间差异在缩小；③水资源利用相对效率的差异是区域经济社会发展的普遍现象，它的存在与演变是区域经济和社会发展等诸多方面直接或间接的影响和反映。

英文摘要：

The contradiction between water resources supplement and demand has become more and more acute in China. In the context of water resources shortage, water resources utilization efficiency has become a focus in water sciences, but it is very difficult to appropriately measure water resources utilization efficiency. Water resources utilization relative efficiency （WURE） is considered to be an effective tool to measure water resources utilization efficiency. Based on the panel dataset of water utilization and social economy from 1997 to 2007 in China, the authors selected indexes of domestic water usage, labor, and fixed asset investment as inputs and gross domestic product （GDP） as output. The revised data envelopment analysis model （DEA）, which was developed to analyze efficiency and productivity based on an input-output production function, was utilized to calculate the water relative utilization efficiency for different regions and periods in China. The exploratory spatial data analysis （ESDA） was utilized to investigate the characteristics of the variations in spatial-temporal differences of WURE. Results indicated that： ① the WURE has shown a trend to increase with time, with showing the highest WURE over developed regions, and the lowest one over less-developed regions. Large amount of effort should be made in order to construct water-conserving society in China, which may be a result of reasonable social and economic developments; ② there exists a significant difference in WURE among 31 administrative regions, and the difference first increased and then decreased. Overall, the difference degree was enlarged in the past 11 years, but decreased between the regions having a similar WURE level; ③four types （High-High, High-Low, Low-High and Low-Low） reflecting more intuitive spatial association pattern of the growth of water resources efficiency in China were showed in this paper. The cluster characteristic of the growth rate of WURE is similar to that of GDP per capita. It is a unive