摘要
综合考虑产业用水效率、产业用水结构、经济水平、水资源禀赋及水资源开发利用率5个因素对中国用水效率的影响,基于扩展的Kaya恒等式建立因素分解模型,应用LMDI分解方法对1997~2008年中国用水效率变化进行分解分析,计算各驱动效应因素的相对贡献率,测度它们的影响程度和方向。结果表明:产业用水效率效应和经济水平效应是影响中国用水效率变化的最显著因素。通过计算各驱动效应的绝对贡献率,采用最小方差法将中国31个行政区(港澳台除外)确定为4种空间驱动类型,进而分析用水效率变化驱动效应的分布特点。研究结果对于正确理解中国用水效率的变化与经济发展之间的关系以及产业结构的调整与可持续发展研究具有一定的参考价值。
Five factors respectively named industrial water use efficiency, industrial water use structure, economy development level, water endowment and water resources development ratio are comprehensively considered as the main factors that impact on the water use efficiency, and the factor decomposition model under the extended Kaya identity was established, the decomposition model was solved by the method of LMDI (Longarithmie Mean Divisia Index), then the driving effects of the five factors were calculated and the main factors were determined, which may useful for the understanding of water use efficiency change from 1997 to 2008 in. China. The relative contribution ratio of the five factors were calculated to measure their function degree and direction. The result shows that the industrial water use efficiency and economy development level are the most significant factors to affect water use efficiency in China. This paper calculated the absolute contribution ratio of the five factors, then the 31 regions of China (Hong Kong, Macao and Taiwan excluding)were divided into four spatial driving types by applying the minimum variance method, and the distribution characteristics of driving effects were further analyzed. The result is of significance economy development, and also has certain for exactly understanding the relation of the water use efficiency and reference significance for the regulation of industrial water use strutture and sustainable development.
出处
《地理科学》
CSCD
北大核心
2011年第10期1213-1220,共8页
Scientia Geographica Sinica
基金
国家自然科学基金项目(40501013)
国家社会科学基金项目(11BJY063)资助
关键词
用水效率
驱动效应测度
Kaya恒等式
LMDI分解法
最小方差
water use efficiency
driving effects measurement
Kaya identity
Longarithmic Mean Divisia Index method
minimum variance