水资源的合理开发和充分利用，关系到人类发展的切身利益。随着人口增加、城市化进程加快、经济高速增长，以及随之而来的用水量增加和水环境破坏，使得水资源紧缺问题日益凸显。水资源短缺正成为制约区域经济社会发展的瓶颈，目前世界范围内约有50个国家缺水，约有7亿人面临缺水问题。对水资源进行合理、高效的使用是解决缺水问题的重要手段。产业用水结构是影响国家和区域水资源使用的重要因素，而实现产业用水结构的合理化是产业结构调整和水权改革的目标之一。对产业用水结构进行准确的计算、分析，在此基础上探索节水型的产业用水结构调整路径，有助于全面建立科学合理的产业用水结构，实现水资源的合理高效使用。产业用水结构既包括各产业部门直接用水量比例，又包括各产业部门之间的间接用水关系；目前研究主要关注前者，而对后者研究较少，不利于全面研究产业用水结构的性质与机理；为此，本论文将包含了直接和间接用水的“水足迹”概念引入产业用水结构分析，为其提供了新的视角。基于水足迹的视角，本论文以“现状核算—变化分析—路径选择”为总体研究思路，对产业用水结构进行了系列研究。首先对水足迹计算方法加以改进，以适应产业用水结构分析的需求，分别建立了适用于行政区域尺度和流域尺度的产业用水结构量化模型，计算了各尺度下的产业用水结构（现状核算）。在此基础上，根据水足迹相关指标进行了产业用水结构空间特征的标准化分析；随后在计算了不同年份产业用水结构的基础上，建立了产业用水结构时间变化因素分解分析模型，以探索其变化驱动机制（变化分析）。最后，根据所总结的产业用水结构变化机制，基于历史变化趋势及未来发展规划，对未来不同情景下的产业用水结构变化进行模拟预测，提出节水路径选择的建议（路径选择）。以上研究主要体现在以下几个方面：（1）在产业用水结构现状核算方法方面，本研究先建立了行政区域尺度的产业用水结构量化模型。针对以往方法在分部门定量分析用水在产业链中的流动情况上的不足，本研究基于经济投入产出—生命周期评价模型，发展了产业用水结构量化模型。该方法能够定量地明确各产业部门的直接用水去向和水足迹的来源，有助于定量地划分产业用水责任，为制定用水和节水的规划、法规和政策等提供科学依据。以北京市为实例对模型进行验证，采用最新可获得的2010年数据计算，发现北京市农业（7.80×108 m3）、其他服务业（5.01×108 m3）、运输邮政仓储、信息和计算机服务业（2.12×108 m3）为水足迹排名前三的产业部门，分别占总水足迹的29.1%、18.7%和7.9%。各产业部门的直接用水最终并不完全用于生产该部门自身的最终产品或服务，而会有一部分用于生产提供给其他产业部门的中间产品或服务（即提供给其他部门的虚拟水）；各产业部门的水足迹部分依赖于其他产业部门供给的中间产品或服务（即从其他部门的取得的虚拟水）。从经济系统整体节水的角度来说，各部门不仅应当提高用水效率、节约自身直接用水量，也可以考虑节约使用来自其他产业部门的原材料（中间产品），以实现整体节水。（2）建立了流域尺度的产业用水结构量化模型，将产业用水结构核算推广到流域尺度。流域尺度的产业用水结构研究与行政区域尺度相比，具有更符合水资源的自然属性的优点，但存在难以取得关键的投入产出数据的问题；为此，本研究引入区域投入产出表生成方法，建立了流域尺度的产业用水结构量化模型。以海河流域为实例对模型进行验证，得出海河流域2010年水足迹排名前三的产业部门为食品饮料及烟草业（9.28×109 m3）、建筑业（4.04×109 m3）、其他服务业（3.60×109 m3），分别占总水足迹的27.7%、12.1%和10.8%。从综合考虑直接和间接用水的角度，直接用水量较大的农业和电力、热力及水的生产供应业并不是综合用水量最高的部门，而大量使用这两个部门提供的中间产品的部门水足迹远高于其直接用水量。（3）在产业用水结构变化分析方面，为了总结不同空间尺度下区域产业用水的空间特征及成因，根据所取得的对行政区和流域尺度的产业用水结构的核算结果，从水足迹强度、最终使用类型水足迹、人均水足迹等多个特征指标角度对不同区域产业用水进行了标准化、定量化分析。研究发现，海河流域所有产业部门的水足迹强度均比北京市的水足迹强度要大。这一方面说明北京市属于海河流域中水资源极度匮乏的地区，另一方面也说明北京市比海河流域其它地区更加节约用水。海河流域大部分的水资源是靠本地水资源和调水工程调入的实体水资源提供的，而北京更多地依赖于进口调入产品中蕴含的虚拟水。同时，北京市的人均本地生产产品的水足迹小于人均本地使用的水足迹，说明该地区的虚拟水流动总体上是净进口的；而海河流域的人均本地生产产品的水足迹与人均本地使用的水足迹基本相等，说明该地区的虚拟水流动总体基本处于进出口平衡状态。（4）基于用产业用水结构量化模型对单一年份的产业用水结构的计算，进行了对产业用水结构的时间变化驱动机制的分析。鉴于以往的研究方法的数学精度与因素识别方面的不足，本研究耦合了经济投入产出—生命周期评价模型、人口—经济—技术模型和平均分解模型，发展了基于水足迹的产业用水结构时间变化驱动机制分析方法，精度高，因素选取也具有更好的实践意义。以北京1987年至2010年期间的产业用水结构时间变化为实例进行了分析，发现北京市的产业部门用水由3.92×109 m3下降到2.68×109 m3，水足迹下降最多的三个部门是农业（下降4.51×108 m3）、食品饮料及烟草业（下降3.20×108 m3）和纺织、服装及皮革制品业（下降2.93×108 m3）。第一产业水足迹下降了4.51×108 m3，第二产业水足迹下降了1.28×109 m3，而第三产业水足迹上升了4.85×108 m3。用水效率的进步是产业用水总量下降的最主要原因，而最终使用结构的变动也对产业用水总量下降起了一定的推动作用；另一方面，人口和人均最终使用的增长、产业结构的调整对产业用水总量的作用是促进其增长的，抵消了部分产业用水总量的下降趋势。（5）在产业用水结构未来路径选择方面，在对产业用水结构时间变化驱动机制进行了分析的基础上，利用情景预测法对未来产业用水结构的变化进行了预测。设置了一个各驱动因素遵循历史变化趋势的基准情景，和多个对不同驱动因素进行了调整的比较情景。根据基准情景预测，按照目前北京市发展趋势，未来到2020年，产业用水量将会上升到3.69×109 m3，出现约3.70×108 m3用水缺口。本研究给出的用水效率进步情景、产业结构调整情景、产业节水综合调整情景都给出了调整用水量、弥补用水缺口的办法，在满足用水需要的基础上，分别取得7.37×108 m3、4.14×108 m3、1.33×109 m3的供水富余量。结果表明，要实现未来用水与供水相协调的目标，提高用水效率和进行节水导向的产业结构调整都是可行的。而将提高用水效率和产业结构调整相结合，比仅用一种手段更加科学和效果显著。

The usage of water resources is closely related to the development of human beings. With the increase of population, the accelerated process of urbanization, and the rapid growth of economy, the consequent increase in water consumption and water environment damage are growing worse, which make the shortage of water resources become increasingly prominent. At present there are about 50 countries and 7000 million people suffering from water shortage. Using water rationally and effectively is a solution to the water problem. Industrial water utilization structure is an important factor affecting the usage of water resources. To rationalize the water utilization structure is the target of industrial reconstruction and the reform of water right system. An in-depth analysis of industrial water utilization structure and an exploration to the changes along with the influence mechanism are helpful to build a scientific and efficient water utilization structure comprehensively, which could lead to the rational use of water resources.The concept of industrial water utilization structure includes both the direct water use ratio of each industry sectors and the indirect water relations among them. The current researchers and managers mainly considered the former but ignored the latter, which is not conducive to the establishment of a scientific and efficient water utilization structure. This study introduces the concept of water footprint into industry water utilization structure analysis, which provides a new perspective for it. This study takes a “Current Status - Driving Forces - Path Selection” program. Firstly, this study developed the model of water footprint to build an improved model for industrial water utilization structure analysis. An analysis framework for industrial water utilization structures in administrative areas and river basins were built. The industrial water utilization structure in Beijing and the Haihe River Basin were calculated (Current Status). Moreover, the regional features of industrial water utilization structures in Beijing and the Haihe River Basin were analyzed to conclude the characteristics of water utilization structures. Thirdly, a decomposition analysis was conducted to understand the driving forces for the changes in industrial water utilization structure (Driving Forces). Finally, basing on the models for the industrial water utilization structure and its driving forces, a future scenario forecasting was implemented (Path Selection). A series of meaningful results were provided as follows.（1）Firstly, a developed model for industrial water utilization structure in administrative regions was developed, which was based on Economic Input-output-based Life-Cycle Assessment framework. The developed model could calculate the sectors’ direct water use, indirect water use, water footprints of final products, and water footprints of intermediate products, which is viable for explore the structure and characters of industrial water utilization structure. The model was implemented in Beijing as a test. In 2010 (the latest available data), the sectors of agriculture (7.80×108 m3), other services (5.01×108 m3), and transportation, postal services, information transmission, computer services and software (2.12×108 m3) were the top three sectors with the largest water footprints, which represented 29.1%, 18.7% and 7.9% of the total water footprint. For upstream sectors, the direct water use could partly be transformed into the intermediated products; and for downstream sectors, their water footprints came from both direct water use and intermediated products. Thus, the intermediated products should be used more efficiently to save water.（2）Studying the industrial water utilization structure at the scale of a river basin could better reflect the water resources situation, but the lack of input-output data is a barrier to implementing water utilization structure research from the perspective of a river basin. A Generating Regional Input-Output Tables method was introduced to bulid a model for industrial water utilization structure in the river basin. The model was implemented in the Haihe River Bain. In the Haihe River Bain, the sectors of food, alcohol and tobacco (9.28×109 m3), construction (4.04×109 m3), and other services (3.60×109 m3) were the top three sectors with the largest water footprints, which accounted for 27.7%, 12.1% and 10.8% of the total water footprint. From the perspective of water footprint, the agriculture and electricity, heat and hot water, which used most water directly, were no longer the sectors with the largest footprints, while those sectors largely used their intermediate prducts had larger water footprints than their direct water use.（3）After calculating the industrial water utilization structure in an administrative region (Beijing) and a river basin (the Haihe River Bain), a standardized comparison was conducted for the characteristics of water utilization structures in different regional scales. All the sectors in the Haihe River Bain had larger water footprint intensities than those in Beijing. It showed that Beijing was the most water-scarce region in the Haihe River Bain; and Beijing was more water-saving than other regions due to the lack of water. The Haihe River Bain mainly relied on local water resources, while Beijing more relied on imported virtual water. The per capita water footprint of local produced products in Beijing was less than that of local consumed products, which meant that the virtual water imports were larger than the exports. The per capita water footprint of local produced products in the Haihe River Bain generally equaled that of local consumed products, which showed a nearly balanced virtual water exchange.（4）To analyze the driving factors of industrial water utilization structure, a decomposition framework combined Economic Input-output-based Life-Cycle Assessment method, “Impact-Population-Affluence-Technology” model and Weighted Average Decomposition model was built. The decomposition framework for the driving factors of industrial water utilization structure had better accuracy and more realistic significant factors than previous methods. The change in industrial water utilization structure in Beijing was taken as a case. During 1987 to 2010, the industrial water use in Beijing decreased from 3.92×109 m3 to 2.68×109 m3. The sectors with the largest decrease were agriculture (-4.51×108 m3), food, alcohol and tobacco (-3.20×108 m3), and textile, clothing, and leather products (-2.93×108 m3). The development of water use efficiency is the dominant reason of industrial water utilization decrease, and the final use structure can help water saving as well. The growth of population and per capita final use is promoting the growth of industrial water utilization, which can partly offset the decline trend of water use. The adjustment of industrial structure can make either positive or negative change to water saving, so the industry reconstruction plan should take the water use structure into account.（5）A reference scenario and four comparative scenarios were set to predict the potential future water use structure. According to the development trend of Beijing city at present, the future industrial water use will rise to 3.69×109 m3 and cause a water resources gap of 3.70×108 m3 in 2020. To prevent this potential problem, the water use efficiency developing scenario, industrial reconstruction scenario, and comprehensive water-saving scenario have given possible paths, which would achieve water surpluses of 7.37×108 m3、4.14×108 m3、1.33×109 m3, respectively. To achieve the goal of relatively balanced water use and supply, improving the water use efficiency and reconstruction in industrial and consumptive structures towards water saving are both feasible; and the combining of them is easier to implement and more effective than relying only on one aspect of the two strategies.