能-水耦合已成为当前生态环境问题研究和政策制定关注的焦点。一方面，整个社会水循环的取水、供水、输配水、用耗水、排水、以及污水回用等过程均会消耗大量的能源，各经济部门和居民消费终端的水资源使用过程也离不开能源的供应。因此，提高各部门取水用水阶段的耗能效率，优化产业间的水资源配置，改进空间区域输配水方案，对于减少能源消费，实现水资源系统的能源协同效益具有重要意义。另一方面，能源的开采、加工、生产和使用等各个环节都需要大量的水资源投入。因此，降低各类能源的耗水强度，发展低耗水的能源技术，优化能源结构，加强区域间能源生产和消费协作关系，是改善能-水空间分布不匹配，缓解能-水资源供需矛盾的重要途径。

京津冀地区作为我国重要的战略核心区，近年来经济发展与能、水短缺之间的矛盾突出。能源方面，京津冀地区能耗约占全国的10%，然而其中有70%依赖于其它地区的供应；水资源方面，该地区水资源储量仅为全国的0.6%，但消耗量占全国的8%，且51%的消耗为地下水，导致该地区地下水位持续下降，水资源压力不断攀升。此外，京津冀三地之间能、水利用效率差异较大，整体利用效率低，且由于地区产业布局不合理，产业对接难度大，没有配套完整的产业链条，未能形成资源的良性流通机制，进一步加剧了资源的过度消费和低效利用。在此背景下，围绕京津冀地区能-水耦合关系进行系统建模，并开展能-水协同管理和优化配置政策研究，对于该地区协同和可持续发展具有重要意义。

本文结合能-水耦合网络模拟、路径分析、因素分解和多目标优化等方法，构建针对区域能-水耦合问题研究的多目标优化配置模型，对耦合系统内部资源流通关系进行刻画，揭示能-水耦合的时空变化特征和社会经济驱动因素，进而识别能-水在部门间的直接和间接耦合路径，最后对能源、水资源管理和产业结构政策进行模拟优化，确定多情景能-水优化配置方案。本文选取京津冀地区为具体研究对象，相关主要研究内容和结果如下：

（1）构建了能-水耦合多目标优化配置模型。首先，通过集成混合生命周期分析、多区域投入产出分析和生态网络分析等方法，建立各部门资源消耗账户，构建多区域多部门的能-水耦合网络，并进一步建立耦合网络分析框架用以刻画耦合对部门动态和系统特征分析的影响，识别具有协同效益的部门和路径；其次，构建能-水耦合影响评估模型，量化区域间资源转移带来的积极影响和消极影响，及其对资源短缺压力的影响程度；然后，基于标准椭圆方法，构建能-水耦合的特征椭圆分析体系，辨识长时间序列下能-水耦合关系在空间分异的变化特征；同时，识别能-水直接和间接耦合路径，分解不同最终需求终端拉动下的具体耦合路径；最后，结合GAINS能源预测模块，构建能源发展的能-水耦合情景模拟框架，并将Social Group Optimization (SGO)多目标优化算法引入能-水耦合产业政策结构调整模式，对涵盖能、水、经济目标的区域资源经济调控提供方案参考。

（2）建立京津冀地区能-水耦合网络，分析了耦合影响下的部门动态和系统特征，具体包括：1）构建基于生命周期分析及投入产出分析的耦合清单核算框架，从部门层面和区域层面建立直接和间接能-水耦合清单；2）基于多区域投入产出分析，量化京津冀地区区域间部门间的能-水耦合流通关系，进一步构建能源网络、水网络、能耗水网络和水耗能网络；3）运用生态网络分析方法，量化部门间的控制和依赖关系，分析网络资源循环率和系统稳定性，识别具有显著能-水协同效益的经济部门。结果显示：1）在京津冀水网络中回流到城市部门的流量占总水资源转移流量的26.2%，而耦合水网络中的回流比例为27.1%。绝大部分循环通流由服务交通业贡献，为9.2%；农业和制造业虽然直接耗水量较高，但只占了11%左右的循环流，因此，农业和制造业在提高用水效率方面仍有很大空间；2）耦合对水网络的影响（±35%）小于对能源网络的影响（±150%）；另外，根据控制/依赖变化结果，受耦合关系影响，各部门的生产活动更加依赖制造业；3）耦合网络系统的鲁棒性分析结果显示，北京社会水循环系统最稳定、系统效率最低，而天津的社会水循环系统稳定性最差、效率最高；水耗能的网络系统稳定性高于能源网络，能耗水网络系统稳定性则低于水网络。结果说明能-水耦合关系可以增加能源网络的稳定性、降低系统效率；同时降低了水网络的稳定性、提高了系统效率；4）农林牧渔业、电力生产、采矿业和冶金等部门对耦合影响有放大效应（存在协同困境），是减少能源和水消耗的关键部门，相反，食品制造、服务业和批发零售业对耦合影响有缩小效应（存在协同效益）。

（3）评估京津冀地区能-水耦合影响。结合不同地区的资源禀赋差异，构建多区域能-水耦合影响评估模型，评估了京津冀地区同其他地区实体资源流动和隐含资源流动对其资源压力的影响。结果显示：1）对于耦合网络中的能源转移，京津冀地区从外界流入的能源中有84.3%对缓解区域间资源分布不均产生了积极影响，而流出的能源中有43.1%产生了消极影响；在京津冀地区内部，大部分产生积极影响的能源流出来自河北，产生消极影响的能源流出则主要来自北京；2）天津对外耦合关系呈现资源和增加值同时流出的情况，使其在资源和经济方面双重受损；北京则刚好相反，其对外耦合关系使其在资源和经济方面双重受益，河北则表现为资源输出和增加值流入，即在资源方面受损但在经济方面受益。

（4）辨识京津冀地区能-水耦合时空变化特征和分解经济驱动因素。基于空间特征椭圆系列定量刻画了长时间序列下能-水耦合关系在空间分异的变化特征；进而运用对数平均迪氏分解方法（LMDI），分析了京津冀地区能耗水和水耗能在长时间尺度下变化的驱动因素。经济驱动因素分析的结果显示：1）相比1997年，2015年北京、天津及河北的能源生产相关水耗分别增长为原来的2.41、3.94和2.83倍，而能源消费相关水耗分别增长为原来的2.92、3.78和4.30倍；整体上，促进能源相关水耗增加的主要原因是经济活动的变化，而促进其降低的主要原因是能源效率的提升；2）相比2010年，2018年北京、天津及河北的工业废水相关能耗分别增长为原来的1.91、1.50和1.11倍；整体上，电力消费强度和经济活动的变化是工业废水相关能耗增长的主要原因，而工业结构变化是其降低的主要原因。空间分异变化特征的结果显示：1）京津冀地区能耗水的空间分布总体呈现“东南-西北”的空间格局；在时间维度上，1997-2015年期间，能耗表现出明显的空间演化特征，大体向西南发展，且空间收缩；水耗能则在向西南收缩的同时，向东南扩张；2）京津冀地区供电和供水企业存在空间交错分布的特征，说明能-水的空间分布匹配程度较低，增加了二者相互间的直接消耗，从而在空间上呈现一定的能-水协同困境；但从相关结果随时间的变化趋势来看，能-水分布范围和方向角逐步趋近，表明这种空间上的协同困境在逐渐改善。

（5）识别京津冀地区的直接和间接能-水耦合路径。采用结构路径分析识别京津冀能-水耦合路径，辨识社会经济系统中具有显著协同效益的生产链条。结果显示：1）最终消费隐含的能源、水、能耗水以及水耗能超过60.0%分布于一、二级耦合路径；34.2%的农村能源消费，39.5%的城市能源消费，45.0%的政府能源消费，19.6%的固定资本形成的能源消费分布于一级耦合路径，水资源消费的路径结构同能源大致相同；2）供应路径中，能耗水主要集中在冶金业与其下游产业（建筑业等）形成的路径当中；水消费和水耗能集中在农业及其主要下游产业（如食品加工业）的路径当中；3）对于其他产业链条，虽然总体能耗不高，但水耗能占总能耗的比重较高，说明链条内部能-水耦合关系紧密，如批发零售业-石油加工炼焦及核燃料加工-煤炭开采和洗选业，减少这些链条的水耗，会对能耗的降低产生显著的协同效益；类似地，一些产业链条尽管耗水较低，如电器机械及器材制造业-金属制品业-金属冶炼及压延加工，但能耗水占比高，因此减少这些链条的能耗，会对降低水耗产生显著协同效益。

（6）京津冀地区能-水政策情景分析与优化配置。结合GAINS能源情景分析和京津冀多目标优化，量化了不同能源政策、水资源政策和产业结构政策对能-水耦合系统的响应，进而确定多情景能-水调控和配置方案。结果显示：1）根据能源政策情景模拟结果，相比转移能源生产地，增加节水能源技术的使用将产生更好的节水效果；2）区域间贸易结构优化在降低整体能源和水资源消耗方面有较大潜力，优化后全国整体的能源消耗下降了4.1%，水消耗下降了10.9%；北京、天津、河北的能源消耗则分别下降了17.8%、20.2%、15.5%，水消耗分别下降了12.1%、22.2%、11.4%；3）通过京津冀三地的产业结构调整和区域间产业转移方案设计，在水优先情景下，京津冀地区水消耗较优化前下降了20.6%，能源消耗下降了11.5%；在能源优先情景下，京津冀地区能源消耗较优化前下降了19.8%，水消耗下降了18.8%。

Energy-water nexus has become the focus of current research and policy development on ecological and environmental issues. On the one hand, the processes of water withdrawal, water supply, water distribution and distribution, water consumption, drainage, and reuse of sewage in the entire social water cycle consume a large amount of energy. The water utilization for various economic sectors and end users also needs energy inputs. Therefore, improving the energy efficiency of each process at the water withdrawal and water consumption phase, optimizing the water resource allocation among industries, and improving the water redistribution plan spatially are of great significance for reducing energy consumption and realizing the nexus synergy benefits of the water system. On the other hand, energy extraction, processing, production, and use require a large amount of water resources. Therefore, reducing the water consumption intensity of various energy types, developing energy technologies with low water intensity, optimizing the energy structure, and strengthening the cooperative relationship between energy producer and consumer is critical to improve the energy-water spatial distribution mismatch and alleviate the energy-water supply and demand contradiction.

As the strategic core area of China, the Beijing-Tianjin-Hebei (BTH) area has been facing increasing contradictions between economic development and energy and water shortages in recent years. The energy consumption in the BTH region accounts for about 10% of China, but 70% of which depends on the supply from the other regions. Considering water issues, the water resources reserve in this region is only 0.6%, while the consumption accounts for 8% of China, 51% of which is groundwater, resulting in the falling groundwater level and rising water scarcity. Besides, the energy and water use efficiencies among the BTH are quite different with low overall efficiency. Due to the unreasonable regional industrial layout, scarce connections among industries, and incomplete industrial chains, the efficient resource circulation mode cannot be formed, which further exacerbates the excessive consumption and inefficient usage of energy and water. In this context, integrated model construction for analyzing energy-water nexus relationship in the BTH region, and research on coordinated energy-water management and optimal allocation policies are of great significance for the regional coordinated and sustainable development.

By integrating network analysis, structural path analysis, decomposition analysis, and multi-objective optimization, this study sets up the regional integrated nexus model to explore the interwoven connections of energy consumption and water use in specific areas, reveal the temporal and spatial characteristics of energy-water nexus and its socio-economic driving factors. Then, the direct and indirect linkage pathways among sectors in the nexus networks are identified. Finally, based on the energy, water, and industrial structural scenarios, multi-objective optimization is conducted to figure out the optimized scheme of energy-water coordinated allocation. Taking the BTH region as a case, the major outcomes of this thesis are listed as follows:

(1) Regional energy-water nexus integrated model construction. Firstly, by integrating life cycle analysis, multi-regional input-output analysis, and ecological network analysis, multi-level energy and water consumption accounts are established. Then, a multi-regional and multi-sectoral energy-water nexus network containing energy network, water network, energy-related water network, and water-related energy network is constructed. Nexus network analysis framework is proposed to quantify the potential nexus impacts on sectoral dynamics and system characteristics, thereby identifying the sectors and pathways with synergetic benefits. Then, the energy-water nexus impact evaluation model is constructed, depicting the positive and negative impacts brought by multiregional resource transfer and their influences on the regional resource scarcity. Subsequently, the standard ellipse method is introduced to describe the spatial feature of energy-water nexus, based on energy consumption distribution ellipse, water consumption distribution ellipse, energy-related water distribution ellipse and water-related energy distribution ellipse, revealing the variation of energy-water nexus relationship in long time series. Meanwhile, the direct and indirect nexus pathways are identified, and the concrete socio-economic driving factors are decomposed for the energy-water nexus of the BTH. Finally, combining the energy scenario analysis of the GAINS model with the Social Group Optimization (SGO) model, the various energy-water related industrial structure adjustments policy implication are provided.

 (2)  Constructing energy-water nexus network, to analyze the nexus impacts on sectoral dynamics and system characteristics, including: 1) Building nexus accounting framework based on hybrid life cycle analysis from production and consumption perspectives; establishing a multi-scale direct and indirect energy-water nexus inventory at the sectoral level and regional level. 2) Quantifying the interregional and cross-sectoral energy-water nexus flow in the BTH based on multi-regional input-output analysis. Then, constructing the regional energy network, water network, energy-related water network, and water-related energy network. 3) Using ecological network analysis to quantify the control and dependence relationship among sectors, analyze the resource cycling rate and system robustness, and identify the critical sectors with notable energy-water positive synergies. The main results show that: 1) Finn cycling index (FCI) in the water network of the BTH is around 26.2%, which is lower than that of the energy network (27.1%). Most of the cycling flows are attributed to Services and Transport, with the percentage of 9.2%. While Agriculture and Manufacture only account for 11% of total cycling flows, their water consumption ranks the top. Thus, Agriculture and Manufacture should be critical sectors to implement energy-water nexus regulations, which will have an obvious positive nexus synergy effect. 2) Energy-water nexus has a great influence on the control relationship of Agriculture and Transportation in the energy network and has a greater impact on Mining, Manufacture, Electricity and gas supply, and Services and Transport Transportation in the water network. Besides, the impact of nexus on the water network (±35%) is generally less than that of the energy network (±150%). 3)As for the system robustness analysis of the nexus network, the results show that the water system of Beijing has the highest robustness but lowest efficiency. In contrast, the water system of Tianjin has the highest efficiency but lowest robustness. The robustness of the water-related energy consumption network is higher than that of the energy network. In comparison, the energy-related water consumption network has lower system robustness compared to the water network. Therefore, the energy-water nexus relationship will enhance the robustness of the energy network but decrease efficiency. In contrast, the nexus relationship lowers down the robustness of the water network while increases its efficiency. 4) Agriculture, Electricity and gas supply, Mining are the key sectors for the reduction of energy and water consumption, which have obvious negative trade-offs on energy-water nexus (coordination dilemma). Food processing, Service, and Wholesale and retails have obvious positive synergy impacts (synergy benefits).

 (3) Assessing energy-water nexus impact. Considering regional resource endowment into network analysis, the multi-regional energy-water nexus impact evaluation model is constructed to distinguish the impact degree of positive/negative physical and virtual flows on regional resource scarcity. The results show that: 1) The largest negative energy flows come from Shanghai and Beijing, which are 418.82 billion kWh and 248.74 billion kWh, respectively. Meanwhile, the amounts of positive energy inputs are the largest in Shanghai and Beijing, accounting for 31.3% and 18.6% of the total positive energy input, respectively. Most of the provinces in western China are positive energy exporter, while the eastern provinces are mostly positive energy receiving provinces. 2) Both resource and value-added show outflow in the external nexus relationship of Tianjin, which leads to resource and economic loss of the region itself. In contrast, Beijing has an inflow of both resources and value-added. Hebei has resource outflow and value-added inflow, indicating that the region losses from resource but benefits the economy.

 (4) Identifying spatial and temporal characteristics and decomposing socio-economic driving factors of energy-water nexus. The multi-year energy consumption, water consumption, energy-related water, and water-related energy in the BTH region are calculated. The LMDI method is used to analyze the driving factors of water-related energy and energy-related water. The results of driving force analysis illustrate that: 1) From 1997 to 2015, the water consumptions related to energy production in Beijing, Tianjin and Hebei increase by 2.41, 3.94 and 2.83 times, respectively, while the water consumptions related to energy consumption increases by 2.92, 3.78 and 4.30 times, respectively. The economic activities have increased the water consumption related to energy consumption, while energy efficiency has decreased the related water consumption of energy consumption. 2) From 2010 to 2018, the energy consumptions related to industrial wastewater in the BTH increase by 1.91, 1.50, and 1.11times, respectively. The power consumption intensity and economic activity are the main drivers for the increase of industrial wastewater-related energy consumption. In contrast, the industrial structure is the main driver for its decrease. The main results of spatial ellipse include: 1) The distribution position of standard deviation ellipse shows an overall spatial pattern of "southeast-northwest." The energy consumption shows obvious spatial evolution characteristics from 1997 to 2015, which can be summarized as the development to the southwest and spatial contraction. 2) The spatial and temporal distribution of energy-water nexus demonstrates a notable mismatch. As a result, the energy consumption of water withdrawal and water consumption of energy supply is increased, showing energy-water nexus negative trade-off at the spatial scale.

 (5) Identifying direct and indirect energy-water nexus pathways. Based on structural path analysis, the energy-water nexus path model is constructed to identify the direct and indirect nexus pathways within the BTH area. The results show that: 1) More than 60% of the energy, water, energy consumption, and water consumption implied in the final demand is distributed in the PL0 (production layer 0) and PL1 (production layer 1). 34.2% of rural energy consumption, 39.5% of urban energy consumption, 45.0% of government energy consumption, and 19.6% of energy consumption driven by fixed capital are distributed in the PL0. The path structure of water resource consumption is roughly the same as that of energy. 2) In the energy supply chain, direct energy and energy-related water are concentrated in the Mining sector and its corresponding downstream industries (Construction, etc.); direct water consumption and water-related energy consumption are concentrated in Agriculture or its downstream industries (Food processing sector, etc.). Compared with other final demand consumption, the major contributing pathways driven by fixed capital formation are distributed at a higher path level, mainly because the products consumed by fixed capital formation are more complex. 3) For other industrial chains, although the overall energy consumption is not high, the proportion of water energy consumption in the total energy consumption is relatively high. It indicates that the energy-water nexus of these chains is intensive, such as Wholesale and retail-Petroleum processing coking-Coal mining and washing, reducing the water consumption of these chains will have a significant synergy effect on reducing energy consumption. Similarly, some industrial chains have low water, but relatively high energy-related water consumption, such as Electrical machinery and equipment manufacturing-Metal products-Metal smelting processing, reducing the energy consumption of these chains will have a significant synergy effect on reducing water consumption.

 (6) Energy-water scenario analysis and multi-objective optimization. Combined with various energy mix scenarios based on the GAINS model and the multi-objective optimization for the BTH, the responses of energy-water nexus of the BTH to various energy policies, water resource management policies, and industrial adjustment regulations are illustrated. The results show that: 1) The sectors of Agriculture, Electricity, Chemical Manufacture, Domestic and Metal Manufacture have synergy effects on nexus impact, which should be considered as the critical sectors of reducing energy and water consumption. While Food processing, Service, and Wholesale and retails can reduce energy and water consumption via nexus impact. 2) Optimization of the inter-regional trade structure can reduce the overall water consumption of the BTH. The water consumption in the BTH would decrease by 11.6% compared to that before the optimization, and the total energy consumption would decrease by 16.9%. Specifically, after optimization, water consumption in the BTH is expected to decrease by 12.1%, 12.2%, and 11.4%, respectively. The energy consumption in the BTH would decrease by 7.8%, 10.2%, and 15.5%, respectively. The burden of water consumption would be shifted from water-scarce areas to water-abundant areas to a certain degree. 3) In the water-saving priority scenario, the water consumption decreases by 20.6% compared with that before the optimization, while the energy consumption decreases by 11.5%. In the energy-saving priority scenario, after optimization, energy consumption decreases by 19.8%, and water consumption decreases by 18.8%.