水分利用效率(Water Use Efficiency，WUE)是认识干旱区生态系统水循环和碳循环之间的耦合关系、指导干旱区水资源利用和生态系统恢复重建的重要指标。探究区域WUE的时空分布规律及调控手段，可以更好地应对全球变化对生态系统的影响。已有研究对区域WUE的时空变异特征及影响因素进行了探讨，但在干旱区开展相关研究相对较少，且WUE在气候及土地利用变化下的响应尚不明确。本文以中亚干旱区典型内陆河流域艾比湖流域为研究区，探索流域WUE对气候变化以及土地利用的响应，并结合RS-DTVGM水文模型对未来气候变化情景下的WUE进行了预测。主要研究内容及结论如下： 1、利用MODIS GPP和ET数据计算艾比湖流域WUE，分析艾比湖流域水分利用效率的时空变异特征规律。流域生长季多年平均WUE为2.65 g C/kg H2O，总体呈现出越远离艾比湖WUE越低的空间分布规律，流域四种地貌类型的WUE大小规律为：绿洲区＞高山区＞山前冲洪积扇区＞尾闾湖区。艾比湖流域2001-2016年生长季年均WUE变化区间为2.24~3.12 g C/kg H2O，总体呈下降趋势。整个流域WUE年际变化率区间为-1.15~0.69 g C/kg H2O，81.96%的区域WUE年际变化率区间为-0.05~0.05 g C/kg H2O。根据WUE slope空间分布特征结合地貌类型，将流域划分为五个分区。 2、明确GPP与ET对艾比湖流域WUE的主导贡献，探究现状年WUE对气候与土地利用变化的响应。2016年与2001年相比，土地利用累积变化面积约为2172km2，绿洲扩张和草地退化是最主要的变化类型。在ET和GPP两个因子中，ET对全流域WUE的变化具有更大贡献。气温和降水与洪积扇持平区和高山持平区的ET存在强相关性，降水与洪积扇持平区和高山持平区的GPP存在强相关性，说明气候变化是影响这两个区域WUE变化的主要因素。土地利用变化对绿洲区的WUE产生了显著影响，非耕地向耕地转化以及普通耕地向节水耕地转化是绿洲内分区WUE发生变化的主要原因。 3、构建艾比湖流域RS-DTVGM模型空间数据库，验证径流模拟结果的可靠性。精河站在率定期和验证期日径流量模拟结果与实测值的相关系数分别为0.82和0.76，纳什效率系数分别为0.52和0.57。温泉站在率定期和验证期日径流量模拟结果与实测值的相关系数分别为0.65和0.66，纳什效率系数分别为0.32和0.28。 4、分析2001-2016年艾比湖流域生长季水循环各要素的年际与年内变化。2001-2016年艾比湖流域生长季降水量总体呈微弱的下降趋势，生长季的降水量主要集中在5-8月份。融雪量总体呈上升趋势，月平均融雪量呈现出先增加后减少的趋势。蒸散发量总体呈微弱的上升趋势，月平均蒸散发量总体基本呈现出逐渐下降的趋势。产流量总体呈上升趋势，月平均产流量总体呈现先增加后减少的变化规律。 5、构建各类型区WUE数值模型，基于IPCC三种气候变化情景，预测2050年艾比湖流域WUE的变化情况。利用未来气候变化情景的气温和降水数据模拟2050年艾比湖流域博尔塔拉河和精河上游来水情况，未来气候条件下艾比湖上游来水径流量有明显的增加。如果维持现状年的入湖量水平，三种未来气候情景下，2050年生长季绿洲区用水量将分别上升51.31%、56.69%和62.21%。预测结果显示，2050年艾比湖全流域WUE平均将下降8.15%，绿洲上升区WUE平均将下降11.56%，绿洲持平区WUE平均将下降15.85%，绿洲下降区WUE平均将下降16.34%，洪积扇持平区WUE平均将下降2.75%，高山持平区WUE平均将下降2.98%。

Water use efficiency (WUE) is an important indicator to understand the coupling relationship between the water cycle and carbon cycle in the arid region ecosystem, and to guide the use of water resources in arid regions and the restoration and reconstruction of ecosystems. Exploring the spatial and temporal distribution of regional WUE and its control measures can better cope with the impact of global changes on the ecosystem. There have been studies on the spatial and temporal variability and influencing factors of regional WUE, but relatively few researches have been carried out in arid regions, and the response of WUE under land use changes is not yet clear. This paper chose the Ebinur Lake Basin, a typical inland river basin in the arid region of Central Asia, as the study area, explored the WUE response to climate and land use change, and combined the RS-DTVGM hydrological model to predict WUE under future climate change scenarios. The main research contents and conclusions are as follows: 1. Using the MODIS GPP and ET data to calculate the WUE in the Ebinur Lake Basin, the temporal and spatial variability of the water use efficiency of the Ebinur Lake Basin was analyzed. The average WUE in the growing season of the basin was 2.65 g C/kg H2O, which shows that the farther away from the Ebinur Lake, the lower the spatial distribution pattern of WUE. The order of WUE for the four landforms types in the basin is: oasis > mountain area> alluvial-pluvial fan > tail lake zone. The average annual WUE of the Ebinur Lake Basin during the 2001-2016 growth season ranged from 2.24 to 3.12 g C/kg H2O, showing a declining trend overall. The interannual rate of change of WUE in the whole basin varies within the range of -1.15-0.69 g C/kg H2O. In most regions, the range of annual WUE changes was within the range of -0.05-0.05 g C/kg H2O. According to the WUE slope spatial distribution characteristics combined with landform types, the watershed was divided into five subregions. 2. Identified the dominant contribution of GPP and ET to WUE, and explore the current status of the WUE response to climate and land use changes. Compared with 2001, the cumulative area of land use change was about 2172 km2, and the expansion of oasis and grassland degradation are the main types of change. Among the two factors of ET and GPP, ET has a greater contribution to the change of WUE in the whole basin. There were strong correlation between temperature and precipitation and the ET of the flat area in alluvial-pluvial fan and flat area in mountain area, and there were strong correlation between precipitation and the GPP of these two regions, which explained that climate change was the main factor affecting WUE in these two regions. Land use change had a significant effect on the WUE of the oasis area. Conversion of non-cultivated land to arable land and conversion of ordinary arable land to water-saving arable land were the main reasons for the change in WUE within the oasis. 3. Constructed a spatial database of RS-DTVGM model in the Ebinur Lake Basin to verify the reliability of the runoff simulation results. The correlation coefficients between simulated and measured values of runoff at the regular and verification phases of Jinghe station rate were 0.82 and 0.76, respectively, and the Nash efficiency coefficients were 0.52 and 0.57, respectively. The correlation coefficients between simulated and measured values of runoff at the regular and verification phases of Wenquan station rate were 0.65 and 0.66, respectively, and the Nash efficiency coefficients were 0.32 and 0.28, respectively. 4. Analyzed the interannual and intraannual variations of the water cycle elements in the growth season of the Ebinur Lake Basin from 2001 to 2016. In the growing season during 2001-2016, the precipitation generally showed a slight downward trend, and was mainly concentrated in May-August. The total amount of snowmelt showed a rise trend, and the average monthly snowmelt volume showed an increase first and then a decrease. The evapotranspiration overall showed a slight upward trend, and the monthly mean evapotranspiration generally showed a gradual decline. The total runoff showed an upward trend, and the average monthly runoff showed an increase first and then decreases. 5. Constructed WUE numerical model for each type of subregion, and forecasted the changes in WUE of the Ebinur lake basin in 2050 based on the three IPCC climate change scenarios. Based on the three IPCC climate change scenarios, the changes in WUE of the Ebinur lake basin in 2050 were predicted. Using the temperature and precipitation data of future climate change scenarios to simulate the incoming water in the Bortala River and the Jinghe River in the Ebinur Lake Basin in 2050, the runoff in the upper reaches of Ebinur Lake would increase significantly in the future climate conditions. If the level of lake intake in the current year was maintained, under the three future climate scenarios, the water consumption in the oasis area of the 2050 growth season would increase by 51.31%, 56.69% and 62.21%, respectively. The forecast resulted show that in the 2050, the WUE in the whole basin would drop by 8.15%, the WUE in ascending area in oasis would drop by 11.56%, the WUE in flat area in oasis would drop by 15.85%, the WUE in descending area in oasis would drop by 16.34 %, the WUE in flat area in alluvial-pluvial fan would drop by 2.75 %, and the WUE in flat area in mountain area would drop by 2.98%.