受干旱-半干旱气候的影响，巴基斯坦的水资源非常稀缺，沿河流域的农业灌溉和居民用水均极度依赖于印度河水系的供给。印度河流域是气候变化的敏感区，同时也是研究的空白区。流域内水文气象观测站点数量稀少，多数研究仅集中在印度河上游流域。为了探究印度河流域水文气象要素的变化趋势及其关键驱动因素，论文广泛收集了包括上中下流域在内的地面监测站点数据以及卫星/同化降水资料。此外，论文还基于不同类型水文模型开展了建模，探讨不同子流域径流变化及其成因。首先，论文利用微波降水监测卫星TRMM中心发布的TMPA数据以及气候预报系统降水数据CSFR，评估了气候灾害群红外降水卫星数据CHRIPS。结果表明：CHRIPS降水数据集的性能因地区和季节不同而存在差异，对海拔较高的印度河上游流域模拟精度较差，但在中游和下游流域表现较好。将上述多源降水数据输入SWAT模型，模拟研究区的径流变化，结果发现：即使区域雨量站空间分布十分稀疏，其模拟的径流量比使用3种开源卫星/同化降水数据（CHIRPS、TMPA和CSFR）的精度要高。后三者中，利用TRMM数据集模拟的径流量整体精度最佳，是唯一能对日径流量进行较为准确模拟的数据集。第二，论文对印度河流域的实测水文气象要素（降水、温度和径流量）进行了时空趋势分析。结果表明：印度河流域总体有变暖趋势；流域上游年径流量呈上升态势，主要与当地气温升高有关，温度升高加速了冰雪融化，从而导致径流量增加；中游和下游流域的年径流量均呈下降变化，但原因有所不同，前者是由于降水减少，而后者则是由于其上游地区的水量调蓄和分流，显著减少了下游流域的入流量。第三，研究进一步利用abcd水文模型、Budyko模型和SWAT模型3种水文模型，模拟和分析了印度河流域内的Soan和Gilgit流域的径流变化。abcd模型和Budyko模型模拟的径流量均有下降趋势，原因是Soan流域内的气候和土地利用类型变化，降雨减少和潜在蒸散增加对径流变化的贡献率为68%，土地利用类型变化的贡献率为32%。而Gilgit流域的降水、气温和径流呈明显上升的趋势，气候变化是当地径流发生变化的主要因素，贡献率为97%。综上所述，本研究收集了多源水文气象资料，探讨了印度河流域水文气象要素的变化趋势及其相互作用机制。在探寻区域气候变化及其对印度河流域的影响研究中，卫星遥感技术起到了不可或缺的作用，之后研究需进一步加强流域地面观测网点的建设和生态水文过程的模拟。
The Indus River is one of the largest rivers in the world, with 47% of its total drainage area falling into Pakistan. As a country controlled by arid and semi-arid climate, the local agricultural and domestic sectors heavily rely on the water sourced from the Indus River and its tributary. However, limited ground hydrometeorological data were involved, and most studies only focused on the Upper Indus Basin (UIB). In order to understand the dynamics of hydrometeorological conditions and the driving factors behind in the Indus Basin, this study not only involved a rich source of in-situ observations but also the satellite/reanalysis rainfall products for investigation. In addition, multiple hydrological models were introduced to interpret the runoff changes in the target basin. Firstly, the Climate Hazard Group InfraRed Precipitation Satellite’s (CHIRPS) data was evaluated in the Indus basin. The results of CHIRPS were evaluated with the precipitation data sourced from the Tropical Rainfall Measuring Mission (TRMM), the Multi-Satellite Precipitation Analysis (TMPA) and the Climate Forecast System Reanalysis (CFSR) data. Results show that the performance of CHIRPS rainfall dataset differed from regions and seasons; poor performance has been observed in the upper Indus basin, but it performed well in the Middle Indus Basin (MIB) and Lower Indus Basin (LIB). These precipitation data were further imported into the Soil and Water Assessment Tool (SWAT) model to test the simulated runoff in the Gilgit basin. It has been observed that the measured precipitation data, which are even sparsely distributed, could yield a better performance in the streamflow simulation than using three open-source precipitation datasets (CHIRPS, TMPA and CSFR). Comparatively, TRMM satellite precipitation data generally yielded the best performance, and it was the only one that can achieve a satisfactory simulation of daily streamflow.Secondly, the hydro-meteorological trends were tested in the Indus basin. Trend analysis indicated that a significant warming trend has been observed throughout the Indus basin. The annual precipitation presented significant increasing trends in the LIB. The annual streamflow presented significant increasing trends for UIB and it decreased in LIB. The increase in streamflow of UIB may be due to the warming temperature and glacier melting. However, the decreased streamflow in LIB could possibly be attributed to the water regulation in the upstream and middle stream.Thirdly, the runoff was simulated in the Soan and Gilgit sub-basins using different hydrological models to understand the streamflow trends, including the abcd model, the Budyko framework and the SWAT model. Both the abcd model and the Budyko framework detected a decrease in runoff in the Soan basin due to the changes in climate change and land use change; decreased precipitation accompanied with increased potential evapotranspiration contributed 68% of the change in runoff. The Gilgit catchment showed a significant upward trend in the precipitation, temperature and streamflow. It was attributed that climate change contributed 97% of the change in runoff within the Gilgit basin, and the other 3% was by land use change. It has been concluded that the high relative contribution of climate change to runoff change is mainly due to the change in climate variables. Overall, this work collected multisource hydrometeorological data to explore the hydrometeorological trend and their relationship in the Indus basin. Results demonstrated that the runoff in the UIB was experiencing a significant increasing trend while it was decreased remarkably in the MIB and LIB. The increase in streamflow of UIB can be attributed to temperature warming, but the reasons for the shrinkage in runoff in the MIB and LIB are different; the former is related to a decreasing trend in precipitation and the major contributor for the latter case is associated with the water regulations for agricultural irrigation and power generation. The study highlights an irreplaceable place of satellites in complimenting ground observations to quantify regional hydrometeorological changes. Moreover, enhancement of ground monitoring and ecohydrological modeling’s are called for, due to the sensitivity and complexity of the Indus Basin to climate change.
