基于GLASS数据的青藏高原2001—2018年蒸散发时空变化分析

准确认知青藏高原蒸散发时空变化特征，为当地可持续农业的水资源规划及理解高原气候变化具有重要现实意义。研究基于GLASS陆表潜热通量产品，采用Mann-Kendall趋势分析方法，结合青藏高原生态地理分区方案，分析了2001—2018年青藏高原蒸散发的时空变化特征及其与气温、降水和植被的关系。结果表明：(1)GLASS ET产品可以较好地表征青藏高原蒸散发的时空分布特征；(2)青藏高原多年平均蒸散发为296.52 mm，整体上呈现出东南高西北低的空间格局，其中东喜马拉雅南翼最高（690.94 mm），柴达木盆地最低（163.47 mm）；(3)近18 a来，青藏高原蒸散发年际变化呈波动性上升，只有东喜马拉雅南翼在下降；(4)研究期间，青藏高原蒸散发以显著性增长趋势为主，占47.44%，主要位于高原东部边缘和中西部腹地，呈显著性减小趋势的地区占3.82%，主要集中于东喜马拉雅南翼；(5)蒸散发的空间分布在干旱区与气温呈负相关，在湿润区呈正相关，与降水空间格局总体呈正相关；(6)蒸散发与NDVI的空间分布呈较好的正相关，与NDVI的变化趋势相关性较为复杂，大部分呈正相关，小部分呈负相关。

Spatiotemporal Changes of Evapotranspiration on the Qinghai-Tibet Plateau from 2001 to 2018 based on GLASS Data

It is of great practical significance for the local sustainable agricultural water resources planning and to understand the plateau climate change to study the characteristics of the spatiotemporal changes of evapotranspiration in the Qinghai-Tibet Plateau. Based on the GLASS land surface latent heat flux product to analyze the characteristics of the spatiotemporal changes of evapotranspiration and its relationship with temperature， precipitation， and vegetation in the Qinghai-Tibet Plateau from 2001 to 2018， with the Mann-Kendall trend analysis method， in consideration of China's ecogeographical divisions. The results showed that： ①GLASS ET can reasonably simulate the distribution characteristics of evapotranspiration over the Qinghai-Tibet Plateau. ②The multi-year annual average evapotranspiration in the Qinghai-Tibet Plateau is 296.52mm， with higher values in the southeast but lower values in the northwest， the southern wing of the Eastern Himalayas is the highest （690.94 mm） and the Qaidam Basin is the lowest （163.47 mm）. ③The inter-annual variation of evapotranspiration in the Qinghai-Tibet Plateau has increased volatility， and only the southern flank of the Eastern Himalayas has been declining in the past 18 years. ④During the study period， the evapotranspiration of the Qinghai-Tibet Plateau with a substantial increase trend， accounting for 47.44%， mainly located at the eastern edge of the plateau and the Midwestern hinterland； with a significant decreasing trend accounted for 3.82%， mainly concentrated in the southern wing of the Eastern Himalayas. ⑤The spatial distribution of evapotranspiration is negatively correlated with temperature in arid areas， and positively correlated with humid areas； it is generally positively correlated with the spatial pattern of precipitation. ⑥Evapotranspiration has a good positive correlation with the spatial distribution of NDVI； The correlation with the change trend of NDVI is more complicated， mostly positively correlated， and a small part are negatively correlated.