【摘要】 针对遥感能量平衡方法估算区域蒸散发（ET）存在的空间歧义性问题,本文以METRIC模型为例,基于地表温度（T_s）与叶面积指数（LAI）相关趋势图,按照"Ts频率直方图占比前5%的最高温度区、LAI频率直方图占比后5%的裸土区为‘干点’,Ts频率直方图占比后5%的最低温度区、LAI频率直方图占比前5%的湿地或农田为‘湿点’"的提取方案,综合确定研究区的干湿限,并通过下垫面涡度相关系统实测数据验证了模型估算精度。经检验,校准期和验证期的区域ET估算控制精度平均为15%,与以往研究成果相比,结果相对合理,提取方案在一定程度上可为METRIC模型主观确定干湿限提供帮助。将ET估算结果按土地利用分类提取,研究区不同土地利用类型表现为ET_水体> ET_林地> ET_耕地> ET_{未开发利用土地}> ET_{城乡工矿居民用地}≈ET_草地,成果对区域遥感ET精准估算研究具有一定参考意义。更多还原

【Abstract】 The METRIC model was chosen to reveal the spatial ambiguity issue encountered during the calculation of regional evapotranspiration（ ET） by remote sensing energy balance method. The study was based on the trend graph of surface temperature（T_s） and leaf area index（ LAI）. The dry pixel was chosen as the bare soil area where the Tsfrequency histogram accounted for the top 5% of the high-temperature region,and the LAI frequency histogram accounted for 5% of the region. The wet pixel was chosen as the wetland or farmland where the Tsfrequency histogram accounted for the last 5% of the low-temperature region,and the LAI frequency histogram accounted for the top 5% of the region. After that,the measured data of the eddy covariance system were used to verify the estimation accuracy of the model. The results showed that the average estimation accuracy of evapotranspiration in the calibration period and verification period was 15%,which was relatively reasonable compared with the previous research results. The selection scheme could help the Mapping Evapotranspiration with Internalized Calibration（ METRIC） model to subjectively determine the dry-wet edge. The evapotranspiration values of different land use types were calculated, and they were in an order of ET_water> ET_forest> ET_{arable land}> ET_{unused land}>ET_{urban and rural industrial land}≈ ET_grassland,which could be referred in researching the precision estimation of regional remote sensing evapotranspiration.更多还原