The aim of this work is to develop and evaluate a performance comparison for estimating satellite-based actual evapotranspiration (AET) using the Penman–Monteith (PM) and Priestley–Taylor (PT) approaches and to create a spatial AET map in an ungauged sub-humid tropical river basin. A few studies have compared the PM and PT model performance with a MODIS evapotranspiration data product (MOD16A2). Estimated AET values by the PT approach (AET_PT), PM model with aerodynamic conductance (Ga) computed using the Leuning equation (AET_PM), and Ga computed using the Choudhury equation (AET_PMCH) were extracted for each of the 304 pixels for each day, and pixelwise comparisons were made with the corresponding MOD16A2 AET estimates for validation. As shown by the low RMSE values (0.19–0.23 mm/day), the PM model suggested in this analysis with Ga computed using the (AET_PMCH) equation turned out to be the better model. Also, using the (AET_PMCH) equation significantly reduced PBIAS values for all days examined. Topography, land use and land cover (LU/LC), temperature, and moisture availability conditions appear to influence AET variations across the basin. For the eight total Julian days in summer and winter for selected wet (2007) and dry (2012) years the for period 2006–2017, pixelwise maps depicting spatial variability were developed using MATLAB for the AET_PMCH approach for selected available cloud-free MODIS image data.

这项工作的目的是开发和评估使用 Penman-Monteith (PM) 和 Priestley-Taylor (PT) 方法估算基于卫星的实际蒸散量 (AET) 的性能比较，并在未测量的子区域创建空间 AET 地图- 潮湿的热带河流盆地。一些研究将 PM 和 PT 模型性能与 MODIS 蒸散数据产品 (MOD16A2) 进行了比较。通过 PT 方法估计的 AET 值 (AET _PT )、使用 Leuning 方程 (AET _PM )计算的空气动力传导 (Ga) 的 PM 模型和使用 Choudhury 方程 (AET _PMCH ) 计算的 Ga) 为每天 304 个像素中的每一个提取，并与相应的 MOD16A2 AET 估计值进行逐像素比较以进行验证。正如低 RMSE 值（0.19-0.23 毫米/天）所示，本分析中建议的 PM 模型和使用 (AET _PMCH ) 方程计算的 Ga被证明是更好的模型。此外，使用 (AET _PMCH ) 方程显着降低了所有检查天数的 PBIAS 值。地形、土地利用和土地覆盖 (LU/LC)、温度和湿度可用性条件似乎会影响整个盆地的 AET 变化。对于选定的湿（2007 年）和干（2012）年的夏季和冬季的 8 个儒略日，在 2006-2017 年期间，使用 MATLAB 为 AET _PMCH开发了描绘空间变异性的像素地图 选择可用的无云 MODIS 图像数据的方法。