An Automated and Improved Methodology to Retrieve Long-time Series of Evapotranspiration Based on Remote Sensing and Reanalysis Data

The large-scale quantification of accurate evapotranspiration (ET) time series has substantially been developed in recent decades using automated approaches based on remote sensing data. However, there are still several model-related uncertainties that require precise assessment. In this study, the Surface Energy Balance Algorithm for Land (SEBAL) and meteorological data from Global Data Assimilation System (GLDAS) were used to estimate long-term daily actual ET three endmember selection procedures: two land cover-based models, one with (WF) other without (WOF) morphological functions, Allen method (with default percentiles) 2270 Landsat images. Models evaluated 23 flux tower sites four main vegetation cover types as well different climate types. Results showed functions (WF_ET) generally performed better than approaches. Climate-based classification assessment provided clearest discrimination between performance humid category. For zones, methods, especially WF, appropriately outperformed Allen. was similar sub-humid, semi-arid arid climates together; approach therefore recommended avoid need dependency maps. Tower-by-tower validation also WF best at 12 sites, WOF 5 6, suggesting use maps alone does not explain differences models approach. Additionally, satisfactory error metrics results when comparing EC estimations measurements, root mean square (RMSE) ≈ 0.91 1.59 mm·day−1, coefficient determination (R2) 0.71 0.41, bias percentage (PBias) 2% 60% crop non-crop respectively, supports GLDAS forcing datasets estimation Overall, given thorough evaluation large scale confirmed validity types, study can be considered an important contribution global retrieval long ET.