The high spatial and temporal resolution of recently developed evapotranspiration (ET) products facilitates agricultural water-savings in irrigated areas as well as improved estimates of crop yield, especially in arid and semi-arid regions. However, cloud cover interferes with ET estimates, in particular when using thermal-infrared-based models in temperate and tropical regions. Previous studies have shown that the two-source energy balance (TSEB) model coupled with soil moisture (TSEB-SM) has great potential for estimating surface ET by overcoming this issue. In this study, the TSEB-SM model was first used to generate a spatiotemporally continuous 1 km daily ET dataset across the Heihe River Basin in China from 2000 to 2020, which was then evaluated against four spatially distributed sites (Arou, Huazhaizi, Daman, and Sidaoqiao) and further compared with the two most widely used daily ET datasets (PML-V2 (Penman–Monteith–Leuning) and SEBAL (surface energy balance algorithm for land)). The results showed that the newly developed ET dataset agrees well with ground-based observations and outperforms the PML-V2 and SEBAL products in precisely characterizing the seasonal fluctuations and spatial distribution as well as the spatiotemporal trends of ET. In particular, ET in the Heihe River Basin exhibits clear regional differences. The upstream and midstream grassland and irrigated oasis areas provide much higher annual ET than the downstream desert areas, with a difference of up to 600 mm/year. A three-cornered hat (TCH)-based pixel-by-pixel analysis further demonstrated that the TSEB-SM and PML-V2 products have substantially smaller relative uncertainties as compared to SEBAL ET. In general, the proposed ET datasets are expected to be more beneficial for irrigation scheduling and to provide more efficient water management across the Heihe River Basin.

Remote sensing-based models usually have difficulty in generating spatio-temporally continuous terrestrial evapotranspiration (ET) due to cloud cover and model failures. To overcome this problem, machine learning methods have been widely used to reconstruct ET. Therefore, studies comparing and evaluating the accuracy and effectiveness of reconstruction among different machine learning methods at the basin scale are necessary. In this study, four popular machine learning methods, including deep forest (DF), deep neural network (DNN), random forest (RF) and extreme gradient boosting (XGB), were used to reconstruct the ET product, addressing gaps resulting from cloud cover and model failure. The ET reconstructed by the four methods was evaluated and compared for Heihe River Basin. The results showed that the four methods performed well for Heihe River Basin, but the RF method was particularly robust. It not only performed well compared with ground measurements (R = 0.73) but also demonstrated the ability to fully reconstruct gaps generated by the TSEB model across the entire basin. Validation based on ground measurements showed that the DNN and XGB models performed well (R > 0.70). However, some gaps still existed in the desert after reconstruction using the DNN and XGB models, especially for the XGB model. The DF model filled these gaps throughout the basin, but this model had lower consistency compared with ground measurements (R = 0.66) and yielded many low values. The results of this study suggest that machine learning methods have considerable potential in the reconstruction of ET at the basin scale.

Remote sensing-based models usually have difficulty in generating spatio-temporally continuous terrestrial evapotranspiration (ET) due to cloud cover and model failures. To overcome this problem, machine learning methods have been widely used to reconstruct ET. However, studies comparing and evaluating the accuracy and effectiveness of reconstruction among different machine learning methods remain scarce. In this study, four popular machine learning methods (deep forest, deep neural network, random forest, extreme gradient boosting) were used to reconstruct the ET product, addressing gaps resulting from cloud cover and model failure. The ET reconstructed by four methods were evaluated and compared in Heihe River Basin. The results showed that four methods performed well in the Heihe River Basin, but the RF method was particularly robust. It not only performed well compared with ground measurement (R = 0.73), but also reconstructed ET throughout the basin. Validation based on ground measurement showed that DNN and XGB models performed well (R > 0.70). However, few gaps still existed in the desert after reconstruction, especially for the XGB model. The DF model filled these gaps throughout the basin, but the model had lower consistency compared with ground measurement (R = 0.66) and yielded many low values. The results of this study suggested that machine learning methods had considerable potential in reconstruction of ET at regional scale.