Evapotranspiration (ET) is a vital variable for land‐atmosphere interactions that links surface energy balance, water, and carbon cycles. The in situ techniques can measure ET accurately but the observations have limited spatial and temporal coverage. Modeling approaches have been used to estimate ET at broad spatial and temporal scales, while accurately simulating ET at regional scales remains a major challenge. In this study, we upscale ET from eddy covariance flux tower sites to the regional scale with machine learning algorithms. Five machine learning algorithms are employed for ET upscaling including artificial neural network, Cubist, deep belief network, random forest, and support vector machine. The machine learning methods are trained and tested at 36 flux towers sites (65 site years) across the Heihe River Basin and are then applied to estimate ET for each grid cell (1 km × 1 km) within the watershed and for each day over the period 2012–2016. The artificial neural network, Cubist, random forest, and support vector machine algorithms have almost identical performance in estimating ET and have slightly lower root‐mean‐square error than deep belief network at the site scale. The random forest algorithm has slightly lower relative uncertainty at the regional scale than other methods based on three‐cornered hat method. Additionally, the machine learning methods perform better over densely vegetated conditions than barren land or sparsely vegetated conditions. The regional ET generated from the machine learning approaches captured the spatial and temporal patterns of ET at the regional scale.

Land surface evapotranspiration (ET) is an important component of the surface energy budget and water cycle. To solve the problem of the spatial‐scale mismatch between in situ observations and remotely sensed ET, it is necessary to find the most appropriate upscaling approach for acquiring ground truth ET data at the satellite pixel scale. Based on a data set from two flux observation matrices in the middle stream and downstream of the Heihe River Basin, six upscaling methods were intercompared via direct validation and cross validation. The results showed that the area‐weighted method performed better than the other five upscaling methods introducing auxiliary variables (the integrated Priestley‐Taylor equation, weighted area‐to‐area regression kriging [WATARK], artificial neural network, random forest [RF], and deep belief network methods) over homogeneous underlying surfaces. Over moderately heterogeneous underlying surfaces, the WATARK method performed better. However, the RF method performed better over highly heterogeneous underlying surfaces. A combined method (using the area‐weighted and WATARK methods for homogeneous and moderately heterogeneous underlying surfaces, respectively, and using the RF method for highly heterogeneous underlying surfaces) was proposed to acquire the daily ground truth ET data at the satellite pixel scale, and the errors in the ground truth ET data were evaluated. The Dual Temperature Difference (DTD) and ETMonitor were validated using ground truth ET data, which solve the problem of the spatial‐scale mismatch and quantify uncertainties in the validation process.

In this study, land surface temperature (LST) and leaf area index (LAI) observations are merged with a coupled two‐source surface energy budget–vegetation dynamic model (TSEB–VDM) via a variational data assimilation (VDA) system to predict turbulent heat fluxes and gross primary productivity (GPP). The TSEB and VDM are coupled by relating photosynthesis in the VDM to transpiration in the TSEB equation. Unknown parameters of the VDA approach are the neutral bulk heat transfer coefficient (C_HN), evaporative fractions for soil and canopy (EF_S and EF_C), and specific leaf area (c_g). The VDA approach is evaluated at six AmeriFlux sites with distinct vegetative and climatic characteristics. The modeled sensible (H) and latent (LE) heat fluxes, and GPP agree well with the corresponding eddy covariance measurements in different environmental conditions. The six‐site average root mean square error (RMSE) of estimated daily H, LE, and GPP is 42.2 W m⁻², 51.5 W m⁻², and 1.8 gC m⁻² d⁻¹, respectively. The outcomes show that the developed VDA approach is able to exploit the implicit information in the sequences of LST and LAI measurements to estimate H, LE, and GPP. Our findings also indicate that the estimates of the H and LE are more sensitive to uncertainties in LST measurements, while the GPP retrievals are more affected by uncertainties in the LAI observations.

Validation of remotely sensed evapotranspiration (RS_ET) products is important because their accuracy is critical for various scientific applications. In this study, an integrated validation framework was proposed for evaluating RS_ET products with coarse spatial resolution extending from homogenous to heterogeneous land surfaces. This framework was applied at the pixel and river basin scales, using direct and indirect validation methods with multisource validation datasets, which solved the spatial mismatch between ground measurements and remotely sensed products. The accuracy, rationality of spatiotemporal variations, and error sources of RS_ET products and uncertainties during the validation process were the focuses in the framework. The application of this framework is exemplified by validating five widely used RS_ET products (i.e., GLEAM, DTD, MOD16, ETMonitor, and GLASS) in the Heihe River Basin from 2012 to 2016. Combined with the results from direct (as the priority method) and indirect validation (as the auxiliary method), DTD showed the highest accuracy (1-MAPE) in the vegetation growing season (75%), followed by ETMonitor (71%), GLASS (68%), GLEAM (54%), and MOD16 (44%). Each product reasonably reflected the spatiotemporal variations in the validation dataset. ETMonitor exhibited the highest consistency with the ground truth ET at the basin scale (ETMap) (R = 0.69), followed by GLASS (0.65), DTD (0.63), MOD16 (0.62), and GLEAM (0.57). Error sources of these RS_ET products were mainly due to the limitations of the algorithms and the coarse spatial resolution of the input data, while the uncertainties in the validation process amounted to 15–28%. This work is proposed to effectively validate and improve the RS_ET products over heterogeneous land surfaces.

Accurate estimation of surface evapotranspiration (ET) in the Heihe River Basin using remote sensing data is crucial for understanding water dynamics in arid regions. In this paper, by coupling physical constraints and machine learning for hybrid modeling, we develop a hybrid model based on surface conductance optimization. A hybrid modeling algorithm, two physical process-based ET algorithms (Penman–Monteith-based and Priestley–Taylor-based ET algorithms), and three pure machine learning algorithms (Random Forest, Extreme Gradient Boosting, and K Nearest Neighbors) are comparatively analyzed for estimating the ET. The results showed that, in general, the machine learning model optimized by parameters was able to better predict the surface conductance of the hybrid model. Driver analyses showed that radiation, normalized difference vegetation index (NDVI), and air temperature had high correlations with ET. The hybrid model had a better prediction performance for ET than the other five models, and it improved the R₂ of the two physical process-based algorithms to 0.9, reduced the root mean square error (RMSE) to 0.5 mm/day, reduced the BIAS to 0.2 mm/day, and improved the Kling–Gupta efficiency (KGE) to 0.9. The hybrid model outperformed the others across different time scales, displaying lower BIAS, RMSE, and higher KGE. Spatially, its ET patterns aligned with regional vegetation changes, with superior accuracy in annual ET estimation compared to the other models. Comparison with other ET products shows that the estimation results based on the hybrid model have better performance. This approach not only improves the accuracy of ET estimation but also improves the understanding of the physical mechanism of ET estimation by pure machine learning models. This study can provide important support for understanding ET and hydrological processes under different climatic and biotic vegetation in other arid and semi-arid regions.

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.