A Global Implementation of Single‐ and Dual‐Source Surface Energy Balance Models for Estimating Actual Evapotranspiration at 30‐m Resolution Using Google Earth Engine

Abstract

Evapotranspiration (ET) provides a robust connection between hydrological cycles and surface energy balance. Accurate and near‐daily ET estimation has utility in water resources, agricultural management applications, crop yields and drought monitoring. This study describes the implementation of an ET modeling system based on a Priestley‐Taylor version of the Two‐Source (soil and vegetation) Energy Balance Model (TSEB‐PT) within Google Earth Engine environment. TSEB‐PT performance was compared with the simpler single‐source HSEB (Hybrid Surface Energy Balance) ET model to assess relative advantages and disadvantages for operational application. Results were evaluated across multiple biomes and climatic zones across the US, Europe, and Australia in comparison with eddy covariance data from 30 flux tower sites. Both models produced similar results when considering all biomes at daily, weekly, and monthly timescales. Daily ET metrics for all sites combined yielded comparable results for both models, with a slightly lower root‐mean‐square error for TSEB‐PT (HSEB) of 1.2 (1.3) mm/d and a higher correlation (r) of 0.83 (0.80), but a larger mean percent bias error (MPBE = −9%) than HSEB (MPBE = 1%). TSEB‐PT performance was lowest for sites in warm summer humid continental and hot semi‐arid climates and in evergreen broadleaf forest cover, while HSEB showed lowest performance in tropical savanna hot semi‐arid climates and in savanna covers. Model performance was improved for both cropland and non‐cropland sites when TSEB‐PT and HSEB ET estimates were combined through simple averaging due to cancellation of opposing errors, showing a promise as potential tools for water resource management on a global scale.

Improved 30‐m Evapotranspiration Estimates Over 145 Eddy Covariance Sites in the Contiguous United States: The Role of ECOSTRESS, Harmonized Landsat Sentinel‐2 Imagery, Climate Reanalysis, and Deep Neural Network Postprocessing

“…For example, Jaafar et al. (2022a, 2022b) evaluated ECOSTRESS ET over 29 EC sites, whereas Kohli et al. (2020) used 5 CIMIS sites in their evaluation.…”

“…Based on this principle, different approaches have been developed for estimating spatial ET over time using remote sensing observations, particularly from spaceborne sensors (J. M. Chen & Liu, 2020), such as the Surface Energy Balance Algorithm for Land (SEBAL) (Bastiaanssen et al., 1998), Mapping Evapotranspiration at High Resolution with Internalized Calibration (METRIC) (R. G. Allen et al., 2007), Priestley‐Taylor Jet Propulsion Laboratory (PT‐JPL) (Fisher et al., 2008), and Atmosphere‐Land Exchange Inverse model/ALEXI disaggregation (ALEXI/DisALEXI) (Martha C Anderson et al., 2012; Martha C. Anderson et al., 2007). Correspondingly, there have been an increasing number of remote sensing based ET being developed during the recent decades (Mohan et al., 2020), such as Moderate Resolution Imaging Spectroradiometer (MODIS) (Mu et al., 2007, 2011a), the Global Land Evaporation Amsterdam Model (GLEAM) (Martens et al., 2017), the Global Land Surface Satellite (GLASS) (Xie et al., 2022), Landsat (R. Allen et al., 2011; R. G. Allen et al., 2007; Martha C Anderson et al., 2012), the hybrid single‐source energy balance model (HSEB) (Jaafar et al., 2022a), and TSEB‐PT (Jaafar et al., 2022b). While remote sensing‐based ET has advantages over site‐specific ET measurements, that is, the estimation of ET at a specific location or site (Jiménez et al., 2011; McShane et al., 2017), to achieve its full potential, the high spatial and temporal resolutions need to be simultaneously addressed (Fisher et al., 2017; Wen et al., 2022).…”

“…Correspondingly, there have been an increasing number of remote sensing based ET being developed during the recent decades (Mohan et al, 2020), such as Moderate Resolution Imaging Spectroradiometer (MODIS) (Mu et al, 2007(Mu et al, , 2011a, the Global Land Evaporation Amsterdam Model (GLEAM) (Martens et al, 2017), the Global Land Surface Satellite (GLASS) (Xie et al, 2022), Landsat (R. Allen et al, 2011;R. G. Allen et al, 2007;Martha C Anderson et al, 2012), the hybrid single-source energy balance model (HSEB) (Jaafar et al, 2022a), and TSEB-PT (Jaafar et al, 2022b). While remote sensing-based ET has advantages over site-specific ET measurements, that is, the estimation of ET at a specific location or site (Jiménez et al, 2011;McShane et al, 2017), to achieve its full potential, the high spatial and temporal resolutions need to be simultaneously addressed (Fisher et al, 2017;Wen et al, 2022).…”

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

“…They reported an RMSE ≈ 1.40 and 1.24 mm•day −1 for crop and non-crop land covers, respectively, which are higher than those obtained for the same categories in this study of about 0.90 and 1.60 mm•day −1 , respectively. Jaafar et al [29] results were also compared with those of a Two-Source (soil and vegetation) Energy Balance Model (TSEB) in [78] which was given the comparable performance with slightly better metrics in favor of TSEB (RMSE ≈ 1.30 mm•day −1 ). The difference between the performances of [29,78], and our study can be attributed to different SEB models natures and their sensitivities to reanalysis meteorological data.…”

“…Jaafar et al [29] results were also compared with those of a Two-Source (soil and vegetation) Energy Balance Model (TSEB) in [78] which was given the comparable performance with slightly better metrics in favor of TSEB (RMSE ≈ 1.30 mm•day −1 ). The difference between the performances of [29,78], and our study can be attributed to different SEB models natures and their sensitivities to reanalysis meteorological data. Moreover, using the morphological operators can be effective in the performance.…”