An accurate estimation of spatially and temporally continuous latent heat flux (LE) is essential in the assessment of surface water and energy balance. Various satellite-derived LE products have been generated to enhance the simulation of terrestrial LE, yet each individual LE product shows large discrepancies and uncertainties. Our study used Extremely Randomized Trees (ETR) to fuse five satellite-derived terrestrial LE products to reduce uncertainties from the individual products and improve terrestrial LE estimations over Europe. The validation results demonstrated that the estimation using the ETR fusion method increased the R² of five individual LE products (ranging from 0.53 to 0.61) to 0.97 and decreased the RMSE (ranging from 26.37 to 33.17 W/m²) to 5.85 W/m². Compared with three other machine learning fusion models, Gradient Boosting Regression Tree (GBRT), Random Forest (RF), and Gaussian Process Regression (GPR), ETR exhibited the best performance in terms of both training and validation accuracy. We also applied the ETR fusion method to implement the mapping of average annual terrestrial LE over Europe at a resolution of 0.05 ◦ in the period from 2002 to 2005. When compared with global LE products such as the Global Land Surface Satellite (GLASS) and the Moderate Resolution Imaging Spectroradiometer (MODIS), the fusion LE using ETR exhibited a relatively small gap, which confirmed that it is reasonable and reliable for the estimation of the terrestrial LE over Europe.

The methods for accurately fusing medium- and high-spatial-resolution satellite reflectance are vital for monitoring vegetation biomass, agricultural irrigation, ecological processes and climate change. However, the currently existing fusion methods cannot accurately capture the temporal variation in reflectance for heterogeneous landscapes. In this study, we proposed a new method, the spatial and temporal reflectance fusion method based on the unmixing theory and a fuzzy C-clustering model (FCMSTRFM), to generate Landsat-like time-series surface reflectance. Unlike other data fusion models, the FCMSTRFM improved the similarity of pixels grouped together by combining land cover maps and time-series data cluster algorithms to define endmembers. The proposed method was tested over a 2000 km² study area in Heilongjiang Provence, China, in 2017 and 2018 using ten images. The results show that the accuracy of the FCMSTRFM is better than that of the popular enhanced spatial and temporal adaptive reflectance fusion model (ESTARFM) (correlation coefficient (R): 0.8413 vs. 0.7589; root mean square error (RMSE): 0.0267 vs. 0.0401) and the spatial-temporal data fusion approach (STDFA) (R: 0.8413 vs. 0.7666; RMSE: 0.0267 vs. 0.0307). Importantly, the FCMSTRFM was able to maintain the details of temporal variations in complicated landscapes. The proposed method provides an alternative method to monitor the dynamics of land surface variables over complicated heterogeneous regions.

The long-term estimation of grassland aboveground biomass (AGB) is important for grassland resource management in the Three-River Headwaters Region (TRHR) of China. Due to the lack of reliable grassland AGB datasets since the 1980s, the long-term spatiotemporal variation in grassland AGB in the TRHR remains unclear. In this study, we estimated AGB in the grassland of 209,897 km² using advanced very high resolution radiometer (AVHRR), MODerate-resolution Imaging Spectroradiometer (MODIS), meteorological, ancillary data during 1982–2018, and 75 AGB ground observations in the growth period of 2009 in the TRHR. To enhance the spatial representativeness of ground observations, we firstly upscaled the grassland AGB using a gradient boosting regression tree (GBRT) model from ground observations to a 1 km spatial resolution via MODIS normalized difference vegetation index (NDVI), meteorological and ancillary data, and the model produced validation results with a coefficient of determination (R²) equal to 0.76, a relative mean square error (RMSE) equal to 88.8 g C m⁻², and a bias equal to −1.6 g C m⁻² between the ground-observed and MODIS-derived upscaled AGB. Then, we upscaled grassland AGB using the same model from a 1 km to 5 km spatial resolution via AVHRR NDVI and the same data as previously mentioned with the validation accuracy (R² = 0.74, RMSE = 57.8 g C m⁻², and bias = −0.1 g C m⁻²) between the MODIS-derived reference and AVHRR-derived upscaled AGB. The annual trend of grassland AGB in the TRHR increased by 0.37 g C m⁻² (p < 0.05) on average per year during 1982–2018, which was mainly caused by vegetation greening and increased precipitation. This study provided reliable long-term (1982–2018) grassland AGB datasets to monitor the spatiotemporal variation in grassland AGB in the TRHR.

Surface all-wave net radiation (Rn) is a crucial variable driving many terrestrial latent heat (LE) models that estimate global LE. However, the differences between different Rn products and their impact on global LE estimates still remain unclear. In this study, we evaluated two Rn products, Global LAnd Surface Satellite (GLASS) beta version Rn and Modern-Era Retrospective Analysis for Research and Applications-version 2 (MERRA-2) Rn, from 2007–2017 using ground-measured data from 240 globally distributed in-situ radiation measurements provided by FLUXNET projects. The GLASS Rn product had higher accuracy (R² increased by 0.04–0.26, and RMSE decreased by 2–13.3 W/m²) than the MERRA-2 Rn product for all land cover types on a daily scale, and the two Rn products differed greatly in spatial distribution and variations. We then determined the resulting discrepancies in simulated annual global LE using a simple averaging model by merging five diagnostic LE models: RS-PM model, SW model, PT-JPL model, MS-PT model, and SIM model. The validation results showed that the estimated LE from the GLASS Rn had higher accuracy (R² increased by 0.04–0.14, and RMSE decreased by 3–8.4 W/m²) than that from the MERRA-2 Rn for different land cover types at daily scale. Importantly, the mean annual global terrestrial LE from GLASS Rn was 2.1% lower than that from the MERRA-2 Rn. Our study showed that large differences in satellite and reanalysis Rn products could lead to substantial uncertainties in estimating global terrestrial LE.

Ocean latent heat flux (LHF) is an essential variable for air–sea interactions, which establishes the link between energy balance, water and carbon cycle. The low-latitude ocean is the main heat source of the global ocean and has a great influence on global climate change and energy transmission. Thus, an accuracy estimation of high-resolution ocean LHF over low-latitude area is vital to the understanding of energy and water cycle, and it remains a challenge. To reduce the uncertainties of individual LHF products over low-latitude areas, four machine learning (ML) methods (Artificial Neutral Network (ANN), Random forest (RF), Bayesian Ridge regression and Random Sample Consensus (RANSAC) regression) were applied to estimate low-latitude monthly ocean LHF by using two satellite products (JOFURO-3 and GSSTF-3) and two reanalysis products (MERRA-2 and ERA-I). We validated the estimated ocean LHF using 115 widely distributed buoy sites from three buoy site arrays (TAO, PIRATA and RAMA). The validation results demonstrate that the performance of LHF estimations derived from the ML methods (including ANN, RF, BR and RANSAC) were significantly better than individual LHF products, indicated by R² increasing by 3.7–46.4%. Among them, the LHF estimation using the ANN method increased the R² of the four-individual ocean LHF products (ranging from 0.56 to 0.79) to 0.88 and decreased the RMSE (ranging from 19.1 to 37.5) to 11 W m⁻². Compared to three other ML methods (RF, BR and RANSAC), ANN method exhibited the best performance according to the validation results. The results of relative uncertainty analysis using the triangle cornered hat (TCH) method show that the ensemble LHF product using ML methods has lower relative uncertainty than individual LHF product in most area. The ANN was employed to implement the mapping of annual average ocean LHF over low-latitude at a spatial resolution of 0.25° during 2003–2007. The ocean LHF fusion products estimated from ANN methods were 10–30 W m⁻² lower than those of the four original ocean products (MERRA-2, JOFURO-3, ERA-I and GSSTF-3) and were more similar to observations.

Reliable estimates of terrestrial latent heat flux (LE) at high spatial and temporal resolutions are of vital importance for energy balance and water resource management. However, currently available LE products derived from satellite data generally have high revisit frequency or fine spatial resolution. In this study, we explored the feasibility of the high spatiotemporal resolution LE fusion framework to take advantage of the Moderate Resolution Imaging Spectroradiometer (MODIS) and Chinese GaoFen-1 Wide Field View (GF-1 WFV) data. In particular, three-fold fusion schemes based on Enhanced Spatial and Temporal Adaptive Reflectance Fusion Model (ESTARFM) were employed, including fusion of surface reflectance (Scheme 1), vegetation indices (Scheme 2) and high order LE products (Scheme 3). Our results showed that the fusion of vegetation indices and further computing LE (Scheme 2) achieved better accuracy and captured more detailed information of terrestrial LE, where the determination coefficient (R²) varies from 0.86 to 0.98, the root-mean-square error (RMSE) ranges from 1.25 to 9.77 W/m² and the relative RSME (rRMSE) varies from 2% to 23%. The time series of merged LE in 2017 using the optimal Scheme 2 also showed a relatively good agreement with eddy covariance (EC) measurements and MODIS LE products. The fusion approach provides spatiotemporal continuous LE estimates and also reduces the uncertainties in LE estimation, with an increment in R² by 0.06 and a decrease in RMSE by 23.4% on average. The proposed high spatiotemporal resolution LE estimation framework using multi-source data showed great promise in monitoring LE variation at field scale, and may have value in planning irrigation schemes and providing water management decisions over agroecosystems.