A synergistic integration of physics-based and data-driven approaches has emerged as promising research field for terrestrial evapotranspiration (ET) estimation, enabling robust modeling of land-atmosphere interactions. This study proposes a hybrid model by integrating machine learning (ML)-based canopy surface resistance (r_s,c) estimation into the Shuttleworth-Wallace (S-W) dual-source scheme under the ETMonitor framework, replacing traditional physics-based r_s,c parameterization. Three ML algorithms, Random Forest (RF), Gradient Boosting Regression Tree (GBRT) and Deep Neural Network (DNN) were tested in the hybrid model. A reference dataset of r_s,c was derived by inverting S-W dual-source model with in-situ flux measurements. The model was trained on 179 global flux tower sites and independently validated on 45 sites. Three full ML-based models based on DNN, GBRT and RF, were also developed to estimate ET directly for comparison. The DNN-integrated hybrid model outperformed the original physics-based model, with Kling-Gupta Efficiency (KGE) increasing from 0.7 to 0.84 and coefficient of determination (R²) increasing from 0.66 to 0.72. The three full ML models showed comparable performance to the hybrid models. Notably, the physics-ML hybrid framework balances physical interpretability with data-driven efficiency, minimizing reliance on prior knowledge and avoiding over-parameterization.

Water volume, a fundamental characteristic of lakes, serves as a crucial indicator for understanding regional climate, ecological systems, and hydrological processes. However, limitations in existing estimation methods and datasets for water depth, such as the insufficient observation of small and medium-sized lakes and unclear temporal information, have hindered a comprehensive understanding of global lake water volumes. To address these challenges, this study develops a machine learning (ML)-based approach to estimate the dynamic water depths of global lakes. By incorporating various lake features and employing multiple innovative water depth extraction methods, we generated an extensive water depth dataset to train the model. Validation results demonstrate the model’s high accuracy, with the bias of −0.08 m, a MAE of 1.09 m, an RMSE of 4.78 m, and an R² of 0.95. The proposed method provides dynamic monthly estimates of global lake water depths and volumes in 2000~2020. This study offers a cost-effective and efficient solution for estimating global lake water dynamics, providing reliable data to support the monitoring, analysis, and management of regional and global lake systems.

Evaluating the performance of irrigation water use is essential for efficient and sustainable water resource management. However, existing approaches often lack systematic quantification of irrigation water consumption and fail to differentiate between the use of precipitation and anthropogenic appropriation of water flows. Building on the green–blue water concept, consumptive water use, assumed equal to actual evapotranspiration (ET_a), was partitioned into green ET (GET) and blue ET (BET) using remote sensing data and the Budyko hypothesis. A novel BET metric of consumptive irrigation water use was developed and applied to the irrigated lands in northwest China to evaluate the performance of irrigation from 2001 to 2021. The results showed that in terms of total available water resources (precipitation + gross irrigation water (GIW)) compared to irrigation water demand, estimated as reference evapotranspiration (ET₀), Ningxia has sufficient water supply to meet irrigation demand, while the Hexi Corridor faces increasing risks of unsustainable water use. The Hetao irrigation scheme has shifted from a fragile supply–demand balance to a situation where water demand far exceeds availability. In Xinjiang, the balance between water supply and demand is tight. Furthermore, when considering the available water (GIW) relative to the net irrigation water demand (ET₀-GET), the Hexi Corridor faces significant water deficits, and Ningxia and Xinjiang are close to meeting local irrigation water demands by relying on current water availability and efficient irrigation practices. It is noteworthy that the BET remains lower than the GIW in northwest China (excluding the Hexi Corridor in recent years). The ratio of the BET to GIW is an estimate of the scheme irrigation efficiency, which was equal to 0.54 for all irrigation schemes taken together. In addition, the irrigation water use efficiency, estimated as the ratio of BET to net irrigation water, was evaluated in detail, and it was found that in the last 10 years the irrigation water use efficiency improved in Ningxia, the Hetao irrigation scheme, and Xinjiang. However, the Hexi Corridor continues to face severe net irrigation water deficits, suggesting the likelihood of groundwater use to sustain irrigated agriculture. BET innovatively separates consumptive use of precipitation (green water) and consumptive use of irrigation (blue water), a critical advancement beyond conventional approaches’ estimates that merge these distinct hydrological components to help quantifying water use efficiency.

The Advanced Microwave Scanning Radiometer-Earth Observing System/National Aeronautics and Space Administration (AMSR-E/NASA) daily global soil moisture (SM) product (2002–2011, 25-km resolution) has been widely used but exhibits limited sensitivity to intra-annual and interannual variability in many regions. This limitation is mainly attributed to inaccurate parameter values (A₀ and A₁), which account for vegetation transmittance and emission in the AMSR-E/NASA SM retrieval algorithm. To address this issue, we recalibrated A₀ and A₁ using in situ SM measurements from 13 observation networks (192 sites) and established their empirical relationships with fractional vegetation cover (FVC). Four dominant land cover types (i.e., bare soil, grassland, cropland, and forest) were considered due to their global representativeness and extensive coverage with in situ SM measurements. Based on these relationships, we generated global maps of A₀ and A₁ and produced an improved Global Daily AMSR-E SM dataset (GD_AMSR-E_SM; 2002–2011, 25-km resolution) using the Global Land Surface Satellite (GLASS) FVC dataset and AMSR-E observations. Validation against in situ SM measurements within six independent networks shows that the GD_AMSR-E_SM dataset achieves greater consistency with in situ SM measurements, with mean absolute error (MAE) and root-mean-square error (RMSE) values of 0.026 and 0.032 cm³/cm³, respectively. This represents average reductions of 20% and 26% compared with the AMSR-E/NASA, AMSR-E/Japan Aerospace Exploration Agency (JAXA), and AMSR-E/Land Parameter Retrieval Model (LPRM) SM products. The enhanced algorithm improves the accuracy and reliability of AMSR-E observations for long-term global SM monitoring.

Water depth, a fundamental characteristic of a lake, is important for understanding climatic, ecological, and hydrological processes. However, lake water depth data are still scarce due to the high cost of in-situ measurements and the limitations of remote sensing observations. In this study, a novel method was developed to estimate time series of pixel-wise water depths of lakes that have ever exposed their bottom by remote sensing observations. Lake water depths were calculated as the difference between the elevations of the dynamic water surface and the historical lakebed elevations using optical images and DEM data. The method was applied in the Sahel-Sudano-Guinean region of Africa where complex climatic conditions and rare in-situ measurements. Experiments showed that the proposed method could get consistent water depths compared with the HydroLAKES data, i.e. with a MAE of 0.86 m and a RMSE of 1.69 m, and water surface elevations similar to the estimates derived from ICESat/ICESat-2 measurements with a MAE of 3.79 m and a RMSE of 5.92 m. The method can provide pixel-wise information on lake water depth at high temporal frequency, and is expected to provide an efficient solution to gather essential information on lakes.

Accurate knowledge of the at-surface solar irradiance (SSI) is essential for retrieving surface and atmospheric properties using satellite measurements of backscattered and reflected radiance. The latter is affected by surface-atmosphere interactions, including the effects of terrain. The SSI is affected by the same processes. This study proposes a method to estimate the components of instantaneous SSI: direct, isotropic and circumsolar diffuse, and terrain irradiance, which is expected to improve the simultaneous retrieval of aerosol optical depth (AOD) and surface reflectance. The method takes into account the coupled effects of topography and atmosphere by combining parameterization and the lookup table (LUT) approaches. The method was applied to rugged terrain over the Tibetan Plateau using Moderate Resolution Imaging Spectrometer (MODIS) atmosphere and surface data, the fifth generation European Centre for Medium-Range Weather Forecasts reanalysis (ERA5) data, Cloud-Aerosol Lidar With Orthogonal Polarization (CALIOP) aerosol data, and a digital elevation model (DEM). The results showed that the SSI estimates were in satisfactory agreement with ground observations at four stations over the Tibetan Plateau (TP) in 2018 with R2 values of 0.61, 0.44, 0.41, and 0.49, respectively, and root mean square error (RMSE) of 205.7, 176.9, 186.0, and 201.2 W/m2, respectively. Estimations of the diffuse irradiance were evaluated separately against the only available in situ observations at the Dali Station, and the results were better than our SSI estimates with R2, RMSE, and relative bias (BIAS) being 0.71, 94.98 W/m2, and 31%, respectively. The isotropic and circumsolar diffuse irradiances accounted for 37.57% and 7.68% of the total annual SSI, respectively, while diffuse irradiance accounted for 46.48% of the total annual SSI. Under clear skies, every 0.1 increase in AOD caused about a 35-W/m2 increase in diffuse irradiance and a decrease of about 25 W/m2 of SSI.

Global-scale surface soil moisture (SSM) products (e.g. SMAP L3.0, ASCAT V3.0, ESA/CCI V7.1 and GLDAS V2.2) are vital for applications in hydrology, climate variability, and agriculture. This study uses a new SSM evaluation approach by combining temporal evolution, Coefficient of Variation (CV), Cumulative Distribution Function (CDF), evaluation metrics, and Triple Collocation Analysis (TCA) to assess SSM accuracy and spatial–temporal variability, particularly the impact of footprint mismatch when comparing retrieved SSM with in-situ measurements. Results revealed significant spatial variability and seasonal patterns in SSM, as indicated by the CV values and temporal evaluations at different resampling scales. The variability captured by in-situ measurements was comparable to that of SSM products. The impact of footprint mismatch between in-situ measurements and data products, particularly for SMAP and ASCAT SSM, was more significant and led to substantial differences in evaluation metrics between smaller and larger spatial scales. TCA alone cannot reliably assess the accuracy of global-scale SSM products without in-situ SSM measurements. Overall, our findings highlight the critical role of footprint mismatch on the estimated accuracy of SSM products and underscore the need to combine multiple evaluations into an overall scoring indicator, as proposed in this study.

Accurate and continuous estimation of surface albedo is vital for assessing and understanding land–surface–atmosphere interactions. We developed a method for estimating instantaneous all-sky at-surface shortwave upwelling radiance and albedo over the Tibetan Plateau. The method accounts for the complex interplay of topography and atmospheric interactions and aims to mitigate the occurrence of data gaps. Employing an RTLSR-kernel-driven model, we retrieved surface shortwave albedo with a 1 km resolution, incorporating direct, isotropic diffuse; circumsolar diffuse; and surrounding terrain irradiance into the all-sky solar surface irradiance. The at-surface upwelling radiance and surface shortwave albedo estimates were in satisfactory agreement with ground observations at four stations in the Tibetan Plateau, with RMSE values of 56.5 W/m² and 0.0422, 67.6 W/m² and 0.0545, 98.6 W/m² and 0.0992, and 78.0 98.6 W/m² and 0.639. This comparison indicated an improved accuracy of at-surface upwelling radiance and surface albedo and significantly reduced data gaps. Valid observations increased substantially in comparison to the MCD43A2 data product, with the new method achieving an increase ranging from 40% to 200% at the four stations. Our study demonstrates that by integrating terrain, cloud properties, and radiative transfer modeling, the accuracy and completeness of retrieved surface albedo and radiance in complex terrains can be effectively improved.

Improving irrigation water management is a key concern for the agricultural sector, and it requires extensive and comprehensive tools that provide a complete knowledge of crop water use and requirements. This study presents a novel methodology to explicitly estimate daily gross and net crop water requirements, actual crop water use, and irrigation efficiency of center pivot irrigation systems, by mainly utilizing the Sentinel-2 MultiSpectral Instrument (MSI) imagery at the farm scale. ETMonitor model is adapted to estimate actual water use (as the sum of canopy transpiration and evaporation of water intercepted by canopy and evaporation from soil) at daily/10-m resolution, benefiting from the high-resolution Sentinel-2 data and thus to assess the irrigation efficiency at the farm scale. The gross irrigation water requirement is estimated from the net crop water requirement and the water loss, including the water droplet evaporation directly into the air during application before droplets fall on the canopy and canopy interception loss. The method was applied to a pilot farmland with two major crops (wheat and potato) in the Inner Mongolia Autonomous Region of China, where modern equipment and appropriate irrigation methods are deployed for efficient water use. The estimated actual crop water use showed good agreement with the ground observations, e.g. the determination coefficients range from 0.67 to 0.81 and root mean square errors range from 0.56 mm/day to 1.24 mm/day for wheat and potato when comparing the estimated evapotranspiration with the measurement by the eddy covariance system. It also showed that the losses of total irrigated volume were 25.4% for wheat and 23.7% for potato, respectively, and found that the water allocation was insufficient to meet the water requirement in this irrigated area. This suggests that the amount of water applied was insufficient to meet the crop water requirement and the inherent water losses in the center pivot irrigation system, which imply the necessity to improve the irrigation practice to use the water more efficiently.

Information on crop yield is important for food security, in particular under the conditions of climate change and growing population worldwide. We developed a new fully distributed, high spatial resolution, model of biomass accumulation and crop yield applicable to a highly heterogeneous desert-oasis agroecosystem. The bulk of required input data is obtained by retrieving pixel-wise biogeophysical variables from a suite of very diverse satellite data. Both temperature and water stress conditions at field-scale are given full consideration, while the model was designed to strike a balance between model applicability and satisfactory characterization of the heterogeneous desert-oasis system to estimate field-scale yield. The development of this model relies on three main innovations. First, the start and end of the growing season were estimated for each pixel by calibrating the high spatial and temporal resolution observations of Normalized Difference Vegetation Index (NDVI) by Sentinal-2 (S2) MSI (Multi-Spectral Instrument) against limited local phenological information. Second, to monitor crop water stress, account taken of irrigation, a process-based water and energy balance model was applied to estimate the actual evapotranspiration (ET). This requires knowledge of soil water availability, which is characterized by downscaling the ASCAT (Advanced SCATterrometer) soil moisture data product. To capture the dominant features of the eco-hydrological conditions in the desert and oasis agroecosystem, ET was further downscaled from the 1 km resolution. Third, likewise the water stress indicator, the air temperature stress indicator was mapped after characterizing the thermal contrast and heterogeneity of the desert-oasis system, by generating time series of air temperature at 1 km spatial resolution using the MODIS (Moderate Resolution Imaging Spectroradiometer) Land Surface Temperature (LST) data product. In the temporal dimension, gaps were mitigated by applying time series analysis techniques to reconstruct cloud-free time series of LST, NDVI, fAPAR and albedo. These innovations add up to a high resolution characterization of crop response to the geospatial variability of weather and climate forcing in the desert-oasis agroecosystem. The model was applied to the dominant crops, i.e., spring wheat, maize, sunflower, and melon, in the oases of the Shiyang River Basin (northwestern China) characterized by a rather fragmented land use. The high resolution of pixel-wise ecohydrological parameters, i.e., crop phenology, temperature stress and water stress factors successfully reflect differences of crops with different phenology and location in the oases. The relative errors for wheat and maize yields compared to the census data are less than 5% at district level. At the county level, the relative errors of wheat yields of Liangzhou, Minqin, Gulang, Jinchuan, and Yongchang equal to 0.87%, 24.2%, 9.7%, 12.5%, and 7.2%. For maize, the dominant crop, the error on estimated yields was less than 5%, except in Gulang. The relative error on estimated yield for sunflower was less than 10% compared to agricultural census data. The relative error on estimated melon yield was 16%. This performance highlights the applicability of the model to estimate field-scale yields in agroecosystems characterized by fragmented land use.

The ASCAT (Advanced SCATterometer) soil moisture product with 10-km spatial resolution was retrieved based on the soil water index (SWI) algorithm from the data acquired by the scatterometer on board the Meteorological OPerational (MetOP) satellites (MetOP-A, MetOP-B). In this study, the ASCAT product was downscaled from 10-km to 1-km spatial resolution based on the Apparent Thermal Inertia (ATI) estimated from MODIS Land Surface Temperature (LST) and Albedo retrievals in 54 grids (1 degree 1 degree) around 54 FLUXNET stations. First, the ATI was estimated at 1-km spatial resolution by using MODIS LST and Albedo data at the same spatial resolution and then resampled to 10-km. Second, the relationship between ASCAT soil moisture and ATI at 10-km spatial resolution was established. Finally, the spatiotemporally continuous soil moisture at 1-km spatial resolution was retrieved using the obtained relationship between ATI and ASCAT at 10-km spatial resolution, and the ATI data at 1-km spatial resolution. However, there were many missing values in the MODIS LST maps leading to spatiotemporal discontinuity in LST and calculated ATI data. To obtain spatiotemporal continuous ATI data, this study first reconstructed the MODIS LST data by finding similar points that had the same land cover type and similar NDVI (the Normalized Difference Vegetation Index) value. In this study, we found that the LST data of similar points in a pair of temporal adjacent LST images had a linear relationship. The LST data of these similar points in a pair of temporal adjacent LST images were used to establish a linear relationship and then used to reconstruct the pair of temporally adjacent LST images. The reconstructed LST data were used to obtain the spatiotemporal continuous ATI data at 1-km and 10-km spatial resolutions. In this study, downscaled 1-km spatial resolution soil moisture product within the 54 grids around the FLUXNET sites were obtained in 2013. Results indicated that the spatial distribution of the downscaled soil moisture using the reconstructed MODIS LST data is better than that using original MODIS LST data. Additionally, the downscaled soil moisture was evaluated against in-situ soil moisture measurements at 54 FLUXNET stations. The average of RMSE (the Root Mean Square Error) was 0.098 m³m^-3 and the average of MAE (the Mean Absolute Error) was 0.08 m³m^-3.

Observing and understanding changes in Africa is a hotspot in global ecological environmental research since the early 1970s. As possible causes of environmental degradation, frequent droughts and human activities attracted wide attention. Remote sensing of nighttime light provides an effective way to map human activities and assess their intensity. To identify settlements more effectively, this study focused on nighttime light in the northern Equatorial Africa and Sahel settlements to propose a new method, namely, the patches filtering method (PFM) to identify nighttime lights related to settlements from the National Polar-orbiting Partnership Visible Infrared Imaging Radiometer Suite (NPP-VIIRS) monthly nighttime light data by separating signal components induced by biomass burning, thereby generating a new annual image in 2016. The results show that PFM is useful for improving the quality of NPP-VIIRS monthly nighttime light data. Settlement lights were effectively separated from biomass burning lights, in addition to capturing the seasonality of biomass burning. We show that the new 2016 nighttime light image can very effectively identify even small settlements, notwithstanding their fragmentation and unstable power supply. We compared the image with earlier NPP-VIIRS annual nighttime light data from the National Oceanic and Atmospheric Administration (NOAA) National Center for Environmental Information (NCEI) for 2016 and the Sentinel-2 prototype Land Cover 20 m 2016 map of Africa released by the European Space Agency (ESA-S2-AFRICA-LC20). We found that the new annual nighttime light data performed best among the three datasets in capturing settlements, with a high recognition rate of 61.8%, and absolute superiority for settlements of 2.5 square kilometers or less. This shows that the method separates biomass burning signals very effectively, while retaining the relatively stable, although dim, lights of small settlements. The new 2016 annual image demonstrates good performance in identifying human settlements in sparsely populated areas toward a better understanding of human activities.

Land Surface Models which determine evapotranspiration (ET) by neglecting the sub-grid heterogeneity of land-atmosphere parameters will cause aggregation biases in spatially-averaged ET estimates, considering the nonlinear dependences of ET on the heterogeneous land-atmosphere parameters. One frequently adopted strategy clusters the heterogeneous surface within a model grid into several tiles, assumed to be homogeneous, usually based on high-resolution land cover data. While the differences in bulk-averaged parameters between different tiles are considered, the heterogeneity within each tile is neglected. This study evaluated in detail the aggregation biases in the tile mean ET estimates due to applying bulk-averages of the Saturation degree of surface Soil Moisture (SSM) and Leaf Area Index (LAI) for each tile through numerical experiments. Four types of Probability Distribution Function (PDF) were used to simulate different scenarios on the heterogeneity (within a tile) of SSM and LAI, i.e., from water scarcity to wet, and from sparse to dense vegetation covered surfaces. Aggregation bias was calculated by comparing ET estimates based on bulk-averaged SSM and LAI with the one obtained by aggregation of the flux estimates based on the PDFs, which complies with energy conservation. In addition, a wide range of meteorological conditions was applied in our numerical experiments and the impacts were evaluated by binning results according to the reference ET (ET ₀ ). We found that potentially significant bias can be found in semi-arid areas. Neglecting the actual spatial variability of both SSM and LAI within tiles can lead to both large relative error (> 20%) and absolute error (> 1 mm/day) in the estimated ET. A negative bias is expected at low ET / ET ₀ and a positive bias is expected at large ET / ET ₀ , regardless of climate conditions (i.e., ET ₀ ).