Evaporation of water bodies is significantly controlled by the available energy (net radiation R_n minus heat storage G). Compared to deep lakes or lakes in the high‐latitude or high‐altitude regions, the heat storage effects of shallow lakes in subtropical regions are less well studied. In this paper, we tested the heat storage effect on the latent heat fluxes estimation using the Priestley‐Taylor (PT) equation and numerical water temperature simulations in an ephemeral and shallow lake, Poyang Lake, in China. Results showed that the estimation bias was significantly correlated with heat storage changes (r = 0.97). Overlooking the impacts of the heat storage changes will induce an increase in the RMSE (root‐mean‐square error) of the evaporation estimation from 10.1 to 54.4 W m⁻². Evaporation was overestimated in June and underestimated in August because the daytime R_n tended to be stored in water in June, whereas the heat storage during daytime tended to be released as nighttime evaporation in August, when the vapor gradient between the saturated surface and the air was generally larger than in other months. Water temperature simulations showed that the diurnal changes in the heat storage in the layer that were less than 2 m deep govern G of the total water column. PT estimation of a shallow lake can be improved if the temperature of the top 2 m of water is known. Also, from the modeling perspective, because G was linearly correlated with R_n (r = 0.77), a simple equation (a*R_n + b) that approximates the available energy can improve the evaporation estimate.

The implementation of accurate water balance assessment in glacier basins is essential for the management and sustainable development of water resources in the basins. In this study, a hybrid modeling framework was constructed to enhance runoff prediction and water balance assessment in glacier basins. An improved physical hydrological model (SEGSWAT+) was combined with a machine learning model (ML) to capture the relationship between runoff residuals and water balance components through the Shapley additive explanations (SHAP) method. Based on the enhancement of the runoff fitting results of the existing model, the runoff residuals are decomposed and used to correct the hydrological process component values, thus improving the accuracy of the water balance results. We evaluated the performance and correction results of the method using various ML methods. We analyzed the results for two consecutive periods from 1959 to 2022 for the glacial sub-basins of three tributaries of the Upper Ili River Basin in central Asia. The results show that the hybrid framework based on extreme gradient boosting (XGBoost) with an average NSE value of 0.93 has the best performance, and the bias based on the evapotranspiration component and soil water content change component is reduced by 3.2–5%, proving the effectiveness of the water balance correction. This study advances the interpretation of ML models for hydrologic assessment of areas with complex hydrodynamic characteristics.

Surface heat fluxes are crucial to the energy and water exchanges between surface and the atmosphere. Hysteresis effect poses a challenge in model simulation of evaporation (latent heat flux, LE) at hourly or intradaily scales. The hidden mechanisms behind the hysteresis remain unclear, given the diverse reports in various climates, geographical locations, and underlying surfaces. This study took advantage of eddy covariance flux observation over Poyang Lake of China, an ephemeral lake experiencing seasonal shifts from land to water. It depicted the hysteresis behavior of LE to environmental forces for the period from May to July in 2018, a hydrologic transition period from land to water surface. Our results showed obvious clockwise hysteresis loops between LE and water vapor pressure deficit (VPD), surface temperature (Ts), and air temperature (Ta). Counterclockwise loops exhibited between LE and surface net radiation (Rn), wind speed (WS), and temperature difference (Ts‐Ta). Rn, WS, and Ts‐Ta dominants control the hysteresis effects and regulate LE variations. The hysteresis loops changed in direction, shape, and area when the surface shifted from land to water. It was primarily related to the lengthened lag time between LE and other variables except for WS, which was attributable to enhanced impacts from WS and VPD, and weakened impacts from Rn and Ts‐Ta. The findings should help to improve our understanding of the complicated water‐atmosphere interactions at the intradaily scale.