Empirical Estimation of Daily Evaporation from Shallow Groundwater with a Temperature Coefficient

Shallow groundwater evaporation (E_g) is a major component of the hydrological cycle, especially in semiarid and arid locations. Existing E_g estimation processes mainly rely on three approaches: direct measurements, numerical models, and empirical methods. Empirical methods are more commonly used in practical applications due to good performances with more accessible inputs and simple forms. However, most of commonly used empirical methods can only weakly represent E_g variations along the soil depth and do not consider the energy driver. Therefore, a temperature coefficient was proposed and incorporated into two preferred empirical models to characterize the impacts of soil temperature and air temperature lags on E_g. The method was evaluated using in situ daily data obtained from nonweighing bare soil lysimeters. The results indicated that the models that considered the temperature gradient variable (T) conformed to the changes in the actual E_g values with depth more appropriately than the original models, accompanied by 4.3%–8.8% accuracy improvements overall. Shallow groundwater evaporation E_g was found to be influenced by the water table depth (H), T, and pan evaporation (E₀) in descending order, and strong interactions were found between H and T. Moreover, bias of E_g measurement results from precipitation was investigated; measurements from dry days without precipitation revealed the actual E_g process, the relative errors in the cumulative E_g values derived at different depths demonstrated a positive relationship with infiltration recharge, and the errors related to precipitation induced 6.7%–8.3% E_g underestimations. These results contribute to a better understanding of evaporative losses from shallow groundwater and the typical E_g situation that occurs simultaneously with recharge, and they provide promising perspectives for corresponding integrated hydrologic modeling research.