Comparison and modification of methods for estimating evapotranspiration using diurnal groundwater level fluctuations in arid and semiarid regions

In arid and semiarid regions, vegetation growth largely depends on groundwater, and causes diurnal fluctuations of shallow groundwater levels. Diurnal groundwater level fluctuations have been widely used to estimate groundwater evapotranspiration (ETG) in several methods. This study compared ETG estimated by three commonly used methods. A groundwater flow model was created to generate synthetic diurnal groundwater level fluctuations caused by a given evapotranspiration. The model also calculates the change in groundwater storage and net groundwater inflow at locations of observation wells. The White method, the Hays method, and the Loheide method were applied to estimate ETG with the model-generated diurnal groundwater levels. The comparison of the actual and estimated ETG revealed the accuracy of each method and indentified the applicability of the methods. When the recovery limb of the groundwater level hydrograph is nonlinear, these existing methods underestimate daily ETG. The Loheide method is comparatively better and can be improved by representing the rate of water table increase in the recovery limb of the hydrograph using an exponential equation. When the recovery limb of the groundwater level hydrograph is linear, all three methods can accurately estimate the daily ETG. The modified White method can provide hourly ETG estimates and is recommended for general use. In practical applications, the analysis of the shape of the water table recovery limb and the up and down gradient groundwater head differences can be used to identify the proper method for estimating ETG.