Revisiting the White method for estimating groundwater evapotranspiration: a consideration of sunset and sunrise timings

The well-known White method (A method of estimating ground-water supplies based on discharge by plants and evaporation from soil: Results of investigation in Escalante Valley, Utah. Washington D.C, US Geological Survey. Water Supply Paper 659-A United States Department of the Interior, 1932 ) based on diurnal water table observations has been widely applied to estimate groundwater evapotranspiration (ET G ) from phreatophyte vegetation. One of the limitations of this method is its large uncertainties in quantifying the daily groundwater recovery rate ( r ), which is assumed to be equal to the average rate of groundwater level rise between midnight (i.e., 00:00 h) and 04:00 h. Recent studies pointed out that ET G is highly dependent on the shape and duration of the diurnal clear-sky solar radiation curve and that using the groundwater recovery rate over a short interval of nighttime hours to represent the daily r may lead to large uncertainties in ET G estimates. In this study, we analysed the dependence of the estimated daily r on the sunset and sunrise timings. Numerical experiment results showed that the estimated r is highly sensitive to the duration between sunset and sunrise, which varies seasonally. Instead of using fixed time spans (TS s ), e.g., from midnight to 04:00 h, we recommend a more universal method for determining the TS, which is associated with the sunset and/or sunrise timings and used to estimate the daily r . This dynamic TS approach was tested at a Tamarix ramosissima -dominated riparian site with a hyper-arid climate (precipitation of 35 mm a −1 ) in northwestern China. Compared with the observed evapotranspiration (ET), our approach showed better performance and less subjectivity in estimating ET G than the traditional White approach.