A statistical approach to estimating evapotranspiration from diurnal groundwater level fluctuations

Over the last few decades, automatic sensors that record groundwater levels at high-frequency intervals have become widely used in groundwater monitoring practice. These sensors provide large amounts of data regarding diurnal groundwater fluctuations, which can be treated as stochastic periodic time series. In this study, a simple relationship between the average standard deviation of diurnal groundwater level fluctuations and the daily evapotranspiration over relatively short periods (days or weeks) was developed for estimating groundwater consumption by phreatophytes in arid/semiarid areas. Our approach allows estimating groundwater evapotranspiration ( ETg) using stable statistical characteristics of diurnal groundwater fluctuations, and it is useful for analyzing large amounts of data obtained from digital groundwater level monitoring sensors. A comparison of the ETg results from a synthetic set of groundwater level fluctuations with predefined values shows that this technique behaves consistently and is robust. A numerical analysis of one-dimensional saturated-unsaturated water flow to a root system using Richards' equation indicates that this method provides a reliable estimate of ETg when the basic assumptions of the White method are met. The method was also applied to two phreatophyte-dominated riparian sites in New Mexico to demonstrate its usefulness, which provides better results than the commonly used White method. Key Points <list id="wrcr20810-list-0001" list-type="bulleted"> A new approach to estimating groundwater evapotranspiration is proposed Diurnal groundwater fluctuation has relative stable statistical characteristics The approach is useful for analysis of large amount of data