Assessing the reliability of exponential recession in the water table fluctuation method

•Water table fluctuation method analyzed through comparison to analytically based recharge.•The water table fluctuation method mostly under-estimates recharge.•Current approaches to the extrapolation of the antecedent recession curve need to be revised. Distributed recharge is commonly predicted from groundwater level data by adopting the water table fluctuation method (WTFM). The simplicity of the technique makes it attractive for groundwater management applications seeking sustainable levels of extraction. While there are variations to the WTFM, the classic approach extends the antecedent recession curve (prior to recharge events) to allow for the estimation of the gross recharge. This is achieved using either the previous (local) recession or a master recession curve obtained from multiple recession events. The most common function used for the recession extension is exponential. Despite the wide application of the WTFM, remarkably, a validation of this approach against known recharge values has not been previously attempted. This is the goal of the current study, which also compares local recession and master recession curve approaches adopting an exponential function for estimating recharge using the WTFM. Stochastic analysis applying an existing analytical solution to water table fluctuations from intermittent recharge was used to produce 1000 hypothetical hydrographs. From these, WTFM-based recharge was estimated for three recession periods of differing lengths, producing 6000 estimates of recharge (1000 simulations, two recession curve approaches, three recharge-recession events). The WTFM produced an average under-estimation error of 14%. The WTFM is more likely to obtain recharge errors within 5% of the true value using the master recession curve approach. This study demonstrates the need to revise the WTFM to eliminate bias, especially in the extrapolation of antecedent recession curves.