ESTIMATE HISTORIC GROUNDWATER FLUCTUATIONS IN WETLANDS WITH EFFECTIVE MONTHLY RECHARGE (WEM) MODEL

Estimates of long-term water-level fluctuations in aquifers that discharge to wetlands may provide guidance to wildlife habitat managers, wetland design professionals, and agency regulators about the probable availability of future groundwater supply. The Effective Monthly Recharge (W_em) model generates a synthetic hydrograph of historic water table elevations for precipitation-driven wetlands and similar groundwater systems on broad interfluves and toeslopes. A time-weighted averaging technique, the W_em water-budget model simulates recharge fluctuations over time using monthly weather data. Recharge equals precipitation, adjusted for interception (I), minus FAO Penman-Monteith evapotranspiration estimates (ET). Model users calibrate calculated monthly W_em values against monthly head data by varying the number of months of weather data (n) used in the W_em calculation, and the weight (d) applied to antecedent conditions. The combination of n and d that generates the best correlation of W_em vs measured heads is used to estimate monthly head values for preceding years; the estimates are generated with historic weather data. Twenty-five years of local weather data and daily head data from a USGS well in Suffolk allow us to test and validate the model using the head on first day of each month. Data from 2003-2005 suggest n=18 months and d=0.9 provide the best fit (R²=0.89) for this site; this correlation made using filtered data (remove all data affected by rain greater than 0.2 inches during the five days before the first of each month) greatly exceeds the correlation generated by using unfiltered data (R²=0.56). Comparison of all monthly USGS well data (unfiltered) and the model-generated head values for 1981-2005 produces a significant correlation (R²=0.68). Sensitivity analyses of interception estimates suggest an annual average I=0.25 maximizes the correlation coefficient for these data.