Evaluation of Climate Change and Groundwater Pumping Impact on Multidecadal Groundwater Levels Using a Physics-Constrained Wavelet-Aided Statistical Model

A physics-constrained wavelet-aided statistical model is developed to analyze and predict monthly groundwater levels on multidecadal or longer time scales. The approach retains the simplicity of regression model but is constrained by temporal scales of processes responsible for groundwater level variation, including time scales of recharge and aquifer response to groundwater pumping. The systematic approach includes (1) identification of hydrologic trends and correlations using cross-correlation and multi-time scale wavelet analyses; (2) integrating temperature-based evapotranspiration; (3) estimating the length of time for propagating groundwater pumping stresses, and (4) assessing model prediction performances. The approach is applied at three hydrogeologically distinct sites in North Florida using over 40 years of monthly groundwater levels. Model performances are assessed for both forecasting and hindcasting monthly groundwater levels from 7 to 22 years using split calibration-validation methods. All models perform well with less than 2-foot prediction error. We later employed the methodology to extend and analyze historical record of groundwater levels, over 50 to 60 years, back to 1904. To assess the validity of the hindcasts for the period of no-record, documented historical drought and wet periods and the predevelopment groundwater level studies are reviewed. As a result of hindcasting, 110 years of monthly groundwater levels (from 1904 to 2015) become available for all three sites, which are utilized to assess the effect of climate change and groundwater pumping on the frequency of low groundwater levels occurring under drought conditions by a Weibull frequency analysis. At all three sites, the frequencies of the groundwater levels falling below the critical low levels continuously for three months increase significantly from the 1904-1959 to 1960-2015 periods, which could be attributed to both climate change and groundwater pumping.