A method for investigating the potential impacts of climate-change scenarios on annual minimum groundwater levels

Previous approaches to assessing the impact of climate-change scenarios on groundwater levels and groundwater droughts have focused on modelling specific recharge processes or phenomena. However, statistical methods, based on correlations between historic groundwater level and rainfall time-series, provide an alternative and robust approach to predicting minimum groundwater levels and droughts. For the purposes of this study, groundwater droughts are defined in terms of the return period of a given groundwater level. A multiple linear regression model (regression of monthly rainfall totals for a given period against values of minimum annual groundwater levels for the same period), when used with synthetic rainfall data based on climate-change scenarios, enables changes in future annual groundwater-level minima to be modelled. The method is illustrated at three sites on the Chalk, Permo-Triassic sandstone and Jurassic limestone aquifers.