Using Coupled Hydrogeophysical Modeling to Assess the Relative Value of Proposed Gravity and Water Level Observations to Support Water Resources Decision Making

Abstract

Decisions about the permitting of new groundwater extractions often depend on the perceived impact of those withdrawals on groundwater levels in wells and flow in nearby streams. Owing to subsurface abnormalities and varying hydraulic flow conditions that can occur in the subsurface, there is much uncertainty when assessing possible impacts. Using an ensemble modeling approach can better inform these decisions, which quantifies both the most likely outcome, and the associated uncertainty given, limits on subsurface hydrogeologic information. The ensemble approach encompasses the uncertainty that is possible in the domain by varying parameters in the model. Each model in the ensemble is weighted to a degree that reduces the uncertainty of future predictions, thereby improving decision making. Groundwater levels in wells are one of the most common hydrologic measurements, but it can be prohibitively expensive to drill wells to add new observation points to inform decision making. Time-lapse gravity measurements provide a proxy method to gain insight into the subsurface hydrologic conditions. While gravity measurements are less direct than groundwater levels, it can be considerably less expensive to add monitoring points. In this study, an ensemble of models is developed for a synthetic catchment. Forecasts of drawdown in one well due to the addition of another well are the prediction of interest (POI). This POI is then converted to a utility value to maximize the satisfaction for decision making. The accuracy and uncertainty of the forecasts are calculated with and without added observations (water levels and gravity). This is investigated for the addition of another well at three different locations. The result is a panel of maps of the basin showing the relative expected value of an added observation at each location for improving the satisfaction of decision-makers. This panel can be used to choose among monitoring well or gravity measurement locations before data are collected, dependent on the location of the additional well. The same approach can be extended to consider multiple measurements of different types.