Monday, May 7, 2012
Non-point source pollution is a worldwide emerging threat for water resources. The role groundwater is typically understood as a “black box” of intermediatate nitrogen storage within a watershed. Little work has been done to understand and highlight the dynamics of an aquifer system as the link between spatio-temporally distributed pollution sources and spatially distributed pollution-affected systems such as wells, springs, and streams. In this study we employ a recently developed groundwater modeling framework to predict the efficiently simulated transport and fate of non-point source pollutants within large groundwater basins. The model consists of a spatially detailed, highly resolved groundwater flow model to simulate long-term average flow dynamics within the groundwater basin; and an efficient streamline transport model, which is used to link spatially distributed contamination sources with spatially distributed water supply systems (e.g. wells). We compute travel time distributions of contaminants, then use the unit response function concept to rapidly simulate long time breakthrough curves in a large number of water supply systems based on historic and alternative pollutant loading management scenarios. The model is applied to a large groundwater basin in the southern Central Valley, the Tulare Lake Basin. To derive a detailed average groundwater flow field we used a simplistic domain decomposition method, where an existing coarse groundwater field is refined locally. A hundred-year nitrogen loading history is simulated and results compared against historic water quality data. A sensitivity analysis is performed over various parameters of the model, in an attempt to improve the predictive capability of the model.