A Modeling Approach Appropriate For Managing Seawater Intrusion Under Uncertainty
Abstract
Managing groundwater in coastal basins is often accomplished with groundwater models calibrated to known conditions. To overcome the problem of an unknown seawater interface location in the Soquel-Aptos Basin of California, we developed target water levels protective of seawater intrusion using quasi-steady state simulations. Cross-sectional models extending offshore helped identify onshore piezometric heads that allow the interface to equilibrate at a known offshore location, regardless of the interface's current position. We attempted to adopt an appropriate level of model complexity that would result in protective, yet relatively low target water levels - the cost of recharging the groundwater basin to meet higher target water levels could be significant. Numerical cross-sectional models were selected over analytical models to minimize target groundwater levels. The numerical models were kept relatively simple; additional complexity may result in less costly target water levels, but the available data do not appear to substantiate additional complexity. Initial efforts to develop optimized water level targets were unsuccessful due to our attempts to achieve an unnecessary level of precision. When the level of model precision was matched to likely management practice, the modeling effort became more efficient. A Monte Carlo analysis was applied to develop a range of possible target water levels. The Monte Carlo approach was intended to allow the managing water district to adopt any one of a range of management decisions, based on their perceived threat of seawater intrusion combined with the cost of raising water levels. We will address whether the selected level of modeling was appropriate, or if simpler models would result in similar decisions. Furthermore, we will discuss the utility and problems of the Monte Carlo analysis; the Monte Carlo analysis provides a clearer picture of risks associated with adopting low target groundwater levels, but it is unclear that the Monte Carlo analysis leads to better management decisions. Finally we will discuss objections that have been raised since the modeling has been completed. While some objections are valid, these objections do not invalidate model results if we assume the level of model complexity is appropriate for making management decisions.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2008
- Bibcode:
- 2008AGUFM.H33G1118W
- Keywords:
-
- 1829 Groundwater hydrology;
- 1847 Modeling;
- 1880 Water management (6334)