Uncertainty in Exploitation Rates and Numerical Modeling of Its Impact on Seawater Intrusion in the Korba Aquifer (Tunisia). Kerrou, J, Lecca, G, Renard, P, & Tarhouni, J In Proceedings 1st SWIM-SWICA Joint Saltwater Intrusion Conference, pages 27–34.
Uncertainty in Exploitation Rates and Numerical Modeling of Its Impact on Seawater Intrusion in the Korba Aquifer (Tunisia) [pdf]Paper  abstract   bibtex   
The Cape Bon peninsula, located 60 km South East of Tunis, is one of the most productive agricultural areas in Tunisia. At the same time, it suffers heavily from water scarcity and salinization due to seawater intrusion especially in the Korba aquifer. In 2002, the Korba aquifer was exploited from more than 9000 wells for a total volume of 54 106 m3. [\n] Groundwater balance estimates in the area show that one of the major sources of uncertainty is the evaluation of the aquifer exploitation. Indeed, no precise information is systematically recorded by the local authority concerning either the current extraction rates or their evolution in time. Knowledge of spatial and temporal exploitation distribution is crucial to understand the dynamic of the aquifer to model the seawater intrusion and to investigate optimal management scenarios. [\n] In this paper, we present a multivariate analysis to evaluate the uncertainty in the estimation of exploitation rates using secondary information including: aquifer geometry and physical parameters, surface water irrigation records and field measurements. Using direct measurements and secondary data, a geostatistical model of exploitation rates was constructed. The impact of the exploitation uncertainty on the seawater intrusion was evaluated with Monte Carlo simulations, based on a 3D density dependent variably saturated groundwater flow and miscible salt transport model. To circumvent the large computing time required to run multiple 3D simulations, the numerical model was run on the GRID infrastructure developed by the EGEE project (Enabling Grid for E Science in Europe). [\n] Results demonstrated the possibility of using secondary information in a geostatistical framework to model the spatial distribution of pumping rates and its uncertainty. The Monte Carlo simulations showed that uncertain pumping rate localization led to a zone of 20.4 km2 where the groundwater heads and concentrations were not known with accuracy. While the size of this uncertain zone remains small with respect to the size of the aquifer, it can be of high importance in terms of potential economic impact because of the risk of agricultural productivity losses when using this water.

Downloads: 0