Simulation and Optimization of Groundwater Use in Coastal Areas


Creative Commons License

Motz L. H., Gördü F., Yurtal R.

American Geophysical Union, Fall Meeting 2001, California, United States Of America, 10 - 14 December 2001, vol.1, pp.1

  • Publication Type: Conference Paper / Summary Text
  • Volume: 1
  • City: California
  • Country: United States Of America
  • Page Numbers: pp.1
  • Çukurova University Affiliated: Yes

Abstract

The simulation of groundwater use in coastal areas requires consideration of seawater intrusion, and optimization can involve maximizing the total pumping rate subject to hydraulic and environmental constraints. Groundwater models that represent freshwater and saltwater as immiscible fluids separated by a sharp interface have been utilized for simulation as well as variable density models that combine the groundwater flow equations with contaminant transport advection-dispersion equations. The response of an aquifer system can be linked to an optimization model by means of the response matrix method, in which influence coefficients representing the responses of heads and chloride concentrations to pumping are determined. In this investigation, the variable density, finite-element numerical code SUTRA (Saturated-Unsaturated Transport) was used to simulate a vertical cross-section of the coastal aquifer system in the Goksu Delta at Silifke, Turkey. Chloride concentrations in the groundwater model range from 0 mg/L in the freshwater to 22,000 mg/L, which represents Mediterranean seawater that has intruded into the aquifer along its bottom boundary. It was assumed that pumping occurs from two wells along the cross-section, and aquifer response coefficients were generated at these two specific well locations using the groundwater model. A linear optimization model was constructed under steady-state conditions to maximize the total pumpage from the two wells subject to water demands and chloride concentration and drawdown limitations. The GAMS (General Algebraic Modeling System) code was used to execute the optimization model, and a trade-off curve was constructed to relate the maximum allowable pumping rate from each well to various chloride concentration levels.