PubAg

Main content area

A coupled stochastic inverse/sharp interface seawater intrusion approach for coastal aquifers under groundwater parameter uncertainty

Author:
Llopis-Albert, Carlos, Merigó, José M., Xu, Yejun
Source:
Journal of hydrology 2016 v.540 pp. 774-783
ISSN:
0022-1694
Subject:
Monte Carlo method, aquifers, equations, expert opinion, freshwater, geophysics, groundwater, groundwater flow, hydraulic conductivity, hydrologic models, mass flow, mass transfer, parameter uncertainty, saline water, saltwater intrusion, surveys
Abstract:
This paper presents an alternative approach to deal with seawater intrusion problems, that overcomes some of the limitations of previous works, by coupling the well-known SWI2 package for MODFLOW with a stochastic inverse model named GC method. On the one hand, the SWI2 allows a vertically integrated variable-density groundwater flow and seawater intrusion in coastal multi-aquifer systems, and a reduction in number of required model cells and the elimination of the need to solve the advective-dispersive transport equation, which leads to substantial model run-time savings. On the other hand, the GC method allows dealing with groundwater parameter uncertainty by constraining stochastic simulations to flow and mass transport data (i.e., hydraulic conductivity, freshwater heads, saltwater concentrations and travel times) and also to secondary information obtained from expert judgment or geophysical surveys, thus reducing uncertainty and increasing reliability in meeting the environmental standards. The methodology has been successfully applied to a transient movement of the freshwater-seawater interface in response to changing freshwater inflow in a two-aquifer coastal aquifer system, where an uncertainty assessment has been carried out by means of Monte Carlo simulation techniques. The approach also allows partially overcoming the neglected diffusion and dispersion processes after the conditioning process since the uncertainty is reduced and results are closer to available data.
Agid:
5247955