An efficient optimisation method in groundwater resource management

Abstract

Uncertainty in input parameters to groundwater flow problems has been recognised as an impediment to designing efficient groundwater management strategies. The most popular approach to tackling this problem has been through the Monte Carlo approach. However, this approach is generally too expensive in terms of computer time because of the number of scenarios required to ensure reliable statistics. Furthermore, solutions obtained through this approach are not necessarily robust. In this paper, it is shown how groundwater management problems, where input parameters are uncertain can be reformulated as second-order cone optimisation (SOCO) problems, which are efficiently solved by recently developed interior-point methods. Results for a real-world case application of a groundwater aquifer found in Kenya are presented.