Groundwater management under uncertainty: a multi-objective approach

A methodology is developed for determining a robust optimal strategy for a groundwater hydraulic control problem posed within the framework of stochastic multi-objective optimisation. The methodology explicitly considers uncertainty in hydraulic conductivity and allows decision-makers (DMs) to evaluate trade-offs among three conflicting objectives: aquifer yield, investment and operational cost, and recourse cost (penalties) incurred if the stipulated constraints are violated. The model includes a two-stage decision process in which the first-stage decision (commonly referred to as "here and now") on how many wells, their pumping rates and their locations is made and implemented immediately without the foreknowledge of the outcome of the uncertain parameters. At a later stage, when the uncertain parameters become known, a second-stage decision is taken using the updated data. Applicability of the methodology is demonstrated through a hypothetical (but realistic) example. A post-optimality Monte Carlo analysis is conducted to examine the performance of the model in terms of robustness (stability). Preliminary results show that robust optimisation can be useful in designing a strategy which performs reasonably well whatever the outcome of the uncertain parameters.