A knowledge based framework for groundwater modeling

One of the major problems of hydrogeology modeling is the basic data preparation namely those related to the geology, hydrogeology, hydrometeorology, delimitation of the study area, and the appropriate choice of model parameters. The experienced user or a newcomer of various existing powerful models (such as MODFLOW, FEFLOW ...) is often lost during the data preparation phase highly required for running a model. Hence, the motivation of this work to develop an intelligent system able to reproduce heuristics mechanisms of hydrogeologists' experts to assist any engineer / researcher in all stages prior to any hydrogeological modeling of the groundwater system studied. A knowledge-based framework, dedicated to this purpose, was built adopting object-oriented approach, and using an Expert System shell. This paper presents the main steps of the framework development, its top-level object hierarchy, and corresponding heuristics rules. Most of the advice given by the intelligent working system was evaluated and validated in the Moroccan Rhis-Nekor aquifer case study to prepare the necessary inputs data for Hydrodynamic model and to guide the user during the modeling process.