The development of groundwater simulation models in the early 1970s provided groundwater planners with quantitative techniques for analyzing alternative groundwater pumping or recharge strategies. The accuracy of a simulation model is dependent, to a certain extent, on the accuracy of the inverse solution, which in turn is determined by the quantity and quality of data. The inverse problem is inherently nonunique and unstable. It has been well understood that the number of unknown parameters must be reduced to obtain a unique and stable solution of the inverse problem. The reduction of the number of unknown parameters is achieved by means of parameterization. It also has become apparent that parameterization and its corresponding parameter values are interdependent and must be estimated simultaneously.
A recent advancement made in groundwater modeling has been the development of a generalized inverse procedure. This procedure allows us to analyze the errors in the parameter, observation, and prediction/management space. The requirement of finding the true parameter values in the classical inverse problem is replaced by a weaker requirement. More importantly, it helps us resolve the following two problems: (1) How complex should a groundwater model structure be for a given model application? (2) Are existing data sufficient for developing a reliable model for the stipulated prediction/management objective? The generalized inverse procedure attempts to find the simplest model structure for a given model application. Such a model requires the minimum amount of data to calibrate.
Was this article helpful?