Identifying the parameters of an aquifer cell model

Cell models are commonly used for forecasting water levels in aquifers. Calibration of such models is achieved through identification of their parameter values, the transmissivities and storativities of all cells, using historical data. Several methods of formulating the identification as a linear or quadratic programing problem are presented. Examples are given, results of the various methods compared, and the sensitivity of these results to errors in the data is discussed. Inflows or outflows during historical periods may also be determined by the same methods; an example for a real aquifer is presented.