Automatic calibration of groundwater models by the head gradient method
Abstract
A groundwater model for flow and transport is reliable only if it is properly calibrated on observations. The best way to ensure a high predictive capability is to limit the number of parameters by defining homogeneous zones. These zones may be defined by an inverse model based on the gradient method. This paper describes a new scheme of inverse model for irregular finite difference 3D models in steady-state. The method determines, in each grid cell, the modification of hydraulic conductivity necessary to adjust the computed head gradients to the observed ones and yields the hydraulic conductivity in the centre of each grid cell. The stability of the results is forced through an optimization scheme simultaneously minimizing the differences of hydraulic conductivity between adjacent grid cells. After determination of the hydraulic conductivity, a procedure derives automatically optimal clusters of values defining geographical zones where homogeneous values of hydraulic conductivity may be imposed. The inverse modelling scheme is operational and applications to four real aquifer systems are presented.
Fichier principal
Publi_Iahs_220_0291_1993_DTH_Head_Gradient_Calibrat.pdf (633.38 Ko)
Télécharger le fichier
Origin : Publisher files allowed on an open archive
Loading...