IMPROVED SUBSURFACE MODEL CALIBRATION USING SOFT DATA ON GEOLOGIC HETEROGENEITY

Even with the many recent advances made in the use of inverse modeling techniques for parameter estimation, in practice, model calibration to a large extent, still relies on trial and error methods. These methods use available field data from observation wells that are limited in number and are not necessarily optimally located. These limitations introduce uncertainty to estimated parameters, thus introducing errors to predications that are made using the calibrated model. Using a series of data sets generated in a three-dimensional synthetic aquifer, how the accuracy of parameter estimation improved with increasing quantities of soft and hard data incorporated in the model construction and inversion was investigated. The synthetic aquifer (208(L) x 117(W) x 57cm(H), constructed using five well-characterized sands) was packed to represent a stationary spatially correlated random field with a moderate heterogeneity and a lens-like layer of fine sand embedded into the stationary field, resulting in the composite aquifer heterogeneity to be non-stationary. The geologic feature of a single layer of fine sand was treated as soft data gathered through geophysical characterization. The pressure distribution within the aquifer was measured at 92 monitoring locations using an automated measuring system. Analysis was conducted using the flow simulator MODFLOW and inversion code UCODE. The results suggest that the accuracy in parameter estimation improves with increasing quantities of both soft and hard data. Although the spatial distribution of the observation wells did not show a clear effect on the estimated parameter accuracy, the pumping location may have an important influence. This suggests a model designed to capture geological complexity in a reasonable manner and a well-designed pumping/observation well network will both help to improve the accuracy of parameter estimation.