2002 Denver Annual Meeting (October 27-30, 2002)

Paper No. 4
Presentation Time: 2:15 PM


DETHAN, Deepu, GRAETTINGER, Andrew J. and LEE, Jejung, Department of Civil & Environmental Engineering, Univ of Alabama, Tuscaloosa, AL 35487, detha001@bama.ua.edu

Quantitatively Directed Exploration (QDE) approaches based on information such as modeling sensitivity, input covariance, and output covariance are presented. Seven approaches for directing exploration are developed, applied, and evaluated on a synthetic model. Spatial input parameter values and their covariance are calculated by a multivariate conditional probability calculation from a limited number of samples. Since geologic data tend to be spatially correlated, an estimation of the variogram structure is required to perform this calculation. Model sensitivity can be determined by perturbing input data and evaluating output response or, as was used in this work, sensitivities can be programmed directly into the engineering analysis model. Output covariance is calculated by the First-Order Second Moment (FOSM) method, which combines the covariance of input data and model sensitivity. A groundwater flow example, modeled in MODFLOW-2000, is chosen to demonstrate the seven QDE approaches. MODFLOW-2000 is used to obtain the piezometric head and the sensitivity of head to changes in input data. The seven QDE approaches are rated based on the “accuracy improvement” of the piezometric head after QDE sample is added. “Accuracy improvement” is evaluated by comparing model head (before and after new data is added) to actual head at the site. For the synthetic site used in this study, the QDE approach that identifies the next sample location based on which location contributes the most to overall site variance proved to be the best method to quantitatively direct exploration.