Identification of Errors in the Hanford Site-Wide Groundwater Flow Model by Inverse Modeling of Alternative Conceptual Models
Abstract
A regional-scale, three-dimensional groundwater flow and transport modeling effort has been undertaken to quantify the environmental consequences of past waste disposal activities and support ongoing environmental management activities at the U.S. Department of Energy's Hanford Site. An important aspect of this effort is the identification and quantification of uncertainties associated with model predictions. It is recognized that such uncertainties arise not only from selection of inappropriate groundwater model parameters (parameter error), but also from the underlying conceptualization of the groundwater system (model error). Therefore, we have adopted an approach to uncertainty characterization that involves the evaluation of multiple alternative conceptual models (ACMs) within an inverse modeling framework. The initial step in implementation of the framework was the development of a multi-processor implementation of the UCODE inverse modeling system and application of the inverse framework to update parameter estimates from a prior deterministic model. A preliminary first-order uncertainty analysis was performed based on the model results. At the same time, site geologists developed an improved conceptual model of the 3D structure of the aquifer system. Inverse modeling of the updated conceptual model led to estimates of some parameters, especially specific yield, that were not plausible, indicating that there were problems with the conceptual model. As a result, additional ACMs were developed and subjected to inverse analysis, including an alternative with modified boundary conditions (leaky underlying bedrock), an alternative incorporating surface recharge modifications based on surface run-on from an adjacent topographic feature, and an alternative incorporating an improved description of the timing and volume of waste discharges arriving at the water table (upper model boundary). Model predictions of transient hydraulic heads under each ACM were compared to 69,000 historical head observations, and estimated parameters were evaluated for plausibility. Based on this analysis, an improved conceptual model that led to decreased predictive uncertainty and utilizes plausible estimated parameters was achieved. Ongoing refinement of the model is focused on a stochastic conceptualization of the structure of low-permeability mud units and parameter zonation for hydraulic properties in the uppermost transmissive layer of the aquifer. Document PNNL-SA-37052. Pacific Northwest National Laboratory is operated for the U. S. Department of Energy under Contract DE-AC06-76RL01830.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2002
- Bibcode:
- 2002AGUFM.H71A0783C
- Keywords:
-
- 1829 Groundwater hydrology;
- 1832 Groundwater transport;
- 3210 Modeling;
- 3260 Inverse theory