A Methodology for Management of Heterogeneous Site Characterizaton and Modeling Data

Scientists, site managers and regulators participating in environmental remediation activities usually need to access, create, and manage large amounts of heterogeneous data, including disparate site characterization data, and models that range from conceptual to numerical. The data are necessary to enable the development of conceptual and numerical models, parameter estimation, and multiple simulations. We designed a methodology based on DOE Savannah River Site F-Area and Hanford databases that included well layout, concentration, groundwater level, lithology, meteorological records, contaminant levels, etc. The approach was necessary to effectively manage the staging and views of environmental data to ensure that site characterization and monitoring data are readily available as inputs for parameter estimation, numerical predictions, uncertainty quantification, and risk analysis. This paper describes the methodology and the application of the developed data management system to the Savannah River F-Area site observations. In particular, we present methods for organizing different databases and data types into a common framework that allows the user to browse the data as a single coherent dataset. Analyses of the Savannah River F-area using the data management system illustrate the utility of the system. The paper also demonstrates the utility of interfaces designed to leverage and integrate existing large-scale data management, provenance and indexing technologies from the scientific community and discuss the next steps.