In a recent model review, DHI’s integrated groundwater and surface water simulation system MIKE SHE was selected as the best tool to model groundwater-surface water related issues. The intercomparison was carried out in connection with a controversial study for the Rocky Flats Environmental Technological Site, Denver, Colorado. MIKE SHE was preferred among 9 integrated hydrological models (mainly based on MODFLOW extensions). MIKE SHE was found to provide the best and most comprehensive description of the surface water - ground water interaction with a full dynamic coupled description of the relevant hydrological processes.
The Rocky Flats Environmental Technology Site (RFETS) is located 16 miles northwest of Denver, Colorado, It encompasses approximately 6,500 acres and is owned by the Department of Energy (DOE). RFETS has earlier been a nuclear weapon research, development, and production complex. The site is currently undergoing aggressive cleanup with the objective to close the site by the end of 2005. Kaiser-Hill has been contracted by DOE to manage and reconfigure the site leading to final closure. The objective of conducting a site wide water balance is to provide RFETS with a management tool for simulating how the area’s hydrology will change from current to final state. An integrated model is required, which couples and simultaneously simulates all the principal components of the hydrologic regime, including snowmelt, overland flow, channel flow, unsaturated zone flow and saturated groundwater flow. DHI was hired as sub-consultants to assist in the model application.
There is tremendous public and scientific attention on the cleanup of the site, which has also affected the model selection. A systematic approach was followed, and a set of general requirements was defined requiring that the code should be commercially available, be able to run with high resolution and predictive accuracy, to work at various scales, to accommodate a very variable and complex site hydrology, be flexible and efficient to comply with project constraints and have the ability to interface with GIS and graphical tools.
Only integrated, deterministic, distributed, physically based, integrated, continuous simulation models were considered. This meant that the basic MODFLOW groundwater model was not considered, but a number of MODFLOW based integrated models were reviewed among a total of nine models. The main reasons for this was their inability to describe all processes in an integrated and physically based manner.