The Calibration of a Complex Watershed Model in Sparse Data Environments

Over two hundred and thirty watersheds throughout West Virginia are currently part of a Cumulative Hydrologic Impact Assessment (CHIA) project to determine the effects of surface and underground coal mining on stream water quality and quantity. The hydrologic watershed model, Hydrologic Simulation Program-Fortran (HSPF), is used, along with a GIS database and associated tools, to simulate stream discharge and those water quality parameters important in measuring the impact of AMD (Acid Mine Drainage). Satisfactory calibration of the stream flow component of HSPF had to be demonstrated, in spite of the data availability being limited to precipitation, air temperature, and potential evapotranspiration (PET) at each site. HSPF is a continuous deterministic model, requiring the estimation of more than twenty physically-based parameters. Because of the very limited availability of stream flow data, five gaged watersheds were selected to serve as calibration watersheds for the 235 CHIA watersheds, based on their geographic locations and similarity in characteristics. Four additional watersheds were chosen to test the validity of transferring the model parameters into non-calibrated watersheds. A joint calibration procedure, following the method used by R. S. Dinicola was used to find a common set of parameters for land use and cover across all watersheds. Two software packages, the USGS interactive Expert System for the calibration of HSPF and the automatic calibration program, Parameter Estimation with HSPF were used to aid in the calibration process. Although the initial calibration, and verification, was viewed as moderately successful, some watersheds still violated specific calibration goals. A new, physically-based, analytical relationship between the SCS Curve Number (CN) and a limited number of HSPF parameters is introduced to effectively reduce the number of parameters required, and to relate HSPF parameters to the CN method as used in mine site hydrology.