WEFTEC 2003 PAPER A NON-POINT MODEL COMPARISON USING HSPF AND WARM; A CASE STUDY ON STEAMBOAT CREEK

摘要

A Non-Point Source Modeling Study was conducted to compare the Watershed Analysis Risk Management Framework (WARMF) model and the Hydrologic Simulation Program – Fortran (HSPF) model. This comparison was completed to assess WARMFs ability to be used as a model to complete TMDL studies. In this study, WARMF’s ability to predict hydrologic and water quality responses of a case subwatershed are compared to HSPF, an industry standard model used for TMDL projects. This detailed comparison was undertaken to: 1) compare the applicability of WARMF and HSPF to a watershed application; and 2) compare non-point source loading estimates and water quality calibration results for dissolve oxygen, nitrogen, phosphorous and total dissolved solids. This paper discusses the flow and water quality calibration methodology and results for both models. Flows and constituent concentrations were calibrated in both models to match observed values at three locations: Galena Gage, Rhodes Road and Kimlick Lane. The target accuracy was +-10 percent for periods with accurate data and +-20 percent for periods with marginal data quality. The HSPF model simulated hydrology relatively well with deviations, generally over-estimated flows, occurring due to snowmelt or irrigation diversions. WARMF simulated hydrology fairly well and captured the peaks and recessions due to snow melt fairly well, but generally over-estimated flows during the irrigation months. HSPF water quality calibration results followed the year to year and seasonal trends reasonably well for all constituents but did over-estimate TN, TP, TDS and TSS during the winter periods. Calibration results for WARMF followed the year to year and seasonal trends reasonably well for all constituents but under-estimated TN, TP, TDS and TSS during storm events. In addition to water quality, pollutant loads were evaluated on a monthly and yearly basis for TN, TP, TDS and TSS.