Parameter Sensitivity Analysis of the HSPF Model on a Pervious Landcover Dominated Watershed

Sensitivity analysis is an important assessment usually performed to gain a better understanding of the hydrological processes impacting model outcomes as well as to identify most dominant parameters for model optimization and calibration. In this study, a watershed model developed using the Hydrologic Simulation Program-Fortran (HSPF) for South Chickamauga Creek watershed was evaluated for parameter sensitivity. Critical HSPF model parameters, i.e., lower zone evapotranspiration (LZETP), index to infiltration capacity (INFILT), base groundwater recession (AGWRC), interflow inflow (INTFW), fraction of groundwater inflow to deep losses (DEEPFR), etc., were investigated for the sensitivity analysis. This paper presents the sensitivity analysis results using sensitivity factor, which is expressed as a ratio of the percentage change in model output (i.e., runoff, peak flow) to percentage change in model input parameter (expressed as percentage) and explains the variability of model outcomes from the selected perturbations of input parameters. Low flow conditions were mostly influenced by AGWRC, DEEPFR, LZETP, and INFILT, whereas INFILT and INTFW had most influence on the peak flows. Annual runoff volume and high flow conditions were highly sensitive to LZETP. However, AGWRC was found to be the most influential parameter in all hydrologic responses in this watershed. Overall, this study helps watershed management by providing insight into model structure and adjustment. Besides, it will work as a basis for future multisite sensitivity and uncertainty analysis.