Improving HSPF Modeling Accuracy from FTABLES: Case Study for the Illinois River Basin

Abstract

An application of the Hydrologic Simulation Program–FORTRAN (HSPF) model was developed for the entire Illinois River basin. During the simulation period from 1985 to 1995, simulated monthly and annual flows correlated well with observed data. However, its performance for daily flows in the lower portion of the basin was not satisfactory because of the complicated hydraulic condition of the Illinois River. In order to compensate the weakness of HSPF for flow routing, we have coupled a calibrated hydraulic model, one-dimensional unsteady flow through a full network of open channels (UNET), to improve flow routing accuracy and the overall performance of the HSPF model. In this study we intended to test how HSPF modeling accuracy can be improved from modifying function tables (FTABLES). We proposed four alternative sets of FTABLES: three (Sets B, C, and D) were based on Manning’s equation and one (Set E) used the flow-stage relationship. Simulations from these four alternatives were compared to the default FTABLES (Set A) and the coupled HSPF-UNET model. A comparison showed that improving the estimation on hydraulic gradients from flood frequency analysis (Set B) or improving the computation of reach geometry from surveyed cross sections (Set C) could only improve model accuracy. However, much better modeling accuracy can be obtained by improving both reach geometry and hydraulic gradients (Set D). Using the stage-discharge relationships simulated from a hydraulic model (Set E) also provided better results. The proposed methods have been successfully applied to watershed modeling with HSPF in the Illinois River basin. This study demonstrated that better HSPF modeling accuracy can be achieved with minimum effort from representative channel geometries and hydraulic gradients.