Investigation of techniques to estimate rainfall-loss parameters for Illinois
An attempt was made by the U.S. Geological Survey to develop parameter-estimation techniques for two rainfall-loss computation methods used in the U.S. Army Corps of Engineers ' flood-hydrograph model (HEC-1). Six rainfall-loss parameters were investigated - four for the Exponential Loss-Rate method and two for the Initial and Uniform Loss-Rate method. Multiple-regression analyses using data from 616 storms at 98 gaged basins were used to attempt to develop parameter-estimation techniques for ungaged basins in Illinois. Techniques were evaluated using 102 storms at 36 uncalibrated gage basins. Estimated unit-hydrograph and rainfall-loss parameters were used to compute discharge hydrograph characteristics, which were compared with characteristics of observed discharge hydrographs. Seventy and 50% of the simulations using the Exponential Loss-Rate and the Initial and Uniform Loss-Rate methods, respectively, produced valid hydrographs. Model-sensitivity analyses to one standard error of estimate indicate that the storm-dependent rainfall-loss parameters are most significant in reproducing the sum of incremental flows and peak discharge. Time of concentrations and storage coefficient are significant in reproducing time to peak discharge and peak discharge, respectively. The small percentage of variation explained by the estimation techniques and the evaluation results using 102 storms indicates that a large degree of uncertainty exists in the computed hydrographs. (USGS)