IMPROVEMENT AND EVALUATION OF CLIGEN FOR STORM GENERATION

Abstract

The program, CLIGEN, generates peak rainfall intensity and storm duration and other daily weather variables for WEPP to predict the rate of runoff and soil loss. Unrealistic peak rainfall intensity simulated by CLIGEN (version 4.2) led to a discovery of a software bug and subsequent modification of the method to estimate the monthly mean of the maximum 30-min rainfall depth for storm generation. To evaluate the modified CLIGEN, break-point rainfall data for 14 sites in the United States were used for periods varying from 4 to 19 years. The modified CLIGEN was then used to generate weather data for a period of 100 years for the 14 sites. WEPP (version 99.5) was run for the 14 sites, using three soil types for each site, so that the simulated mean annual runoff and soil loss can be compared with those using the observed break-point data. For most (> 96%) of the 42 site-soil combinations tested, there is no significant difference in WEPP-simulated mean annual runoff and soil loss at the 0.05 level between the break-point rainfall data and CLIGEN-generated rainfall data. The bias in the mean is less than 2 to 3% for runoff and soil loss when all sites are considered. The minimum bias in the mean annual runoff and soil loss lends support for the modified CLIGEN to generate input for WEPP for the purpose of runoff and soil loss predictions.