Submitted to: International Soil Conservation Organization Conference Abstracts
Publication Type: Abstract Only
Publication Acceptance Date: 8/1/2001
Publication Date: 3/1/2002
Citation: ZHANG, X.J., GARBRECHT, J.D. EVALUATION OF CLIGEN WEATHER GENERATOR AND ITS IMPACT ON WEPP SOIL LOSS ANDRUNOFF PREDICTIONS. INTERNATIONAL SOIL CONSERVATION ORGANIZATION CONFERENCE ABSTRACTS. 2002. Interpretive Summary:
Technical Abstract: CLIGEN is a stochastic weather generator, which generates daily weather time sequences for use in many natural resources management and hydrological models. The WEPP model, which represents a new generation of physically based soil erosion models, has been used and tested worldwide by many scientists. The objectives of this study are to evaluate the ability of the CLIGEN model to reproduce long-term climate data and to explore the potential of using CLIGEN to deliberately synthesize daily weather sequences of desired climate scenarios and or long-term weather forecasts. Daily weather data from the National Weather Service and Oklahoma Mesonet at four Oklahoma locations and measured storm data from other eight sites scattered in the eastern US were used. The non-parametric Wilcoxon and the Kolmogorov-Smirnov tests were conducted. Daily non-precipitation variables generated by CLIGEN at four Oklahoma locations were adequate, though statistically different, for use in the WEPP model. For precipitation variables, generated daily precipitation amounts were somewhat acceptable; however, generated storm duration and relative peak intensity were different from measured at all four locations at P<0.0001 for both tests. More importantly, the generated data showed little or no cross correlations among these variables, while the measured data from all locations showed strong cross correlations. WEPP tests showed that the disparities of generated data, compared with measured data, resulted in an average absolute error of 16% for runoff prediction and 29% for soil loss under conventionally tilled row crop conditions at eight sites. A detailed test on one Oklahoma location revealed that the CLIGEN model is capable of adequately reproducing desired climate scenarios.