부산대학교 도서관

Maozhou watershed is a highly urbanized watersheds in southern China that is a typical watershed in China's rapid urbanization. Changed land use/cover(LUC) within the watershed changed the flood response, that requires the model parameter updating with the LUC changes induced by urbanization to have an accurate flood forecasting. Flooding in Maozhou watershed is serious, which was aggravated by increased urban land area within the watershed. Accurately forecasting the flood of Maozhou watershed is imperative, but currently there is no effective flood forecasting model to be used for the flood forecasting of Maozhou river. Hydrological models used for flood forecasting are usually categorized as lumped hydrological models and distributed hydrological models, with the latter the latest development. Distributed hydrological model derives model parameters directly from the watershed terrain properties, so has the potential to update model parameters with the changed terrain properties if it could be acquired. In this study, the Liuxihe model, a physically based distributed hydrological model is employed to simulate the flood processes of Maozhou watershed, and the latest LUCs were estimated by using SVM algorithm for updating the model parameters. The result shows that Liuxihe model could be used for the flood forecasting of Maozhou watershed. [ABSTRACT FROM AUTHOR]