2006 年 12 巻 p. 329-334
This paper develops the two-cascade storage routing model with four parameters that can generalize the snowmelt runoff for study river basins of interest. First, the present study describes the algorithms that mathematically optimize model parameters, using the Newton-Raphson method in which sensitivity coefficients are analytically derived. Second, model parameters are identified only from the largest flood event during the whole melting season. Finally, the four optimized model parameters are employed to reproduce a series of snowmelt flood hydrographs with high accuracy from April to June.
The proposed methods are applied to analyses of snowmelt runoff in Hoheikyo Dam located in the Toyohira River, which is a tributary of the lshikari River. It was found from simulation results that the proposed two-cascade storage routing model could reproduce snowmelt hydrographs in good accordance for the whole melting season in different water years.