Abstract
Numerical experiments were performed to analyze the generation of a secondary discharge peak from a small watershed. A simple physically based three-dimensional infiltration simulation was adopted at the Minamitani valley head watershed (0.45 ha) in university forest in Aichi, the university of Tokyo. The parameters of hydraulic characteristics of topsoil were determined by sample tests and those of bedrock were sought to simulate the observed hydrographs and groundwater levels. Changing such numerical conditions as soil depth, bedrock depth and hydraulic conductivity showed that the secondary discharge peak from this watershed was generated when the value of soil or bedrock hydraulic conductivity was not changed vertically, or when a comparatively high permeable region was dominant in a watershed. The secondary discharge peak increased with increasing hydraulic conductivity, but the time of the secondary discharge peak did not change with increasing hydraulic conductivity. A secondary discharge peak was not generated in cases of high permeable regions in upper layers and low permeable region in lower layers. Soil water infiltration into bedrock functioned to suppress the secondary discharge peak under some conditions because the hydraulic conductivity of bedrock is commonly low. Increased soil or bedrock depth caused very stable base discharge whether or not the secondary discharge peak was generated. These results demonstrate the relationship among the physical conditions of the intended watershed and the conditions that generate the secondary discharge peak.
