Evaluation of the unique characteristics of runoff with both ‐dq/dt and contributing area rate in recession stage, and estimation of discharge during large‐scale floods

Jun‐ichi KURIHARA, Taro UCHIDA, Takao YAMAKOSHI and Ryosuke OKUYAMA

Abstract

Regarding the runoff analysis for sediment and flood damages (sediment laden floods), a method of identifying the parameters of the storage function method with the relationship between the runoff height (q) and the rate of decrease of runoff height (‐dq/dt) was investigated. In this study, we assumed that when the rainfall amount is larger than a certain level, the entire basin becomes a contributing area for runoff, and the falling limb of the hydrograph after the end of rainfall follows a site‐specific time‐discharge relationship that is independent of the rainfall pattern. Moreover, we assumed that even if the rainfall amount is smaller than the certain level, the falling limb of the hydrograph can be described by the site‐specific time‐discharge relationship and the ratio of contributing area to the whole catchment area. The purpose of this study is to find out whether the relationship between q during recession and ‐dq/dt will be site‐specific if the role of contributing area could be considered even in a basin with a scale of several tens of km2, and to clarify whether it is possible to estimate the discharge during large‐scale floods using this site‐specific q~‐dq/dt relationship. We analyzed observed discharge data in Uchikawa, Miyagi, which has an area of87km2. We found once we tuned the contributing area ratio, there are single q~‐dq/dt relationship for eight falling limb of hydrograph, supporting to our hypothesis. Moreover, we confirmed that the discharge during large‐scale floods can be predicted using the parameter determined based on the q~‐dq/dt relationship. In this analysis, it became clear that since the data in low flow periods are much larger in that of high flow, the parameters were easily influenced by low flow date. Moreover, the date in low flow showed large variation. To solve these problems, we proposed a method to extract data used for identification according to constant rate of decrease in runoff height. Compared to the conventional q‐S graph identification, this method can solve the problem how to determine the initial value of the storage height, and can determine parameters directly from the observed flood level. The relationship between the spatial distribution of rainfall and the contribution area ratio was discussed.

Key words

parameters identification, contributing area rate, rate of decrease of runoff height