Numerical investigation on the appearances of “Spatial Representativity” in rainfall-runoff processes

Abstract

To explore how the spatial representativity of discharge develops, we artificially created spatially distributed discharge data with a distributed rainfall-runoff model that can separately calculate fast, intermediate and slow discharge components. Firstly, we generated spatially distributed discharge data using a distributed rainfall-runoff model. Next, we investigated the relationship between the normalized specific discharge and the subcatchment area based on numerical experiments with different discharge conditions. In addition, we separated the generated river discharge into three discharge components with different recession characteristics and investigated the relationship between normalized specific discharge and subcatchment area for each component. Then we identified discharge components that contribute to the spatial variability of river discharge. As a result, we found that the spatial variability of river discharge originates from the fast discharge component during higher discharge, whereas the slow discharge component dominates during lower discharge. These results suggested that spatial variability in river discharge should depend on the discharge conditions of the river, which has not been explored previously.