Runoff recession features in an analytical probabilistic streamflow model

Abstract

The development of run-of-river hydropower requires the flow duration curve (FDC) to estimate the electricity production. As obtaining an FDC based on gauging station streamflow requires significant cost and time (i.e., more than several years), it is important to predict the FDC in the absence of streamflow data. An analytical probabilistic model of streamflow dynamics (p(Q) model) can calculate the FDC, which requires several rainfall and runoff recession parameters. To estimate the recession parameters in the absence of discharge data, we selected two approaches: the geomorphic recession flow model (GRFM) and geological recession approach (GRA). The GRFM and GRA associate the river network and geology, respectively, with runoff recession features. This study aims to comprehend the characteristics of the GRFM and GRA employed in the p(Q) model by targeting 57 basins throughout Japan. We found different performance trends between the GRFM and GRA that corresponded to specific basin characteristics. While the GRFM performed well in high precipitation basins, the GRA performed well in basins with abundant unconfined groundwater contributions. These results agree with the features in the derivations of the GRFM and GRA. In addition, we found a complementary relationship between the GRFM and GRA based on the baseflow index (BFI), which suggests that the complementary relationship is one method to improve the performance of the p(Q) model in the absence of streamflow data.