Temporary dependency of parameter sensitivity for different flood types

Hydrological and climatic data at finer temporal resolutions are considered essential to model hydrological processes, especially for short duration flood events. Parameter transferability is an essential approach to obtain sub-daily hydrological simulations at many regions without sub-daily data. In this study, the objective is to investigate temporary dependency of parameter sensitivity for different flood types, which contributes to research into parameter transferability. This study is conducted in a medium-sized basin using a distributed hydrological model, DHSVM. Thirty-six flood events in the period of 04/12/2006-07/01/2013 in the Jinhua River basin, China, are classified into three flood types (FF: flash flood, SRF: short rainfall flood and LRF: long rainfall flood) by using the fuzzy decision tree method. The results show that SRF is the dominant flood type in the study area, followed by LRF and FF. Runoff simulations of FF and SRF are more sensitive to parameter perturbations than those of LRF. Sensitive parameters are highly dependent on temporal resolutions. The temporary dependency of LRF is the highest, followed by SRF and FF. More attention should be payed to sensitive and highly temporal dependent parameters in a subsequent parameter transfer process. Further study into this result is required to test the applicability.