Abstract
Floods caused by hurricane storms are responsible for tremendous economic and property losses in the United States. To minimize flood damages associated with large hurricane-season storms, it is important to be able to predict streamflow amount in response to storms for a range of hydroclimatological conditions. However, this is challenging considering that streamflow response exhibits appreciable variability even for hurricane-season storms that deliver similar precipitation amounts. As such, better estimates of event responses require refined understanding of the causes of flood response variability. Here, a physically based, distributed hydrologic model and supporting hydrologic datasets are used to identify and evaluate dominant hydrologic controls on streamflow amount variability. The analysis indicates that variability in flood response in the Lake Michie watershed is primarily driven by antecedent soil moisture conditions near the land surface and evapotranspiration during postevent streamflow recession periods, which in turn is a function of precipitation history and prevailing vegetation and meteorological conditions. Presented results and ensuing analyses could help prioritize measurements during observation campaigns and could aid in risk management by providing look-up diagrams to quickly evaluate flood responses given prior information about hurricane storm size.
