Changes in Future Flash Flood-Producing Storms in the U.S.

Abstract:

Flash floods are high-impact events that can result in massive destruction, such as the May 2010 flash floods in the south-central U.S. that resulted in over 2 billion dollars of damage. While floods in the current climate are already destructive, future flood risk is projected to increase based on work using global climate models. However, global climate models struggle to resolve precipitation structure, intensity, and duration, which motivated the use of convection-permitting climate models that more accurately depict these precipitation processes on a regional scale due to explicit representation of convection. These high-resolution convection-permitting simulations have been used to examine future changes to rainfall, but not explicitly floods. This study aims to fill this gap by examining future changes to rainfall characteristics and runoff in flash flood-producing storms over the U.S. using convection-permitting models under a pseudo-global warming framework. Flash flood accumulated rainfall increases on average by 21% over the U.S. in a future climate. Storm-generated runoff increases by 50% on average, suggesting increased runoff efficiency in future flash flood-producing storms. In addition to changes in non-meteorological factors, which were not explored in this study, increased future runoff is possible due to the 7.5% K^-1 increase in future hourly maximum rain rates. Though this median change in rain rates is consistent with Clausius-Clapeyron theory, some storms exhibit increased future rain rates well above this, likely associated with storm dynamics. In addition to presenting results on flash flood-producing storms across the U.S., we will also show a regional analysis focused on changes in cool-season atmospheric river events in California using the same convection-permitting simulations.