Spatiotemporal Properties of Floods and Extreme Hydro-Climatological Characteristics for Large Reservoirs in Missouri River Basin

Abstract:

Traditional approaches to flood risk assessment are typically indexed to an instantaneous peak flow event at a specific recording gage on a river, and then extrapolated through hydraulic modeling of that peak flow to the potential area that is likely to be inundated. However, property losses tend to be determined as much by the duration and volume of flooding as by the depth and velocity of inundation. We argue that the existing notion of a flood risk assessment and consequent reservoir operations based on just the instantaneous peak flow rate at a stream gage needs to be revisited, especially for floods due to persistent rainfall (>30 day duration). In this study, a hydro-climatologic framework is presented to cluster the flood types based on peak, volume, duration and timing conditional on large-scale atmospheric teleconnections. A joint hierarchical clustering and wavelet spectrum analysis is proposed to identify similarities in spatial and temporal scales for large reservoirs of Missouri River Basin. The results demonstrate that the short and long duration floods occur in different times of the years and are connected separately to specific large scale atmospheric teleconnections processes.