Shifting landscapes of risk: Quantifying pluvial flood vulnerability beyond the regulated floodplain

Abstract

Floods are recognized as the costliest type of natural hazard both worldwide and in the United States, with projected increases in frequency and magnitude in the absence of effective adaptation strategies. In the fall of 2018, Hurricane Florence made landfall in southeastern North Carolina, USA, bringing record rainfall and resulting in widespread inundation that impacted many areas outside of the federally designated Special Flood Hazard Area (SFHA). Much of this flooding was from inland pluvial inundation, which is an understudied component of coastal risk and vulnerability assessments primarily due to the scarcity of infrastructure data and historically lower flooding recurrence rates. This has resulted in severe damages in areas that residents and local officials considered at low risk from flooding. Using nearly-coincident high-spatial, high-temporal resolution CubeSat satellite imagery, we quantified the areal extent of post-Hurricane Florence floodwater within and beyond the 100-year floodplain (SFHA) and the proportion of residential structures exposed to flooding within an eight-county study area. We propose a novel approach to estimate flood risk resulting from this singular event (termed an actualized risk index) when compared to a published empirical model of vulnerability. We show that 24.3% of detected floodwater was outside the 100-year floodplain, 43.4% of exposed residential structures are outside the 100-year floodplain, and communities of highest vulnerability are not only along the coast but also inland along the Cape Fear, Northeast Cape Fear, Trenton, and Neuse Rivers. This suggests that the SFHA may not adequately show the spatial distribution of pluvial flood risk in riverine areas, and that misunderstanding of this risk has led to a pattern of development in which houses have a higher than expected risk of flooding. Moreover, this additional flood risk may disproportionately affect lower-income residents of these largely rural areas. These results have important implications in light of recent policy guidance in southeastern USA states that mandate that predictive coastal vulnerability assessments to sea level rise be conducted relative to 100-year SFHA zones.

Keywords: CubeSat imagery; Inundation; Managed floodplains; Vulnerability.