Measurement Approach to Develop Flood-Based Damage Fragilities for Residential Buildings Following Repeat Inundation Events

Abstract

Floods account for the highest annual average losses from natural hazards across the United States, and the occurrence of repeat flood inundation events in United States communities is increasing. Distinguishing damages caused by distinct flood events in a community that has experienced repeated flooding is difficult, and best practices for repeat flood metrology are needed to better inform and validate flood damage models. This paper presents a longitudinal methodology for measuring impacts from repeated flood inundation through a case study of buildings in Lumberton, North Carolina, where major flood events occurred in 2016 and 2018. Sources of uncertainty encountered in flood damage assessments are presented to inform best practices for future investigations of repeat flood events. A novel initial state parameter is introduced for accurate damage characterization for a repeat flood event. This paper presents the first analysis of statistical distributions of damage conditioned on flood depth for a set of buildings that have been flooded in two consecutive events, and the results show how floods with similar intensities occurring in the same area at two different times can exhibit differing distributions. Flood damage data sets for the two floods are combined to derive flood damage fragilities, and we propose the creation of a flood damage database by aggregating data from various flood events across the United States to enable more robust fragility functions that can be applied across geographies and flood events.