A spatial analytical approach for evaluating flood risk and property damages: Methodological improvements to modelling

Abstract

Empirical evaluations of flood damage from past events are key for identifying the most important factors influencing household vulnerability to floods. However, creating accurate assessments of flood risk with observed records is challenging because of issues associated with data quality, analytical approach, and the successful integration of different information and measurement schemes into one comprehensive model. This paper reviews these issues and proposes ways to overcome them using revised research protocols and spatial analytical methods and techniques. Recommendations are then applied to an empirical evaluation of residential property damages resulting from a major coastal storm in the Houston area, Texas. Results indicate that a parcel‐based, spatial approach to modelling flood risk yields a more detailed and accurate assessment of impacts with an explained variance 2 to 4 times greater than other observation‐based flood damage models available in the literature. Structural improvements to analytical approach and measurement also made it possible to derive direct measures of flood risk at neighbourhood scales and assess the influence of specific land development decisions with respect to major disaster events. Such improvements to flood risk modelling can help reduce the uncertainty of damage estimates and support effective decision‐making in insurance, risk management, and community planning.