Flood depth-damage and fragility functions derived with structured expert judgment

Abstract

Depth-damage relationships of residential typologies have an important bearing on the reliability of flood risk assessments. Yet, the information for creating adequate functions is frequently missing in places where they are needed the most: developing countries where floods inflict severe losses. To overcome the lack of site-specific functions, modelers usually adopt functions from foreign typologies, extrapolate from limited flood damage data, and conduct expert-elicitation surveys. The latter, although not exempt from challenges, are a helpful resource in situations in which there are sufficient experts available. The classical model is a structured expert judgment technique which is one of the most used because it minimizes the number of subjective elements in the ranking of experts’ experience and its results can be compared with similar studies. This paper presents depth-damage and fragility functions developed with the classical model conducted with experts with extensive knowledge in the flood damage mechanism of typologies in northeast Argentina—a region under a severe flood hazard. The depth-damage functions were partially validated with damage data and exhibited good agreement, also they are consistent with the fragility functions. Hence, these functions can be helpful for probabilistic flood risk assessments, provided that the underlying assumptions are deemed acceptable.