Physical vulnerability to dynamic flooding: Vulnerability curves and vulnerability indices

Vulnerability analysis is crucial to assess natural hazard risk. Methods for vulnerability assessment include indices as well as vulnerability curves. Vulnerability curves make use of empirical data to show the relationship between the process intensity and the resulting degree of loss on each affected building whereas vulnerability indices are based on a number of indicators representing building characteristics and their surroundings. In the present paper, damage data from two relatively recent torrential events in the European Alps are used to compare results using a vulnerability curve (Beta model) and a physical vulnerability index (PVI). Following the appli-cation of both methods, their strengths and weaknesses are outlined. Vulnerability curves constitute a valuable quantitative method for the assessment of physical vulnerability but, in the present study, they tend to over-estimate damages. On the other hand, vulnerability indices better support the understanding of local-scale damage patterns but they require detailed data and further research on weighting and indicator selection. The study leads to the conclusion that both methods complement each other providing better insights into the physical vulnerability of buildings exposed to torrential hazards. Furthermore, uncertainties associated with the two approaches are related to the required data. Therefore, a sensitivity analysis is carried out showing that process intensity is a key variable for the assessment of vulnerability, whereas, differences in the calculation of the degree of loss based on different building values are less important. Finally, the paper gives clear recom-mendations for improved event and damage documentation and provides an outlook on future needs in vulnerability assessment, including constant updating of both methods based on recent events.

Automated summary

This study compares the results of vulnerability assessment using vulnerability curves and vulnerability indices. It found that vulnerability curves, while valuable, tend to over-estimate damages, while vulnerability indices provide better understanding of local-scale damage patterns but require more detailed data and further research. Both methods complement each other and provide better insights into the physical vulnerability of buildings exposed to torrential hazards.