Europe - Pluvial Flood Loss Questionnaire - GFZ

The online questionnaire developed within the H2020 Insurance project builds on previous work to collect loss information using computer aided telephone interviews and online surveys. The survey campaigns always focused on a single event and used a questionnaire with about 180 questions regarding aspects of hazard, exposure, vulnerability and residential building and content losses. The datasets collected in various surveys after major floods in Germany have been successfully used to develop and validate complex flood damage models, e.g. Schröter et al. 2014, to better understand flood damage processes Rözer et al. 2016, and build improved models e.g. Rözer et al. (2019). Further it has been shown that the predictive performance of vulnerability models can be improved when models are updated with location specific information (Lüdtke et al. 2019). Within the H2020 project the questionnaire approach has been adapted for the Danube region with a focus on pluvial floods. In a first step, those questions have been selected from the comprehensive questionnaire, which are relevant for the pluvial flood vulnerability model for residential buildings. Hence, the number of questions was reduced to 28. From the answers to these specific questions the input variables for the pluvial flood vulnerability model can be quantified, and model predictions can be compared to actual (observed) loss. These questions have been grouped into hazard characteristics and damage, building characteristics, and personal information. Hazard characteristics include information about the inundation depth, the duration, and possible pollution. The amount of damage is queried separately for building structure and household goods. Regarding building characteristics, information about the building type, usable floor space, heating systems, use of basement and other details are collected. Personal information covers aspects of household size and location for geocoding of the records.