Increasing flood losses over the last decades emphasize the need towards significantly improved and more efficient flood risk management. One key requirement is reliable risk assessment in conjunction with consistent flood loss modeling. Current risk assessments and flood loss estimations for Europe are until now based on regional approaches using deterministic depth-damage function and do rarely report associated uncertainties. To reduce these shortcomings, we present the results of a novel, consistent approach based on the Bayesian Network Flood Loss Estimation MOdel for the private sector (BN-FLEMOps).The dataset is consistent in terms of the input data used to drive the model and because we use the same vulnerability model to derive the flood loss estimation. Essential inputs for any flood loss estimation are hazard (usually water depth), asset (value of objects at risk) and flood experience parameters. The hazard input was given by a European inundation scenario for a continent-wide flood with 100 years return period (Alfieri et al., 2014). Asset values were computed following the the approach by Huizinga et al. (2017) and the flood experience was derived using the database of the Dartmouth Flood Observatory (DFO) (Brakenridge, 2018).The provided dataset comprises a flood loss estimation covering the European continent, spatially aggregated on level three of the standard territorial units for statistics NUTS-3 (https://ec.europa.eu/eurostat/web/nuts/background). The data set reports the summary statistics as a flood loss distribution per NUTS-3 region in 10 per cent quantile steps. The flood loss estimations are given in Million Euro. In addition, the NUTS-3 code, the underlying version of the standard territorial unit and the associated NUTS level are provided. This data publication includes the exact dataset as reported in Lüdtke et al (2019) [filename_1], which is single model application. Supplementary, we provide the summary statistics from an ensemble of 1000 model runs to account for the inherent variability of the probabilistic model [filename_2]. The ensemble model application reports the same statistical measures as the single model application (flood loss distribution per NUTS-3 region in 10 per cent quantile steps), but the given numbers show the median of 1000 model runs for each quantile step (10%, 20%, … 90%).The dateset is provided as a multi-polygon vector. All polygons that belong to the same standard territorial unit share the same attributes. The spatial reference system is defined by EPSG:4326. We provide two formats, (I) an ESRI shape file and (ii) a GEOjson representation. For more information please refer to the associated data description.
Climate change manifests in terms of changing frequency and magnitude of extreme hydro-meteorological events and thus drives changes in urban flood hazard. Flood risk oriented urban planning is key to derive smart adaptation strategies, strengthen resilience and achieve sustainable development. 3D city models offer detailed spatial information which is useful to describe the exposure and to characterize the susceptibility of buildings at risk.This web-based application presents the 3d-city flood damage module (3DCFD) prototype which has been developed and implemented within a pathfinder projected funded by Climate-KIC during 2015-2016. The presentation illustrates the results of the 3DCFD-module exemplarily for the demonstration case in the City of Dresden. Relative damage to residential buildings which results from various flooding scenarios is shown for the focus area Pieschen in Dresden.The application allows the user to browse through the virtual city model and to colour the residential buildings regarding their relative damage values caused by different flooding scenarios. To do so click on 'Content', then on the brush-icon next to 'Buildings' and select a certain style from the drop-down menu. A style represents a specific combination of loss model and flooding scenario. Flooding scenarios provide spatially detailed inundation depth information according to different water stages at the gauge Dresden. Currently two flood loss models are implemented: a simple stage-damage-function (sdf) which related inundation depth to relative loss and the 3DCFD-module which uses additional information about building characteristics available from the virtual city model. A click on a coloured building will display additional information. The loss estimation module has been developed by the German Research Centre for Geosciences (GFZ), Section Hydrology. The web-application has been developed by virtualcitySYSTEMS GmbH. The data consisting of flood scenarios, a virtual 3D city model, and a terrain model were provided by the City of Dresden.