Modeling Flood Risk for the Insurance Industry

Dag Lohmann
Risk Management Solutions Ltd
Abstract:

Recent major flood losses raised the awareness of flood risk worldwide and flood insurance penetration has been anticipated to grow during the upcoming years. In order to serve the growing needs of insurers and reinsurers to adequately assess and price their flood risk in Europe, Risk Management Solutions (RMS) has decided to take the challenge of building a Europe-wide flood model over the next few years. This is done in three major steps. The first step includes building a Pan-European Rainfall event set using a Monte Carlo analogue technique that allows us to correlate flow, and hence flooding, within different rivers in Europe at both high and low return periods under current climate conditions. The second step includes building a high resolution (25 meter) flood / hydrology model for river-flood as well as off-floodplain flooding forced by rain derived using the stochastic pan-European rainfall event set. This allows consistent event correlation between models for different areas of Europe. Coupled to these hazard models are models that define the relationship between hazard and average flood damage, as expressed by a modelled mean damage ratio and a coefficient of variation. A financial module then calculates insured losses for various financial perspectives. I will present the various techniques used in the modeling (e.g. calibration, validation, sampling, etc.) as well as the data sets that were used as forcing variables, boundary conditions and model parameters, among them GSWP/ISLSCP, ERA40, and various other data sets from NASA and NOAA.