An appraisal of the value of simulated weather data for quantifying coastal flood hazard in the Netherlands

Estimates of the risk posed by rare and potentially catastrophic weather events are often derived from relatively short measurement records, which renders them highly uncertain. By replacing measurements with much larger sets of simulated weather data, the statistical error can in principle be made arbitrarily small. However, systematic errors in the meteorological model output can easily outweigh the gain in precision. We assess the value of simulated weather data for a particularly demanding task: the quantitative assessment of coastal flood hazard for the Netherlands. In particular, we analyse how (and by how much) the uncertainty in return values of wind stress and coastal water level for return periods of up to 10 million years can be reduced. Based on insights from physics and extreme value theory as well as evidence from data, we argue that simulated weather data are suitable for estimating the shape of the upper tail of the distribution function of stress, even if stress from present-day weather prediction models may be too high or too low. We extend this argument to simulated data of water level along the Dutch coast. As scale and location parameters can be estimated with sufficient precision from relatively short measurement records of water level, we estimate return values from a combination of measurements (for scale/location) and simulated data (for shape). We assess the reduction in uncertainty achieved and discuss strengths and limitations of the approach as well as prospects for further exploitation of simulated weather data to quantify flood hazard.