Key points are not available for this paper at this time.
Pluvial flood mapping is a common way to identify locations at risk of being flooded in connection with an extreme short-duration rainfall. As rainfall extremes are expected to intensify in a warmer climate, more and more cities and municipalities are performing pluvial flood mapping, in Sweden and elsewhere. Often a rainfall with a return period of 100–200 years is used in the mapping, and the main question underlying this study is to which extent such conventional flood mapping can provide useful indications also of the consequences of a far more severe rainfall. To investigate this question we use the worst urban flooding in Sweden to date as a case study. In Aug 2021, a record-breaking rainfall with a return period > 1000 years hit G¨avle city on the east coast, resulting in widespread flooding and massive damages. We reconstruct the space–time evolution of this event by combining different rainfall observations and using them as input to a high-resolution hydrodynamic model. The reconstruction was overall successfully validated using municipal and citizen observations. Different representations of the rainfall had limited impact on the average flood response, but distinct local differences were found. Comparing the reconstructed flooding with the results from a conventional 100-year flood mapping revealed that overall the worst flooded locations during the actual event were also pointed out in the conventional mapping, albeit with some differences. We recommend to always include assessments of fluxes and water velocities as well as sensitivity tests of very extreme rainfalls in pluvial flood mapping.
Olsson et al. (Sat,) studied this question.