Assessing climatological trends in daily precipitation extremes in Berlin-Brandenburg, Germany, using 43 years of station-based and reanalysis-based data

Precipitation extremes have caused considerable damages in the Berlin-Brandenburg metropolitan region (Germany) during recent decades. To assess the climatological trends in intensity and probability of occurrence of these events, long-term precipitation data sets are an important prerequisite. Often, such investigations are based on measurements or interpolations of ground station networks. However, due to the high spatial variability of precipitation and its extremes, even relatively dense station networks may not be sufficient to accurately represent small-scale events. The utilization of long-term data, derived from numerical model simulations, has the potential to facilitate an area-wide evaluation. Therefore, we analyzed the spatial distribution of precipitation extremes in Berlin-Brandenburg based on observations by ground measurements and the capabilities of a gridded, reanalysis-based data set for assessment of spatial patterns in changes in seasonal return levels.

The data used consists of 43 years of daily rain gauge measurements by 227 stations of the German Weather Service (DWD) and precipitation data of the Central Europe Refined Analysis version 2 (CER v2, Bart et al., 2024). The CER v2 is a WRF-based dynamical downscaling of ERA-5 for the Berlin-Brandenburg region with a maximum spatial resolution of 2 km. For both data sets we fitted a Generalized Pareto Distribution using a time-dependent seasonal threshold and scale parameter to estimate the 2-, 5- and 10-year return levels at each location (station, grid point). After evaluating the goodness-of-fit at each station the magnitude and change in return levels was compared between both data sets.

The estimated changes in return levels for the CER v2 data correspond relatively well to changes estimated from the DWD stations. However, the CER v2 return levels themselves were on average 10-18% higher. The spatial patterns show an increase in the intensity of the 2-, 5- and 10-year events during summer months in both data sets and a decrease across the region during spring. The spatial variability of the rates of change is particularly high in winter and fall. Overall, the results show that reanalysis-based data could provide an important complement in the assessment of changes in regional precipitation extremes.

Bart, F., Schmidt, B., Wang, X., Holtmann, A., Meier, F., Otto, M., Scherer, D., 2024. The Central Europe Refined analysis version 2 (CER v2): evaluating three decades of high-resolution precipitation data for the Berlin-Brandenburg metropolitan region. metz. https://doi.org/10.1127/metz/2024/1233