Implementation and validation of the RegCM5 regional climate model on a Central European domain for extreme precipitation conditions

Due to climate change, extreme precipitation events have become increasingly common over Europe, resulting in flash floods such as in Germany in 2021 or Spain in 2024. Regional climate models (RCMs) are key tools to project the mitigation and adaptation challenges regarding these events in the 21st century. However, to provide accurate results, it is necessary to find the optimal modelling setup for the research domain and validate the simulations against reference datasets of past measurements.

In this study, the validation of several regional climate model simulations using the new RegCM5 RCM with a horizontal resolution of 10 km has been carried out. The model was set up with different parametrization scheme combinations for each run to assess which combination most accurately reproduces the monthly temperature and precipitation fields of the reference datasets. To analyse the model performance in reproducing extreme precipitation, the monthly frequency of days with precipitation over 10 mm (RR10) was also validated.

The simulation period is short, i.e. 2010–2011, with a spin-up year of 2009. However, the period suits the validation purposes as it involves both an extremely wet and dry year with 2010 holding the record for the highest, and 2011 the lowest sum of yearly precipitation in Hungary. The initial and boundary conditions for temperature, geopotential, specific humidity, horizontal wind components and sea surface temperature were obtained from the ECMWF ERA5 reanalysis dataset. HuClim served as the reference for Hungary, which is a quality-controlled, gridded and homogenized database on a horizontal resolution of 0.1°, whereas for the rest of the domain the measurement-based, gridded E-OBS dataset with the same horizontal resolution was used.

The early results for the Hungarian part of the domain show that the simulated temperature fields are acceptable for almost all the combinations. However, there are substantial differences between the different set ups in the monthly sum and extreme precipitation, for example, the Tiedtke convection scheme provides better results for 2010, whereas the Kain-Fritsch convection scheme is more accurate in 2011.