A multi-year sensitivity analysis of microphysical parameterization in convection-permitting simulations over the Black Sea Basin

The Black Sea Basin (BSB) is one of the climate change hot-spots where intense atmosphere–sea interactions and complex topography shape temperature distribution, precipitation regimes, and extreme weather events. Accurate simulation of these regional climate patterns in high-resolution atmospheric models requires microphysical parameterization schemes that realistically represent hydrometeor evolution and temperature tendencies. In this study, we evaluate a suite of convection-permitting WRF simulations at 3 km resolution to identify the microphysics scheme that most reliably reproduces temperature, precipitation, hail occurrence, and snow cover across three meteorologically contrasting years over the BSB. Annual sensitivity simulations for 2008 (a dry year), 2010 (a warm SST year), and 2017 (a wet year) using the Goddard, Milbrandt, NSSL, WSM7, WDM7, and Thompson schemes show that the model reproduces the spatial and seasonal patterns of Tmax and Tmin across the basin. Seasonal temperature patterns are robustly captured by all schemes, with biases generally limited to 2–3°C and partly shaped by mountainous terrain. Similarly, precipitation evaluations indicate that the model represents basin-wide spatial distributions and seasonal cycles with high fidelity, with remaining biases of 6–10 mm/day over the Caucasus and eastern Anatolian highlands, reflecting the observational limitations in complex topography. RMSE metrics show that the microphysics schemes vary in performance, with the Milbrandt scheme standing out for its consistently year-to-year uniform behavior. It maintains low Tmin RMSEs of 1.30–1.40°C and precipitation RMSEs of 0.87–1.15 mm/day across all three years, indicating that it reproduces temperature and precipitation fields with stable accuracy under different meteorological conditions. Hail and snow cover evaluations further reinforce this result. The Milbrandt scheme adequately represents the spatial distribution and spring/early-summer evolution of hail-prone areas across Türkiye. Simulated spring snow cover closely matches satellite observations over the Upper Euphrates Basin, located in eastern Türkiye. Additionally, preliminary results from ERA5-driven simulations strengthen these findings by realistically reproducing the long-term characteristics of temperature, precipitation, and snow cover over the BSB. Overall, the Milbrandt scheme serves as the most suitable microphysical parameterization option for the upcoming long-term fully coupled atmosphere–ocean simulations aimed at improving the representation of two-way air–sea feedbacks, which are crucial for understanding climate processes over the BSB.

Acknowledgment: The numerical calculations reported in this paper were fully performed using the EuroHPC Joint Undertaking (EuroHPC JU) supercomputer MareNostrum 5, hosted by the Barcelona Supercomputing Center (BSC). Access to MareNostrum 5 was provided through a national access call coordinated by the Scientific and Technological Research Council of Turkey (TÜBİTAK). We gratefully acknowledge BSC, TÜBİTAK, and the EuroHPC JU for providing access to these resources and supporting this research.