Detailed microphysics modeling of the intense precipitation episode IOP7a observed during HYMEX experiment : study of the impact of pollution

The French coastline in the Mediterranean Sea is affected by heavy rainfall episodes especially in autumn. Cevennes – Vivarais, which is part of the Massif Central Mountains, is one of the affected regions. The associated heavy precipitation episodes (HPE), namely “Cevenols”, can cause natural disasters with important economic damages and life losses. The prediction of such episodes by Numerical Weather Prediction (NWP) models has been significantly improved; uncertainties remain though, regarding their occurrence and strength. The improvement of microphysical parameterizations in NWP models is one key-component for the reduction of forecast errors. The aim of this study was provide a better understanding of the microphysical processes that govern HPE and their interaction with atmospheric aerosol particles (APs) by exploiting observations from the HYMEX research program.The present study focused on the HPE from the HYMEX Intense Observation Period (IOP) 7a, whose observations were assessed and compared with modelling results from the bin-resolved microphysics scheme DEtailed SCAvenging Model (DESCAM, Flossmann and Wobrock; 2010) with 3D dynamics. This research model uses a detailed representation of the APs. Observations from ground-based instruments, as well as in-situ measurements were used for the evaluation of the model’s performance. The ground-based dataset consists of X-band Radars, Micro-Rain Radars (MRR), disdrometers, but also a rainfall reanalysis by rain gauges and operational radars (Boudevillain et al. 2016). Moreover, hydrometeor probes and the 95GHz cloud radar RASTA provided observations on-board of the French research aircraft Falcon-20.The role of pollution on the development and evolution of the HPE of IOP7a was investigated, as well. Considering that the highest AP concentrations were observed during IOP7a, the followed strategy was to perform model simulations by using less polluted observed AP spectra with lower total number concentrations. The results showed that the initial AP concentration influences the spatial distribution and quantity of rainfall, as well as the vertical properties of the rain water content and the ice water content of the precipitating cloud system. For the studied cases, with increasing the initial number concentration of APs, the total rain amount was decreased. Finally, the present study revealed a critical role of the model’s large-scale configuration necessary to correctly represent the dynamics.