The paper presents a numerical modeling of convective events in the Nizhny Novgorod region, taking into account different aerosol loads in order to improve the quality of thunderstorm forecast. The center of the considered region is the city of Nizhny Novgorod. This is an industrial metropolis located in the central part of the East European Plain. The mesoscale weather forecast model WRF-ARW was used as a basic numerical modeling tool, as well as the parameterization of electric processes was used to determine the electric parameters of thunderclouds. A series of numerical experiments to account for various levels of aerosol load was performed. For this purpose, the parameterization of microphysical processes was used, which explicitly takes into account the activation of aerosols in the form of condensation of cloud droplets or ice particles and which allows us to set an arbitrary initial distribution of aerosols. Thus, by varying the input distribution of aerosols, numerical experiments were conducted in the work to simulate thunderstorms taking into account different aerosol loads; various scenarios were compared and their results were verified; optimal modeling parameters were identified in the context of improving the quality of thunderstorms forecasting.