The extensive damages of extreme rainfall events (EREs) and associated natural disasters on the natural and anthropogenic resources and the enormous economic losses underscore the requirement for developing early warning systems to mitigate the impact of such disasters. However, accurate forecasting of EREs at a regional scale and at higher lead times is challenging due to the uncertainties involved in the model predictions. This study proposes a novel technique to nowcast the heavy and extreme rainfall events using the early signatures of the microphysical evolution of mesoscale convective clouds. The nowcasting method integrates the cloud top temperature (T)-cloud effective radius (re) profile derived using remote sensing methods and logistic regression modelling to estimate the probability for the occurrence of heavy and extreme rainfall events. The capability of the method is demonstrated by nowcasting different recent EREs over the windward slopes of the southern Western Ghats (Kerala, India). The results of the analysis of the T-re profiles of the normal, heavy and extreme rainfall events of August 2018 are significantly distinct and indicates polluted (aerosol-rich) scenario during EREs. The study suggests significant interactions between moisture availability and aerosol concentration during the occurrence of EREs in August 2018, along with their independent effects. The proposed technique shows distinctive competency for nowcasting the EREs at a regional scale with an overall accuracy of 93% and at a lead time of not less than six hours. This study highlights the significance of the aerosol-cloud interactions in the occurrence of EREs of the region and suggests the importance of the aerosol pollution leading to EREs and associated natural disasters.