Characteristics of extreme rainfall from warm clouds associated with a mesoscale convective system: sensitivity of different microphysical and cumulus parameterization schemes

Intense convection triggered by the sustained moisture flow and the veering winds on October 7, 2021, over Gadanki, located in southeast India, produced 92% of the October rainfall in less than 3 h. The 850 hPa relative vorticity is 0.8 x 10(-5) s(-1), indicating weak synoptic forcing during extreme rain. Precipitation echoes observed by L-band radar show up to a maximum altitude of 6 km, indicating that the rainfall is associated with warm rain processes. The drier mid-troposphere on October 7, retarded the vertical growth of convection to deeper altitudes. Raindrop size distribution during the event differs from other October rainfall events with a smaller mass-weighted-mean-diameter (D-m) and larger normalized scaling parameter for concentration (N-w). Compared to the shallow rain D-m reported in the literature, the October 7, event Dm is larger and possesses continental rain characteristics (D-m > 2 mm). The enhanced collision-coalescence due to the strong turbulence associated with the event increased the raindrops' size and Dm. Verification scores and event total rainfall comparisons of Weather Research and Forecasting (WRF) simulations and observations indicate that the Betts-Miller-Janjic (BMJ) (1994) cumulus scheme performed better than the Kain-Fritsch (KF) (2004) scheme. Though the BMJ scheme can predict rain intensity and duration better than the KF scheme, it differs in vertical structure of the precipitation observed from the L-band radar and background atmosphere from ERA5. None of the combinations of cumulus (BMJ and KF) and microphysical (Kessler, WSM6, WDM6, and Thompson) schemes in the WRF model can simulate the observed warm rain extreme event.