Role of PBL and air-sea flux parameterization schemes in the forecast of super cyclone Amphan and ESCS Phailin in the cloud-resolving scale using WRF-ARW model

The present study evaluates the role of turbulence mixing in the boundary layer and surface roughness schemes through parameterization of the planetary boundary layer (PBL) and air-sea flux (ASF) schemes, respectively, in the prediction of Super Cyclonic Storm (SuCS) Amphan 2020 and Extremely Severe Cyclonic Storm (ESCS) Phailin 2013 over the Bay of Bengal region. This study utilized a high-resolution Advanced Research version WRF (WRF-ARW) modelling system with a moving-nested domain in a cloud-resolving scale about 1.667 km horizontal resolution. Six simulations were conducted with two PBL (YSU non-local and MYJ local) schemes and three air-sea flux (FLUX0, FLUX1, and FLUX2) schemes. In the last the time-varying Sea Surface Temperature (SST) was also updated for those simulations having over-predictions in the maximum surface wind (MSW). The model predicted track, intensity, and structures were validated with the Indian Meteorological Department best-fit track data, Doppler Weather Radar (DWR), and Cooperative Institute for Research on Atmosphere (CIRA) multiplatform satellite datasets. Results suggested that model simulations provided a better forecast in MSW using the MYJ-FLUX2 experiment with mean absolute errors of about 5.3 m/s, followed by the MYJ-FLUX1 experiment. The simulated rapid intensification in both cases (Amphan and Phailin) was well captured in the MYJ-FLUX1 and MYJ-FLUX2 experiments. The time-varying SST experiments provided less intensity compared to without SST experiments and showed a positive impact on the forecast of MSW in the first two days with the YSU-FLUX1 experiment. For a better understanding about under-prediction and over-prediction during the entire simulation period were presented and discussed in terms of microphysics latent heating, and divergence. Storm structures in terms of spatial wind speed and vertical structure of temperature anomaly suggested that simulations were varying by changing PBL and ASF schemes. Overall, the YSU-FLUX1 experiment showed a better prediction in terms of track of the storms, with a mean track error of about 63 km. This study suggested that the high horizontal resolution about 1.667 km using YSU-FLUX1 with SST in the WRF model provided a better representation of the intensity and storm structures of ESCS Phailin and SuCS Amphan.