Preliminary results on the simulation of the 1999 Orissa supercyclone using a GCM with a new boundary layer code

Abstract

We present here preliminary results from the simulation of the Orissa supercyclone using a new AGCM code (named Varsha) written as part of a NMITLI project. The simulation is initialized at 00 UTC, 26 October 1999, using ECMWF T-106 initial conditions. The control run is made using the Varsha code at a T-80 resolution with a standard Monin-Obukhov boundary layer code incorporating a gustiness factor. With the horizontal resolution improved to 120 spectral modes with a 78 km grid spacing, and a new boundary layer parameterization at low winds, the code shows substantial improvements: the maximum error is reduced from 350 to 234 km at 36 h after initialization, 310 to 34 km at 48 h, and 410 to 55 km at 96 h. It is suggested that part of the explanation for this improvement lies in the improved estimation of surface forces and torque in the new boundary layer code. The role of torque is particularly interesting as the major contribution to it comes from the outer regions of the cyclone where the winds are relatively low but the area on which the surface force acts and its moment arm are both high. Intriguingly the higher surface forces arise also from the higher winds predicted by the new code. An interesting finding is that, on both track and minimum pressure, the improvement due to higher resolution is greater with the new boundary layer module. Further analysis is necessary to assess the effect of other eddy fluxes (sensible heat, moisture) on cyclone track prediction.