In this study, numerical simulations are performed on the formation of Typhoon Robyn (1993) in the western North Pacific (WNP) during the Tropical Cyclone Motion (TCM-93) field experiment. One objective is to assess the role of tropical mesoscale convective systems (MCSs) in the development. This TCM-93 case is chosen because additional in-situ and aircraft data are available to define the structure of several MCSs associated with the formation process, and for model verification. This study identifies the mesoscale processes responsible for Typhoon Robyn’s formation where the background vorticity associated with the monsoon trough and passages of tropical disturbances are both essential factors.

The Pennsylvania State University (PSU)-National Center for Atmospheric Research (NCAR) mesoscale model MM5 version 3 with nested grids is used for the simulations. The model is initialized with large-scale analyses from the Navy Operational Global Atmospheric Prediction System (NOGAPS). In the control experiment, the Betts-Miller cumulus parameterization is activated on both the coarse grid with 81-km resolution and the fine grid with 27-km resolution. It is found that many of the features associated with the formation of Typhoon Robyn could be simulated in the 27-km resolution grid. These features include the initial near-surface convergence and subsequent frontal line, and then two consecutive convective events that led to the intensification to a tropical depression. However, great sensitivity of the simulations to the model physics is also found. A faster intensification (and closer to the observed intensity) was obtained using the standard Betts-Miller cumulus parameterization compared with the Kain-Fritsch, Grell, or a modified Betts-Miller scheme with a ‘tropical’ reference profile, because the standard Betts-Miller technique is able to produce a larger amount of convective rainfall. The simulations are also sensitive to the choice of boundary layer (BL) parameterization. With certain parameterizations such as the Eta scheme, a false vortex is spun up in the model. In addition, a more sophisticated explicit moisture calculation such as the Reisner scheme, which predicts the number concentration of snow and graupel is indeed essential for predicting the formation of Typhoon Robyn. MM5 simulations with higher resolutions will be performed to better resolve the cloud systems associated with the convective events before Robyn’s formation to gain further understanding of their contributions in the vorticity budget of the cyclone.