| dc.description.abstract | Cumulus convection plays a central role in most regional precipitation predictions, where the representation of cumulus convection, generally called cumulus parameterization, is almost always at the core of efforts to numerically model the atmosphere. The new Weather Research and Forecasting (WRF) model, like other numerical models, has several options for cumulus parameterization schemes (CPS). This study focuses on how different CPSs in the WRF model simulate tropical cyclone convection. Typhoon Fanapi (2010), which brought very heavy rainfall (1131 mm) to the south plain of Taiwan, was selected for this thesis as a study case. A three-day simulation was set up during a time of high precipitation with four CPSs: the Kain–Fritsch (KF), Betts–Miller–Janjic (BMJ), Grell–Devenyi ensemble (GD), Grell three-dimension ensemble (G3D), and a no-scheme. Each of these have their own limitations; however, the results from the KF scheme is likely the best among them, showing good track, precipitation, and vertical velocity location simulations. GD and G3D also exhibited acceptable simulation results. However, the BMJ model did not successfully predict this TC. To further understand the precipitation process in southwestern Taiwan, the KF scheme was chosen as a control run.
Because clouds and their associated physical processes strongly influence the coupling between the atmosphere and oceans (or ground) through modifications of the radiation and planetary boundary layer processes, further investigations applying a new data set of land use and roughness length were conducted and reveals that the application of these data improves the location and intensity of convection in this particular case.
| en_US |