Passage of Tropical Cyclones over Mountainous Islands Part I: Control Experiment

The passage of a tropical cyclone (TC) over a mesoscale mountainous island, such as Puerto Rico, often brings heavy rainfall which produces flooding and landslides. Factors that affect the amount and distribution of this type of orographic rainfall in this region are not well understood. This study investigates the impact of microphysics (MP) and planetary boundary layer (PBL) schemes in a TC simulation using the Advanced Research Weather Research and Forecast (ARW) model. The numerical mesoscale ARW model was adopted to conduct a study of Hurricane Jeanne's (2004) passage over the island. Sensitivity experiments using four different MP and three PBL schemes were performed. Each sensitivity experiment was represented by a combination of a single MP (e.g., Lin et al. scheme, Eta scheme, WRF Single-Moment 6-class scheme, and the New Thompson et al. scheme), and a PBL scheme (e.g., Yonsei University scheme, Mellor-Yamada-Janjic scheme, and BouLac PBL).

Results show strong consistency for the cyclonic track among all experiments with an average distance difference of ~60 km and significant landfall time difference of ~4 hours ahead of observations. Rainfall distribution was well represented, with maxima on the southeastern and higher mountain regions. EXP3 (WSM6-YSU) produced the best simulation in terms of both rainfall distribution and locations with high rainfall magnitude. However, there were sizeable differences between the model cyclone's wind intensity and minimum sea-level pressure at model landfall and those of the observed storm. EXP6 (ETA-MYJ) and EXP7 (WSM6-MYJ) produced a more realistic sea-level pressure between 987 and 995 mb, compared with 991 to 995 mb from observations. EXP6 produces the lowest wind speed of less than 85 mph. In conclusion, the simulation using the WRF Single-Moment 6-Class MP scheme with the Mellor-Yamada-Janjic PBL scheme (EXP7) appears to be strongly comparable with observations, and is the best choice for further investigations of the production of heavy orographic rainfall. Future study will examine some control parameters and common ingredients, such as Convective Available Potential Energy (CAPE), Precipitation Efficiency (E), and the impact of mountain height variations, to help understand the dynamical and physical processes associated with orographic precipitation induced by the passage of a TC.