25th Conference on Hurricanes and Tropical Meteorology

Monday, 29 April 2002
Tropical Cyclone Intensity and Quantitative Precipitation Forecasting
Shuyi S. Chen, Univ. of Miami/RSMAS, Miami, FL; and M. Lonfat, J. E. Tenerelli, R. F. Rogers, and F. M. Horsfall
The objective of this study is to understand the physical processes governing the structure and distribution of precipitation associated with tropical cyclones and, ultimately, to improve storm intensity and quantitative precipitation forecasts (QPF). To resolve the hurricane inner core structure, crucial in intensity forecasting, the model horizontal resolution needs to be at 1-2 km. In this study, we have developed a storm-following nested-grids modeling system in the PSU/NCAR MM5, which allow us to conduct 5-7 day long simulations at grid spacing of 1.67 km. The impact of model grid resolution on hurricane intensity and precipitation forecasts is investigated. We begin by developing a model evaluation and validation procedure using the available rainfall observations, including rain gage network data and the NEXRAD ground based radar-derived rain rates over the coastal regions, the NASA TRMM satellite TMI/PR precipitation estimates, and the NOAA airborne radar data over the ocean. MM5 simulations of Hurricanes Floyd (1999) and Bonnie (1998) are compared with both the GFDL hurricane operational forecasts and the observations. The model results show that the high grid resolution improves the model simulated eyewall structure significantly, which generally produces a more realistic precipitation distribution near the inner core region where the most intense rainfall is observed. The cumulative storm-total rainfall is less sensitive to model grid resolution and storm intensity. To further explore the impact of model physical parameterizations on the storm intensity and QPF, we conduct a series numerical experiments using various surface heat flux and microphysics parameterizations. These simulations show that the modeled hurricane intensity is most sensitive to the surface heat fluxes and less sensitive to the different microphysics parameterizations.

Supplementary URL: