This study focuses on the spatial structure of rainfall in tropical cyclones (TCs), using global estimates from the Tropical Rainfall Measuring Mission (TRMM) Microwave Imager (TMI), and numerical simulations with the fifth generation PSU/NCAR mesoscale model (MM5). Our goals are 1) to characterize the rainfall structure in TCs, globally and 2) to understand the physical mechanisms that are responsible for the rainfall distribution in TCs. We focus on the effects of the vertical wind shear and dry air intrusion on TC structure and rainfall. Observations have shown that they, among other environmental factors, can affect the TC rainfall and induce asymmetries.

The TRMM data extend from 1 January 1999 to 31 December 2003, and includes more than 4500 instantaneous observations of TCs. A Fourier analysis is used to characterize the azimuthal mean and spatial asymmetries of the TC rainfall, globally and as a function of TC intensity. The analysis is conducted both relative to the storm motion and to the vertical wind shear, defined as the difference between the 200 and 850 mb horizonthal wind averaged over a 600 km disk centered on the TC. The Statistical Hurricane Intensity Prediction Scheme (SHIPS) provides the shear estimates. The rainfall asymmetry relative to the TC motion is located ahead of the TC center in all basins and at all TC intensities. The shear-relative rain asymmetry occurs down-shear left (right) in the Northern (Southern) Hemisphere. The asymmetry amplitudes are greater relative to the shear, implying that the shear has a larger impact. The analysis indicates that large asymmetries in the surface rainfall result from the combined effect of boundary layer (BL) processes (linked to the storm motion) and processes acting on the mid- to upper-troposphere (shear effect). These processes need to be in phase (BL convergence upstream of the main updrafts above BL) for a strong asymmetry to develop.

MM5 simulations of Hurricane Floyd (1999) are conducted to test this hypothesis. The objective is to construct a set of idealized experiments to simulate a variety of shear to motion directions and moisture distributions. Mid-level moisture (800 to 500 hPa) in the storm environment is modified using time varying gaussian shape functions to simulate the intrusion of dry air. The vertical wind shear can also be modified using a similar approach. Initial modeling results agree with the TRMM analysis, showing rainfall maxima down-shear left.

This work sheds light on the physical mechanisms responsible for the TC rainfall structure. Our results can provide relevant information to current rainfall forecasting approaches that often handle storms as symmetric systems.