The Distribution and Structure of Precipitation in Hurricane Georges: Observations from TRMM, SSM/I, and GOES

It is recognized that tropical rainfall plays a vital role in the global energy and water cycle and that tropical disturbances (waves, depressions, and cyclones) are among the most prolific rainfall producers. The latter was highlighted during 1998 when heavy rainfall from Hurricanes Mitch and Georges caused devastation in the Caribbean, Central America, and the Southern United States. Unfortunately, accurate measurement of hurricane precipitation remains a critical problem and our knowledge of hurricane precipitation structure is limited. The Tropical Rainfall Measurement Mission (TRMM) satellite and field campaigns have given us the unprecedented opportunity to explore hurricane precipitation in great detail. Specifically, we examine rainfall variability in Hurricane Georges using measurements from the TRMM Microwave Imager (TMI) and Special Sensor Microwave Imager (SSM/I), GOES (infrared), precipitation radar, and gauges.

While it is generally known that Georges produced copious amounts of rainfall, the spatial variation and the reasons for the variation (e.g., terrain interaction, convective bursts) are not well understood. To address those issues, we first compile a storm-total rainfall distribution using the GOES-8 infrared data and a modified version of the Convective/Stratiform Technique (CST). Next, we compute the instantaneous rain rates from the TMI and TRMM Precipitation Radar observations. These values are then combined with the SSM/I derived rain rates to modify the CST estimates when orbits are approximately coincident (±15mins). In this manner we develop statistical relationships among the various sets of observations for the lifecycle of the cyclone. These relationships may then be used to estimate quantitative precipitation for the periods during which only geostationary infrared data is available. The results from our satellite techniques are compared with ground-based observations in the Caribbean and the United States as well as with quantitative precipitation forecasts issued by the National Weather Service. Finally, we investigate the factors that may be influencing the evolution of precipitation during the lifetime of the storm.

The data set produced from this study will be useful for numerical model initialization and data assimilation. This study contributes to a broader effort to apply satellite techniques, combined with radar- and ground based information, in characterizing tropical rainfall and reducing risk from floods.