Mesoscale wind-fields with high spatial and temporal resolution, continuity, and accuracy have been determined for a mature hurricane. An automatic algorithm utilizing a massively parallel computer architecture has been developed for computing cloud-top winds from GOES visible rapid-scan imagery, and a post-processing procedure using spectral digital filter methods has been applied for purposes of noise-removal. Previously satellite-derived winds near tropical cyclones have been determined only at some distance from the eye over a typical time period of 1-2 hours. With the advent of high-resolution and high-dynamic range GOES imagery, new capabilities exist to measure the evolution of hurricane outflow processes, inner-core dynamics, and the mesoscale interactions between them.
The source of data for this study is a unique nearly continuous twelve-hour sequence of one minute images from the GOES-9 satellite on 06 September 1995. The images are centered on Hurricane Luis when it was approximately 250 km NE of San Juan, PR at category-4 strength and undergoing a period of vigorous eyewall convective bursts. Wind fields are computed for each 512x512 image in the sequence. From these wind fields the evolution of the outflow region of Hurricane Luis during this period may be observed. The mesovortex embedded within the inner-core of the storm may be inferred from the rapid rotation of the low-level center of circulation and its interaction with the eyewall. In addition the upper-level outflow region is decomposed into axisymmetic and perturbation flow components, and deformation fields. Conclusions are drawn from observations of the evolution of these fields about their relationships to storm motion, environmental shear, and intensification processes. Results are presented via animated computer visualizations