Abstract
Aircraft, rawinsonde, satellite, ship, and buoy data collected over a 40-h period were composited to analyze the inflow-layer structure of Hurricane Frederic (1979) within a radius of 10° latitude of the storm center. To improve the quality of the composite analyses, the low-level cloud-motion winds (CMWs) employed in this study were assigned a level of best fit (LBF). An LBF was assigned to each CMW by determining the level at which the closest agreement existed between CMW and ground-truth wind data (e.g., rawinsonde, aircraft, ship, and buoy). The CMWs were then adjusted vertically to uniform analysis levels, combined with ground-truth wind data, and objectively analyzed. These objectively analyzed wind fields were used to obtain kinematically derived fields of vorticity, divergence, and vertical velocity. An angular-momentum budget was also computed to obtain estimates of surface drag coefficients.
The low-level CMWs in this study were found to have LBFs ranging from 300 to 4000 m. It was shown that judicious use of this knowledge leads to substantial improvements in the estimates of the radial flow, but relatively insignificant improvement in the estimates of the rotational component of the wind. These results suggest that the common practice of assigning all low-level CMWs in a tropical cyclone environment to a constant level of 900–950 mb (approximately 500–1000 m) is probably appropriate for computations that depend primarily upon the rotational wind component. These findings, however, also indicate that failure to account for variations in LBFs of low-level CMWs could result in substantial errors in calculations that are sensitive to the radial wind.
The kinematic analyses showed that the asymmetric wind structure observed previously in studies of Frederic's inner core extends out to at least 10° latitude radius. Frederic was characterized by strong northeast-southwest radial flow through the storm and a pronounced northwest-southeast asymmetry of the tangential wind field at each analysis level. Analysis of Frederic's surface-560-m angular-momentum budget showed that the mean value of the surface drag coefficient beyond 2° radius was approximately 1.8 × 10−3.