Abstract
This paper describes the boundary layer wind structure and dynamics of Hurricanes Danielle (1998) and Isabel (2003), based on the analysis of high-resolution global positioning system dropwindsonde data and simulation of the flow by a three-dimensional boundary layer model produced by Kepert and Wang. The observations show that the hurricane boundary layer has a complex three-dimensional structure with large variability over small distances. The analysis emphasizes three aspects: the degree of gradient-wind balance, the radially varying depth of the boundary layer, and the strength of the near-surface wind speed relative to that at a higher level. Each aspect is compared both with results obtained in a simulation of the individual storm by Kepert and Wang’s model and with theoretical predictions. The observations show that the boundary layer depth decreases toward the center of the storm, consistent with theoretical arguments. The strongest azimuthal winds occur near the top of, but still within, the frictional inflow layer. These strong azimuthal winds are marginally supergradient in Hurricane Danielle but strongly so in Hurricane Isabel, where the imbalance amounts to approximately 10 m s−1 near the radius of maximum winds and is statistically significantly nonzero. This layer of supergradient flow is surmounted by a layer of outflow, in which the flow returns to gradient balance. The maximum storm-relative azimuthal wind occurs in the left front of Hurricane Danielle, and the strongest inflow is located in the right front. These asymmetries rotate anticyclonically with height, but there is also a clear wavenumber-2 asymmetry superimposed, which shows less rotation with height and is possibly forced by environmental factors associated with the storm’s impending recurvature. In Hurricane Isabel, the azimuthal wind maximum is located in the left rear and the inflow maximum in the left front, with neither showing much tendency to vary in azimuth with height. The ratio of the near-surface wind speed to that farther aloft increases toward the storm center for both storms. The largest values are located near the radius of maximum wind, and in general higher values are found on the left of the storm’s track than on the right. Simulations of the two storms with the boundary layer model are able to explain several of these factors; they also show some ability to reproduce individual dropsonde wind observed profiles. Important is that the model predicts weakly supergradient flow in Danielle and strongly supergradient flow in Isabel, in excellent agreement with the observational analysis. Based on these simulations, physical arguments, and earlier studies, the authors conclude that the differences between these storms in this respect result from their differing radial profiles of gradient wind and argue that the occurrence of supergradient flow in the upper boundary layer of individual hurricanes should be readily predictable.
Assimilation of Himawari-8 Imager Radiance Data with the WRF-3DVAR system for the prediction of Typhoon Soulder
