The Hurricane-customized Extension of the VAD (HEVAD) method: Hurricane wind estimation in the Lower Troposphere

The Hurricane-customized Extension of the VAD (HEVAD) method is a new single-Doppler radar analysis method for hurricanes. It is shown that a second-order Taylor series expansion of the Cartesian wind components and the first 3 coefficients of a Fourier series representation of the VAD data are sufficient to reconstruct the wind field of hurricanes that are predominantly axisymmetric. The HEVAD method assumes that the wind field can be approximated by a modified, Rankine-combined vortex with a superimposed mean-environmental wind. The method retrieves the parameters that specify this vortex so that the Wave 0 (symmetric) components of the tangential and radial winds can be estimated along with the Cartesian components of the mean-Environmental wind averaged over the domain of the VAD circles. The VAD data is taken from PPI scans executed at elevation angles < 11 degrees, and the ratio of the radius of the VAD circle to the range to the circulation center is restricted to < 0.55. Thus, the total wind field estimate is restricted to within the hurricane's Planetary Boundary Layer (~3 km altitude). The method requires estimates of the position of the circulation center and the radius of maximum wind as input. This information is obtained from the first-author's PCA method (another paper also submitted for consideration at this conference).

The HEVAD method was applied to the archive level II data of Hurricane Bret (1999) recorded by the WSR-88D radar facility at Corpus Christi (KCRP), Texas. A single volume scan was selected to coincide with the time of a Triple-Doppler radar synthesis of Bret's wind field provided by the NOAA/AOML/Hurricane Research Division (combined data from the WSR-88D radars at KBRO and KCRP, and a Doppler radar aboard a NOAA research aircraft). The HEVAD method's estimate of the total wind field agreed with that of the Triple Doppler analysis to within 15%. These differences decreased to ~5% when the range to the VAD circles was restricted to < 35 km. These results are very encouraging considering 1) the potential biases that wind asymmetries may inflict on the HEVAD method's estimates, 2) the poor viewing angles between KBRO and KCRP in some regions of the Triple-Doppler wind synthesis, and 3) the different areas used to average the environmental wind,. The information provided by the HEVAD method has potential uses both in the real-time diagnosis of hurricane wind fields and in the data assimilation for hurricane track and intensity forecast models.