Verification of a High-Resolution Model Forecast Using Airborne Doppler Radar Analysis during the Rapid Intensification of Hurricane Guillermo

Abstract

Abstract The NOAA Hurricane Research Division (HRD) P-3 aircraft provided airborne radar observations during the period of rapid intensification of Hurricane Guillermo on 2 August 1997. The inner core structure and evolution of Hurricane Guillermo (1997) over a 120 km by 120 km square area, centered on the storm, was observed by the P-3 aircraft during 10 flight legs at half-hour intervals during a 6-h period from 1800 UTC 2 August to 0000 UTC 3 August 1997. A high-resolution short-term model forecast initialized at 1800 UTC 2 August 1997 was made using the fifth-generation Pennsylvania State University–NCAR nonhydrostatic, two-way interactive, movable, triply nested grid Mesoscale Model (MM5). The weak vortex at the initial time in the NCEP analysis was replaced by a tropical storm–like vortex generated by a 4D variational data assimilation (4D-Var) vortex initialization experiment. The modeled Guillermo followed the observed track with less than a 12-km track error at any time during the 6-h forecast period. The modeled eye is smaller than the observed eye and the modeled vortex is more upright than shown by the radar analysis. The minimum pressure, maximum wind (intensity), and radial profile of tangential winds are close to the radar analysis after 2–3 h of model spinup. A spectral decomposition further reveals that (i) large differences between the model simulation and radar analysis of the asymmetric features are mostly caused by azimuthal phase errors; (ii) the wavenumber 1 component dominates the asymmetric features and remains stationary within the inner core region, as is also observed by airborne Doppler radar; and (iii) although being significantly different from radar analysis, the azimuthal phase of the wavenumber 1 component of modeled reflectivity does not vary greatly with time as the radar data suggest.