Boundary Layer Rolls Derived From SAR Wind Fields in Typhoon Megi

The marine planetary boundary layer (BL) is the lowest layer of the atmosphere and is generally turbulent. In tropical cyclones (TC) the BL is one of the most extreme environments, determines the supply of thermodynamic energy to the storm and is unstably stratified. As a result organized eddies in form of roll vortices form. These rolls are spatially periodic, span the depth of the BL, and are in line with the mean surface wind direction. Below downdrafts there is an enhancement in the surface wind and there is a reduction below updrafts. It is also below these updrafts where less friction occurs. The spacing of the roll axes typically spans 2.4 to 4 times the BL depth. The alignment of the rows also creates areas which are buoyantly more efficient due to the protection of the ambient wind shear. Due to extreme conditions, the BL structure in TC has mostly been analyzed with the help of models (physical parameterization). Nevertheless, a great amount of in situ and remote sensing data is available. In this work, horizontal wind fields are derived from 12 SAR (Synthetic Aperture Radar) images of Super-Typhoon Megi (Pacific: October 15th to 21st, 2010) by Cosmo SkyMed (1, 2, and 3), ENVISAT, RADARSAT-2, and TerraSAR-X and compared to COAMPS (Coupled Ocean/Atmosphere Mesoscale Prediction System) wind fields. After a sea-level-pressure correction of the horizontal wind fields, these are input to a TC BL model to simulate two dimensional vertical wind fields (azimuthal and radial). In this work we are presenting example results from two collocated SAR images (RADARSAT-2 and Cosmo SkyMed-3) showing the vertical distribution of the azimuthal wind and the overturning circulation. The azimuthal wind velocity indicates the position, size, and strength of the vortices (Figure 1). In between rolls we find a reduction in the surface wind below updrafts and an enhancement below downdrafts. In Figure 1 we find such a velocity reduction at about 40 and 60 km distance to the storm's eye. In between these two points the velocity is increasing which indicates a downdraft. As a result we suspect two interacting rolls between 40 and 60 km distance to the storm's center. Further examples and analysis of this work would be presented at the conference.