The Westchester/Greenwich tornado of July 12, 2006 produced an impressive swath of damage from the Hudson River near the Tappan Zee Bridge through backcountry Greenwich north of the Merritt Parkway.
- Time: 19:30 UTC / July 12, 2006
- Towns: Grand View-On-Hudson to Greenwich
- Strength: F-2 (New York) / F-1 (Connecticut)
- Path Length: 13 miles
- Max Gate-to-Gate Shear (ΔV): 76 knots
- Storm Type: Supercell/Multicell Cluster
The storm exhibted supercellular characteristics with a mesocyclone from approximately 15,000ft AGL through the lowest radar slice. Besides the impressive rotation between the ground and occasionally 6,000ft AGL this supercell isn’t winning any beauty prizes. On reflectivity it resembled more of a mish-mash multi cell cluster. No tight reflectivity gradients, unimpressive reflectivity (rarely over 55 dbz) and a rather shallow mesocyclone made this storm one of the more unimpressive tornadic supercells in Connecticut tornado history during the 88d era.
The weather map at 18z featured a weak shortwave trough in the eastern Great Lakes and upstate New York with weak differential cyclonic vorticity advection over southern New England. The atmosphere, however, was juiced. A light southeasterly flow was present off Long Island Sound with dew points of nearly 22C at White Plains and Bridgeport.
Westerly winds at 925 mb increased to near 35 knots during the day (see 00z OKX sounding below) – and with a light easterly flow near the surface 0-1km shear/helicity was quite high. It’s not surprising tornadogenesis occurred with this storm cluster in the locally backed flow where 0-1km helicity likely exceeded 300 m2/s2. Here’s the 00z OKX sounding and hodograph which shows a moist adiabatic profile from the boundary layer through 250mb. MLCAPE on the order of 884 j/kg was present at 00z.
While there’s plenty of CIN on that sounding at 18z White Plains was 79/72 following sunshine earlier in the afternoon – that should have been enough to remove that CIN. Lifted condensation levels were incredibly low with the tropical airmass in place which is a key ingredient for Connecticut tornadogenesis.
Storm Scale Environment / Radar Analysis
15 minutes before tornadogenesis at 19:15 UTC there was little, if any, low level rotation present. In addition, no mid level mesocyclone was noted early in the storm’s lifetime. By 19:20 UTC a small area of rotation develops south a Nanuet, NY in Rockland County both at the 0.4º and 1.4º elevation slices (4200 and 9600ft AGL, respectively). This is the beginning of a “bottom up” development of the mesocyclone (the non-descending paradigm).
By 19:24 UTC, the volume scane immediately prior to tornadogenesis, spectrum width continues to remain high near the low level velocity couplet (gate-to-gate delta V now 46 knots) indicating a turbulent flow.
By 19:39 UTC the tornado was doing F2 damage near Hawthorne. The low level tornado vortex signature includes a gate-to-gate shear of 72 knots (delta V). The tornado entered Connecticut in Backcountry Greenwich by 19:54 UTC and uprooted thousands of trees on 2 rural streets north of the Merritt.
The tornado vortex signature was most intense at this point – a delta V of 76 knots was detected at 3100ft AGL on the state line. The entire mesocyclone was its strongest at this point with 85 knots of gate-to-gate shear at 8900ft AGL in North Greenwich.
Through the storm’s lifetime the tornado vortex signature was quite small – only spanning 2 or 3 azimuthal gates. The few times when the low level rotation weakened spectrum width jumped indicating that radial velocity bins were unable to properly sample the vortex circulation.