It was a classic setup for low topped supercells on Saturday, September 8 across Connecticut. A strongly sheared environment with long and curved hodographs ahead of a sharp 500 mb shortwave and deepening surface low near the Canadian border. A strong squall line weakened as it approached Connecticut as the best dynamics lifted north the morning and midday low topped supercells were just fascinating to watch! 2 tornadoes touched down in New York City – one in Canarsie and another in Breezy Point (near the Rockaways).
Here’s the 12z OKX sounding about 3 hours prior to the first tornadic supercell in New York City.
The 12z sounding certainly isn’t much to get excited about but that hodograph isn’t half bad! In fact as Saturday morning went on hodographs became longer as the low level jet strengthened. The boundary layer was moistening during the morning with dew points ticking up 1-2C between the 12z balloon launch in Brookhaven and tornadogenesis in the City. Here’s the 1451Z JFK METAR that shows the near-storm environment of the first tornado in Breezy Point that occurred 7 minutes after and only a mile or two from where this observation was taken.
METAR KJFK 081451Z 17020G26KT 10SM BKN012CB BKN017 BKN055 OVC130 26/23 A2975 RMK AO2 PK WND 18026/1451 TSB21E42RAB00E42 SLP075 CB W-NW MOV NE P0000 60000 T02560233 56021
The 18z sounding shows a substantially more unstable sounding owing to the heating and moistening in the boundary layer. Also notice the longer and clockwise curved hodograph. The environment was ripe for tornadoes!
There was nothing pretty about either tornadic supercell on doppler radar. While OKX doppler radar showed little rotation the Newark (and JFK) TDWR did. I’ve been critical of the NWS in some cases before (particularly the Springfield tornado debacle) but they did a phenomenal job getting the warning out here for Queens and Brooklyn. The warning came out with 3 minutes of lead time (which is impressive for this kind of storm) for the first touchdown and 10 minutes for the second.
Here’s OKX radar at 1513Z or 8 minutes after the touchdown.
While the OKX radar site was unable to pick up much rotation (likely was occurring under the radar beam or below 4000 feet) the dual polarization products were able to pick up airborne debris! The CC near or below 80% and differential reflectivity near 0 shows a tornado debris signature in this supercell. This is one of the odd cases where dual pol shows the presence of a tornado given the debris signature but the tornado is occurring below the radar and and therefore there’s no TS or TVS on velocity!
Contrast the low topped supercell on Long Island to the supercell that moved through Litchfield County about 2 hours later. Here’s the same 4 panel doppler radar image
Holy supercell! This is about as impressive you’ll see low level rotation as you’ll see. A strong tornado vortex signature is present over the southern tip of Litchfield near the Morris and Thomaston borders. Delta-V is nearly 90 knots gate to gate! Base reflectivity shows an impressive hook echo and nROT has spiked to 1.56. Notice, however, correlation coefficient values remain quite high indicating that there is no debris signature. Miraculously, given the much less impressive signatures over Queens and Brooklyn 2 hours earlier, this supercell did not produce a tornado. We’re lucky.
The squall line across New York and Pennsylvania began to weaken as it approached Connecticut. Poor mid level lapse rates and the fact the thing was running ahead of best upper level dyanamics shut down the severe weather threat quickly.
This is another example of high shear and even relatively marginal CAPE resulting in a severe weather event across the region during the climatological peak of sea surface temperatures. Always beware the 75 degree water when coupled with a rocket of a low level jet!