What New England tornado was on the ground for the longest distance? That’s easy, right? 1953 Worcester.
2011 Hampden County? Wrong again!
It was actually the New Hampshire EF-2 tornado in 2008 that was on the ground for an incredible 52 miles with a nearly continuous damage path. The New Hampshire tornado is an interesting case as it occurred in a low CAPE/high shear environment. While it’s certainly not unusual to see spin ups in this setup getting a 52 mile tornado most certainly is unusual!
The synoptic setup on July 24, 2008 was a classic cut-off upper level low severe setup in New England (not terribly different from the 2012 Connecticut MCV). A strong cut-off low was located at 12z over Lake Huron and western Ontario.
Strong vertical shear is a no-brainer in this scenario (along with low LCLs). The question is normally whether sufficient surface based instability can develop and whether low level winds will back enough to produce enough low level helicity for a tornado threat – hodographs are already long you just need to get them curved!
Here’s the morning 12z GYX sounding showing some instability above the inversion. Not that impressive. The wind fields, however, were quite impressive.
While the boundary layer is stable with a relatively strong inversion breaks of sunshine and the advection of higher dew point air (dew points >22C) eliminated the inversion and CIN by 17z.
This 17z special sounding at GYX shows just shy of 1000 j/kg of CAPE. What is impressive is the 60 knot winds around 600mb that results in strong deep layer shear. Strong low level shear with long hodographs/impressive horizontal vorticity was also present thanks to a 40 knot low level jet around 850mb. Low LCLs (around 500m per SPC mesoanalysis) were supportive for tornadic development.
While storms became surface based there was one storm that produced a powerful mesocyclone and tornado.
The tornado touchdown occurred at 15:30Z in Deerfield, NH producing F2 damage around the time the low level mesocyclone tightened and a gate to gate TVS signature developed. By 15:34Z the tornado vortex signature was quite impressive (90+ knots gate to gate) and a rain wrapped tornado was tearing through the woods of southern New Hampshire.
The tornado was on the ground for a remarkable 80 minutes through a 52-mile swath of New Hampshire. While the low level rotation on radar was present through the storm’s entire path (albeit occasionally unimpressive) the National Weather Service let the warning lapse for 19 minutes while the storm was still on the ground. The initial warning was also issued 16 minutes after the initial touchdown occurred. For the 80 minutes the storm was on the ground it was without a tornado warning for nearly half the time – 35 minutes!
Here’s a look at the 0.5º SRV/Base reflectivity about 3600ft AGL approximately 5 minutes after the initial tornado warning had lapsed and 14 minutes before the next warning was issued. Ironically, the Delta-V associated with the TVS was the second most impressive of the storm’s life with over 80 knots gate-to-gate. On base reflectivity you can also see a hint of a hook echo. Unwarned and rain wrapped is the nightmare scenario for a forecaster.
So what lead to tornadogenesis? There are two good presentations from the 2008 NROW conference that shed some light on tornadogenesis. Numerous storm-scale and topographic interactions (see Cannon, 2008) in addition to an area of enhanced surface based instability and locally backed winds in a small area of clearing (see Cempa, 2008) were some of the storm and mesoscale processes that may have helped move things along.
Here’s a look at a 16z surface plot when the tornado was ongoing east of CON. Notice the higher moisture content in the coastal plain with dew points at KDAW and KSFM (low 70s) in contrast to the upper 60s at KLCI. Also note locally backed winds near the coast. Higher zone of helicity and instability to the right of the storm’s track?
One other thing of note which is interesting about this storm. One would expect a tornado on the apex of a bow echo associated with a QLCS (storm was on the northern/leading end of QLCS) to be relatively short lived. In this case the storm was able to tap into a reserve of untapped unstable air and strong low level horizontal vorticty through its life effectively acting as a discrete storm. Once convection developed east of the line over southern Maine the tornado dissipated. The storm did have echo tops to just over 30,000 feet but the 50dbz echoes just barely scraped 15,000 feet. No hail in this thing!
As I like to say – always beware the cut-off low! They don’t always produce but they always have a few tricks up their sleeve.