September 13, 2001, Isolated Intense Supercell Thunderstorms

A NE-SW oriented cold front with a significant thermal gradient moved across the state through the day... see the 15Z and 00Z surface analysis charts. 850mb temperatures dropped from around 14C ahead of the front... to 2-4C behind it. Several severe thunderstorms formed from late afternoon to mid evening, and became quite intense. Widespread significant convection was not forecast, and did not occur, with one of the largest factors being lack of significant moisture. Surface dew points were generally in the upper 50s to low 60s; see 18Z & 22Z surface dew point contour. Since these maps show isodrosotherms (lines connecting equal dew point) over a wide area, isolated ASOS observations reporting dwpts in the low to mid 60s are not shown in the contour. The AVN and ETA models forecast this well... with the 1000mb panels bringing the 59F isodrosotherm very close to observed conditions... see the 18Z and 00Z 1000mb ETA forecast panel. The NGM did poorly in this respect, keeping low level dew points too low, although this is a bias of the model. Precipitable Water was generally around one inch, give or take from area to area. The convection with this cold front developed, as forecast, along and BEHIND the boundary... unlike the majority of cold fronts which have the convection ahead of them. Models did a good job in showing the forcing behind the front. No moist convection formed before around 19Z (3pm EDT), and models (including commonly used ones such as ETA, NGM, AVN, MESO-ETA) showed a significant increase in forcing, UVV, between the 18Z and 00Z panels. This corresponded well with the time convection would begin to develop, which was mid to late afternoon (>19Z).

The airmass remained unstable for several hours after frontal passage... and then the cooler, drier, more stable air moved in. Widespread sunshine occurred through the morning and afternoon hours before the convection developed... as can be seen in the visible satellite imagery... allowing the airmass to destabilze. Models indicated this well... including the locally run MM5 (from NWS CTP)... see the 15 and 18Z cloud forecast panels. Note the lack of synoptic scale clouds indicated in the lower and mid levels, which turned out to be correct. Surface Based CAPE values over a wide area were low... less than 1000 J/kg across the state, per PSU maps. See the 20Z estimated CAPE. LI's were also nothing impressive, dropping mostly to -4, but enough to support thunderstorms given the presence of other favorable conditions. See 20Z estimated surface LI. KI values were generally in the mid and upper 20s during afternoon and evening before the stable airmass moved in. TT's were in the mid-upper 40s. Each of these values support convection but, like the CAPE and LI, are not impressive. The main feature which made several thunderstorms severe was the vertical wind profile. Rather good speed shear moved into the region during afternoon and evening... with the 00Z KPIT observed sounding showing 55 knot winds reached by the 400mb level. Note the pretty good increase in speed below 400mb, from the lower to mid levels.

Several of the thunderstorms that formed in PA during the afternoon and evening formed rotation. Note in the 00Z PIT sounding, where the cold front has already passed over, the wind shift in the low levels to the north-northwest... while wind remained westerly through the rest of the atmosphere. The forecast ETA sounding for KIPT (Williamsport, PA) at 18Z, just prior to frontal passage, shows a general westerly unidirectional profile. The 00Z forecast sounding for IPT shows the continued WSW profile above 800mb, but a significant veer in low level wind to the NNW... due to the surface front having passing through. Since the convection formed behind the front... this directional shear, plus the added speed shear profile, was sufficient to cause rotating updrafts within thunderstorms.

SPC did not have any of the region in a slight risk... as seen in their day-1 outlook graphic and text. Their probabilistic outlook for damaging convective winds and large hail had the area in the 5% probability, however, which is the lowest percentage given for wind/hail. This event was probably worthy of a slight risk, however, easier seen after the fact. NWS-CTP AFD's mention the possibility of isolated svr wx today, especially the bottom two (2:30pm from the 12th, and 3:35am from the 13th).

The first thunderstorms became severe very rapidly during the late afternoon over north-central PA. The 21:04Z (5:04pm) 0.5 degree tilt (tilt 1) reflectivity image shows two severe thunderstorms; one just west of the Lycoming county border in Clinton county... and the other in northern Lycoming county just south of the Tioga county border. Note 71 dbz with the small cell near the Lycoming/Clinton county border. However the storm in northern/northeastern Lycoming county would soon become the most significant storm. The tilt 1 Storm Relative Velocity (SRV) data by 21:11Z clearly showed rotation within that thunderstorm... and a severe thunderstorm warning had been issued several minutes earlier. Radar indicated hail up to 2.25" in diameter, in addition to a mesocyclone. The 21:16Z Echo Tops image shows 43K feet with the storm. Tilt 1 reflectivity shows 70dbz... and tilt 2 even computes 69dbz at that elevation. While tilt 1 SRV data is not impressive now, tilt 2 SRV looking higher into the thunderstorm cell shows significant rotation... with radar Storm Information Table (SIT) indicating 3-D correlated shear/rotation and hail as large as 2.5"! 21:21Z VIL shows 59 KG/M2... which is quite high. Reflectivity tilt 1 at 21:22Z still has very high dbz... 69. While the tilt 1 SRV data shows rotation, the 21:24Z tilt 2 is certainly more notable, and essential in warning for this storm. Radar continues to indicate a mesocyclone and damaging hail up to 2.5" in diameter, with 44Kft echo tops and 58 VIL. 21:27Z reflectivity data continues to be impressive as well... with tilt 1, tilt 2, tilt 3, and tilt 4 all showing high dbz values. Tilt 3 revealed a core of 70dbz suspended high within the storm... while the three other tilts show a 68dbz. Tilt 1 actually indicates what looks like a hook echo (which implies a rather strong mesocyclone present) in the southwest part of the storm... where mesocyclones typically form. SRV data continues to show rotation.

21:33Z reflectivity tilt 1 and 3 both show a 69dbz, while tilt 2 shows 71dbz. 21:35Z SRV tilt 2 continued to indicate rotation, now becoming very strong actually. NWS-CTP is in the process of issuing a tornado warning for the storm at this time. A strong inbound/outbound couplet can be seen in a zoomed-in image.

21:41Z SRV tilt 2 continues to detect a strong mesocyclone. A zoomed-in image again shows the full extent of it. In addition to the eastern Lycoming county supercell... note the second severe thunderstorm of earlier in western Lycoming county now. It too has rotation within it, however not of the magnitude or organization of the storm east of it. Nevertheless... it is producing large hail and marginally damaging winds.

21:45Z tilt 1 reflectivity data shows a 70dbz echo... and an impressive 73dbz suspended at tilt 2... with the supercell in far eastern Lycoming county at the Sullivan county border. Quickly becoming of more importance is the second severe storm in western Lycoming county... it also possessing a significant elevated dbz core. VIL data from the same time now shows a value of 57 with the western one... higher than the current eastern one. 21:47Z SRV tilt 2 zoomed-in image continues to show strong rotation present with the thunderstorm near the Lycoming/Sullivan county border.

21:50Z tilt 1 reflectivity shows 72dbz... and a VIL as high as 60. The supercell that was in eastern Lycoming county has now moved largely into far southwestern Sullivan county and directly over the Lycoming/Sullivan border. Meanwhile... the severe thunderstorm in western Lycoming county still contains high dbz and VIL, though not nearly as much rotation as the other, indicating the main threat is large hail.

Tilt 2, 21:58Z SRV data shows the rotation continuing with the storm near the Lycoming/Sullivan county line... still capable of tornadic activity. Weak and disorganized rotation also continues with the storm further west in Lycoming county.

22:02Z tilt 1 reflectivity data still shows significant returns. And... for the first time during this event... tilt 1 SRV shows significant rotation with the supercell in far southeast Lycoming and southern Sullivan county. Tilt 2 SRV also shows the rotation as earlier... and also shows rotation with the thunderstorm toward southwestern Lycoming county becoming stronger and more organized. Now also take note of yet a third severe thunderstorm developing... near/just north of KAVP at the Luzerne/Lackawanna county line (eastern most thunderstorm on this radar scope).

The Base Velocity (different than the SRV) from 22:08Z measures tilt 1 wind speeds with the severe storm around the Lycoming/Sullivan/Columbia county line of 71 mph (62 knots). Meanwhile... the tilt 1 SRV continues to show the persistent rotation with that storm. Strong rotation is now also seen with the severe thunderstorm near KAVP at the Luzerne/Lackawanna county line. Both can be seen in more detail with this zoomed-in image. Gusts to at least 51 mph were recorded at the ASOS at KAVP.

22:14Z tilt 1 reflectivity picture shows the three severe thunderstorms; the one now in south-central Lycoming county... the second at the Lycoming/Sullivan/Columbia county border... and the third in Lackawanna county just east of KAVP now. What appears to be a hook echo is noted on the southwestern side of the storm in Lackawanna county... and can be seen in a zoomed-in reflectivity image as well as the full size one. With a zoomed-in tilt 1 SRV image, good rotation in two cells can easily be identified. Tilt 2 SRV also continues to show rotation.

By 22:25Z... the two western storms have weakened significantly... while the severe thunderstorm in Lackawanna county is now very strong... containing a 71dbz core. SRV tilt 1 shows pretty strong rotation within that thunderstorm cell. Large hail to at least golfball size is falling... and smaller hail accumulated to a depth of 2" on the ground, including roads. A man in the Moscow area was struck and killed by lightning when mowing the lawn in a field. Also note rotation with the storm now in far northern Columbia county is weakening... but still present.

Reflectivity from the KBGM RDA site at 22:31Z shows 73dbz at tilt 1 with the thunderstorm in southern Lackawanna county. The two storms which were earlier severe to the west continue to weaken.

22:37Z KBGM reflectivity shows the severe thunderstorm now containing 72dbz entering extreme southern Wayne county from southern Lackawanna county. Both thunderstorms to the west continue to weaken.

22:43Z reflectivity data from KCCX shows the western most of the three storms so far discussed has pulsed back up some... now located at the Lycoming/Northumberland/Montour county border. Also... you may want to take note of the thunderstorm much further to the west in Clearfield county. While not severe at this time... it will become so next hour.

22:57Z reflectivity from KBGM site shows the only intense storm on the scope being on the Wayne/Pike county border... having 67dbz on tilt 1. Hail continues to fall and cover the ground to a depth of a couple inches... causing crop damage... some vehichle damage... and making driving slippery. Note there is not much left of the intense supercell thunderstorm from earlier... now dissipating over west-central Luzerne county.

23:17Z VAD product from the KCCX RDA site shows the vertical speed and directional shear. Typically, you would want the low level winds to be from a southerly direction (SW, S, SE) to provide the vertical directional shear. Today... since the convection formed behind the cold front... low level winds were from a NW-N direction. This is not a typical vertical profile you would look for to expect rotating thunderstorms... but these situations do happen occasionally, with forcing, instability, and shear behind the cold front.

23:23Z reflectivty data shows the thunderstorm mentioned earlier in Clearfield county strengthening near the Clearfield/Centre county line... with a small core of 62dbz noted. A core of dbz values around 60 are also present in another small cell in northern Lycoming county, but this intensity does not last long and high reflectivity is not present through a deep layer.

Reflectivity from 23:29Z now shows the severe thunderstorm over the Clearfield/Centre county line with 65 dbz returns. Unfortunately, SRV data is not available on archive... however it is known that rotation was present with the thunderstorm. A wall cloud was also observed with this severe thunderstorm in southwestern Centre county. See also the 23:35Z reflectivity image. This storm dissipated by 00Z. Through mid evening.. a few other thunderstorms briefly pulsed to strong/marginally severe limits... but after sunset the severe threat had greatly diminished.

A brief LSR from NWS is available... however there are at least several reports that are NOT on the LSR (as there often is). Strangly, even the lightning death victim is not listed there. Other damage from the eastern Lycoming county supercell not listed on the LSR includes several roads blocked by trees and power lines, and at least one house with a large tree into it. More than likely there are even other damage reports. The part of Lycoming county the storm traversed is rural, which also contributed to a lesser number of reports getting to the NWS. Funnel clouds were observed along the path, however it is unknown if a tornado touched the ground anywhere along the path... or the damage done was due to downburst winds. Due to the strength and persistence of the rotation it is probable that a tornado touched down. Yet even with classic supercells in the central United States, sometimes they just don't spawn a tornado, so one can't be sure. NWS never went to survey the area... and I never got up there. A damage swath through forest may very well be found by hunters this Winter. It is fairly common for damage tracks to be found months later in rural/forest areas, and even confirmed by NWS after that long a time period.

A complete listing of all images/text available for this event can be found here.