Long Track Severe Thunderstorm - Union County Damage, September 4, 2001

A). -- Synoptic Scale & Atmospheric Evaluation

During the morning, afternoon, and into the evening hours a NE-SW oriented cold front moved across Pennsylvania, triggering scattered showers and thunderstorms. An old outflow boundry ahead the front also aided in convective development, and actually appears that it was likely responsible for the first and severe wave of convection. Middle and upper air wind speeds were fairly strong (though nothing impressive), and mostly unidirectional averaging from the west. See the morning 12Z soundings from KPIT, KIAD, and KBUF. The surface cold front just crossed over the BUF area as is evident in the low level winds (per sounding) having veered to the northwest... and the 12Z surface chart. The vertical speed shear was forecast to, and ultimately did, create a favorable environment for isolated thunderstorms to become severe, and sustain themselves, should they grow deep enough into the atmosphere.

One factor that played against a widespread severe weather event was a considerable amount of cloudiness across much of northern and central Pennsylvania during the morning. Areas south of Interstate 80 saw a good deal of sunshine all morning... and the atmosphere became moderately unstable over those locations... with CAPE values up to 1200 J/kg and LI's as low as -5C. Across the whole region ahead of the frontal boundary... KI's were generally at or above 30... and TT's in the 40s, both of which are sufficient to support convection. Atmospheric moisture was also available... with dew points in the upper 60s to low 70s... and precipitable water between 1 and 1.5". The ETA model forecast this well overall, with the 1000mb panel for 18Z showing upper 60s/low 70s dew points surging northward across PA (the solid orange line, known as an isodrosotherm, represents dew points at or above 69F). Mid level energy, positive vorticity, was forecast well by the ETA model at 18Z. Convective activity typically occurs along/just ahead of the forward section of the PV area... in this case the southern section. And that matches well with observed conditions. A big factor that played against a solid line of violent thunderstorms with damaging winds was the lack of significant synoptic scale forcing along the frontal boundary. Also low level speed shear was not significant across PA and areas south and east. In addition, while mid-upper level winds were strong... they certainly could have been much stronger. South of New York State... across PA/NJ during the afternoon... speeds reached no higher than 60 knots in the upper levels and were mostly less than 50 knots prior to frontal passage. Across Upstate Western NY... speeds increased significantly aloft (50 knots just above 600mb, 85 knots at 200mb)... however frontal passage occurred during morning, an unfavorable time for convection with not much great snyoptic dynamics in place. Clouds also were widespread across that area... further inhibiting significant convective development.

SPC had the region in a slight risk for severe weather from the previous day... see text and graphic. The new day 1 outlook issued by 2 am on the 4th continued the threat (text / graphic). NWS CTP and PHI issued these Severe Weather Outlooks to alert the public of the threat.

B). -- Regional Thunderstorm & Radar Overview

A band of scattered showers and thunderstorms increased in intensity ahead of the cold front and with the outflow boundary during the course of morning.... and especially afternoon... across PA. Due to the widespread cloudiness across northern PA and early frontal/outflow boundary passage time... thunderstorms did not become strong until they reached the I-80 corridor and southward. Here, isolated intense thunderstorms formed during the afternoon. SPC issued a severe thunderstorm watch during mid afternoon for about the southeastern 1/4 of PA... from the Lower Susquehanna Valley through the Lehigh Valley and Philadelphia Metro area... northern Delaware... and much of New Jersey.

NOTE: UNFORTUNATELY, KCCX RADAR WHICH NORMALLY COVERS THE AREA DISCUSSED BELOW WAS NOT AVAILABLE TO ANYONE OUTSIDE THE NWS OFFICE DUE TO COMMUNICATION PROBLEMS. THUS, RADAR FROM KBGM WILL BE PROVIDED.

What would become the most significant thunderstorm of the day in PA formed initially in southwest Lycoming county during early afternoon. However, its origin appears to have started well northwest of there, in NW PA as a towering cumulus and convective shower. Not until near the Clinton/Lycoming county border did it attain thunderstorm status, though. As this thunderstorm moved southeast into Union county... it produced large hail and a wind damage event. At 17:42Z (1:42pm)... you can see the core of this thunderstorm over the Lycoming/Union county borders. The 17:53Z (1:53pm) reflectivity image shows the thunderstorm moving across eastern Union county, the severe part of it having just entered Kelly Township north of Lewisburg. It is just about to produce wind damage in Kelly Twp around 2 miles north of Lewisburg. The reflectivity image from 17:58Z (1:58pm) continues to show the severe thunderstorm... now about to enter Upper Northumberland county south if Milton. Nickel to quarter size hail (.88 - 1") was falling under the core of it. Wind damage began around 1:55pm... and damage is occurring at the time this radar image was taken (1:58pm). 1803Z reflectivity data shows the severe thunderstorm over the Union/Northumberland county border. The most significant winds are dissipating around now. See the directory for this event for numerous radar images after time. Storm Relative Velocity (SRV) data indicated very weak rotation at a very shallow level, only evident on the lowest level radar scan. This was nowhere near the criteria for a tornado warning, and thus a severe thunderstorm warning was issued for the area. The thunderstorm continued to track southeast for a rather long period of time... until southern New Jersey. A few spots of enhanced damage were noted along the path... such as in eastern Union county (as noted above) and parts of Berks county. KRDG (Reading Regional airport) recorded thunderstorm wind gusts to 59 mph as the storm moved by. Numerous trees were downed across the path of the thunderstorm in parts of western/central Berks county... where peak convective gusts likely exceeded 60mph by a little bit. Otherwise just sporadic damage ocurred southeast across SE PA... more minor compared to areas further northwest. Large hail continued to fall under the precipitation core. The rest of this paper will discuss the damage in Union county... the most significant from this storm (and of the day for that matter). While NWS classifies this a true tornado only, with a funnel cloud extending from the thunderstorm (as opposed to a downburst/gustnado combination), I am aware of several things in their survey that are incorrect for a fact... including the damage width. We will probably never know for sure, and the difficulties in determining it (with pros and cons for both aspects) will be discussed at the bottom of this paper.

C). -- Union/Upper Northumberland County Path Track Information

NOTE: My damage survey, which is provided below, will differ some from the NWS survey. It it clearly seen upon investigation of the damage area that the width of the damage and "touchdown" location are different than what NWS states (both in their PNS and especially to myself over the phone). I advised the NWS surveyer of this and part of their information may be amended for official inclusion in the Storm Data Publication. Perhaps the largest disagreement between us is classification... tornado verses downburst and gustnado. I will not refer to the event as a true tornado in this paper, due to the building evidence against it.

The first damage was noted north of Colonel John Kelly Road along Fairchild Road in Kelly Township... northwest of the Colonel John Kelly Road/Hospital Drive intersection... and this is being used as location for the touchdown of the downburst. The downburst tracked generally in an east-southeast direction... crossing Fairchild Rd... Colonel John Kelly Rd... Hospital Drive... JPM Rd... Saint Mary Street... 1st Street... Zeigler Street... brushing by Hafer Road and then crossing U.S. Route 15. It crossed the West Branch of the Susquehanna River and entered upper Northumberland county in West Chillisquaque Township. The downburst crossed State Route (SR) 405 (which runs parallel to and just a few feet east of the river) and appeared to dissipate shortly east SR 405 (between SR 405 & 147). This very end section of the damage path was not surveyed as extently, thus information as detailed as the rest of the path cannot be provided. The largest width of the damage path was right around 3/4 of a mile... or about 4000 feet... widest toward the middle and end of the path and narrower at the beginning. The length of the damage path was about 2.5 to 3 miles. Intensity rating will be given an F1 (low end of F1) at its maximum time... with peak winds probably somewhere in the 80 mph area (this is one thing myself and NWS agree on). As is typical though, the entire path did not experience these speeds, and intensity across the damage track ranged from F0 to low F1... most commonly in the upper F0 range.

D). -- Damage Information

The more and more I looked over the damage from a map point of view... the more it became likely that this damage was the result of a downburst (a microburst) and gustnado. The width of the damage path widened with its length... a classic sign of a microburst. A narrower track of enhanced damage was seen within the larger wind damage path. The only rotation or convergance noted along the entire damage path was in that narrow enhanced area of damage. And further, as stated above, radar did not indicate any notable rotation with the parent thunderstorm needed to produce a "true tornado".

The first damage noted was along Fairchild Road... north of Colonel John Kelly Road. Here at least one tree was uprooted... a portion of a sheet metal barn roof was peeled back... and the roofing of a porch was torn away. Isolated damage continued to the corner of Fairchild/Col. John Kelly Roads... where several tree limbs were snapped out of trees and at least two houses had gutters partially ripped off. Very little corn damage was noted so far. In the corn fields just south of John Kelly Road and just west of Hospital Drive is where significant corn damage began to be noted... along with several trees snapped. This damage continued through the fields crossing Hospital Drive south of John Kelly Road. A portion of another barn roof was peeled back along Hospital Drive and a large tree snapped on the property. Further down the damage path to JPM Road, significant corn damage continued to be noted. Clear rotational patterns was seen in this area. Along JPM Road around a power sub station... several large Colorado Spruce trees were snapped and uprooted... with a portion of a metal fence blown over. Just south of this sub-station a very large tree was snapped. The most significant damage occurred along Saint Mary Street, 1st Street, and Zeigler Street along/just east of JPM Road. Numerous trees were snapped or uprooted... from Oaks to Pines and Spruces. A large... 2 and a 1/2 foot diameter... tree was uprooted immediately next to a house... luckily falling in the opposite direction. A few houses away a falling tree knocked out the window of a house. A dog house was overturned. All the roads in the area were blocked initially by the debris. Some yards looked like a jungle of downed trees... just anywhere you look you saw piles of trees laying down. Besides the broken window on the house... damage to structures was very limited to the houses here. Gutter damage was noted on a few.

Just a few hundred yards from this area along Zeigler Street sat the Kelly Township Municipal grounds and building. To the rear of the building was a baseball field. A wooden concession stand was oblitterated with its pieces strewn across the field. Trees were snapped. The Township building itself sustained heavy damage... beyond repair. The entire roof was completely torn off... with pieces of it landing anywhere from a few hundred yards (in a field) to a half mile away (along Route 15). In addition the entire western side of the building was pushed inward and collapsed on a dump truck. An employee had left the buiding just a couple minutes before. Large plastic drainage tubes on the Twp grounds were thrown a few hundred yards into the field along with the roof parts.... coming to rest among a grove of trees (some of which were snapped). A 1/4 mile to the south of the Twp Building was the very popular Route 15 Flea Market. Very few trees are on the grounds... but most of the few small ones that were there were snapped. The roof of an overhang next to a shed was torn away... and a trailer was overturned. Many 100 pound wooden benches outside for the vendors were thrown around... some piled three high... others pushed into piles and overturned. At least one was carried a considerabel distance away from all the others... coming to rest upside down in the middle of the parking lot. Deadly missles such as pitch forks and rakes were found laying around the area. A fairly large roof attached to the main building over large door area was ripped off. An air conditioning unit bolted to the roof was ripped off the bolts and landed in a low laying section of the roof. On one building... the metal section on the gable end of the roof was pulled away from the structure by several feet. A carwash on the grounds had the ceiling materials from the inside part of it pulled out and strewn across the area. An antique store also had similar damage... with pieces of the soffit pulled out. Pieces of the wooden trim/fascia were also pulled away from the building. Along Route 15... from in the front of the Flea Market Grounds south to Hafer Lane... 13 large (PP&L) power poles were snapped. At least a couple signs were destroyed.

At the intersection of Zeigler Street and Route 15 (few hundred yards from Flea Market)... a camper dealership had a small portion portion of their sheet metal roofing ripped off... which ended up hanging from power lines. On the other side of the street... a smaller barn next to a large barn was destroyed... with a several foot high concrete block wall knocked completely over. Sheet metal roofing on the main part of the barn was peeled back. 55 gallon barrels were picked up and thrown down an embankment along Route 15. An auto parts store (now out of business) had pieces of the soffit pulled out. A wooden shed had one wall pulled apart from another wall. At the Route 15/Hafer Road intersection... several trees were down, including some large ones. Down Hafer Road... which was along the very edge of the damage path... a few houses sustained minor damage, including one with a few shingles lifted and an antenna bent and another with the gutter and fascia ripped away.

Over to the east side of Route 15 toward the Susquehanna River... more trees snapped. A barn had a part of the sheet metal roofing peeled back. A road construction sign was found laying in a field; it is unknown exactly where this was blown from. Across the river into upper Northumberland county... numerous trees were downed across SR 405... blocking it. A small section of sheet metal roof on a factory was peeled back along SR 405. A main path of damage dissipated east of here.

This event had a high potential of causing injuries and deaths due to the amount of debris with it. The employee that left the Township building just prior to its destruction is an example of pure luck. If the flea market would have been in operation that day... many thousands of people would have been inside and out. Heavy wooden benches, roofing, and other flying objects (not to mention the overturned trailer) could have easily caused a tragic event. As for residents... a number of them interviewed actually sought shelter in the basement... and all had property damage. One mother took her child who was outside playing inside minutes before the storm struck. When they emerged after the storm... downed trees covered their property.

E). -- Conclusions

What actually caused this damage may likely never be known for sure. It could have been a true tornado... or a microburst and associated gustnado(s). The more the case is studied... the latter appears more probable. As stated at the top of section D, the damage path widened with length. This is not uncommon for microbursts, it can be a classic signature actually. Velocity data off the radar did not indicate any notable rotation indicative of a mesocyclone within the thunderstorm which would be responsible for a tornado. Very weak rotation was seen in the lowest scan elevation... and if a weak mesocyclone was present it was very shallow and mostly below the radar beam. The damage appears to have occurred under the precipitation core of the thunderstorm... a common place for microbursts within individual thunderstorm cells (as this one was). Assuming this was a microburst... it appears that at least one gustnado occurred with it... probably along the leading edge of the microburst. It is possible that two gustnados spun up, which would not be that uncommon. The main gustnado path would have resulted in the enhanced damage track within a larger area of lesser damage (from weak downburst). A number of people reported seeing rotations... but none could link a funnel cloud to the thunderstorm base. From all evidence gathered... from radar data to storm surveys... there is nothing that disagrees with a microburst event.

So far we have discussed primarily aspects against a tornado. Now lets assume this was an actual tornado, from weak low level thunderstorm rotation. Weak tornadoes can sometimes have its strongest/most damaging winds on the forward motion side of the tornado. Thus... since only one side of the tornado is causing damage and the other side has lesser winds not capable of causing much damage... usually almost all damage would seem to be divergent (from a downburst). The exception can be over corn fields where even the weaker winds can be capable of blowing down corn stalks... thus showing rotation there only. In this case that was the only area where rotation was noted - over the corn field area. Once those winds move past the corn fields they are usually not capable of causing notable damage... and thus rotation is no longer indicated. With this particular event... that is the other possibility. Due to the width of the damage path... it is highly unlikely that the tornado was that large. It is possible that the tornado itself was smaller, the size of the narrower enhanced damage track, and the surrounding more minor damage was due to inflow around the tornado.

Both of these solutions seem reasonable, and each can be argued nearly equally. What may tilt the scale toward downburst/gustnado classification for me is the fanning out of the damage path, and lack of tornadic signatures on radar. Any further information obtained in the future will be posted on this page. Questions or comments are warmly welcomed, and can be sent to BowEcho@ptd.net.

Go here for a listing of all images/text available for this event.