Deadly thunderstorms of 13 August
By
Richard H. Grumm
And
Ermie Ostuno
National Weather Service State College PA 16803
Introduction:
This page shows some of the distinct radar signatures associated with the severe weather event of 13 August 1999. This was a late afternoon event, which primarily affected the central most counties of central Pennsylvania. The event was dominated by a series of bow echoes. At least three of these bow echoes may have fit the category of single-cell type bow echoes (CBE) as described by Lee et al. (MWR Oct 1992). The first squall line like bow echo (SLBE) of the day downed trees at cross Clearfield, Centre, southern Clinton, southern Lycoming, Snyder, and Northumberland Counties. A lone-long lived pulse type storm downed many trees in northern Cambria County and turned into a bow echo as it moved into southwest center County. This CBE retained its character as it crossed seven mountains and took and a clear bow echo shape in the next southwest to northeast oriented valley in Snyder County. A second SLBE followed about 20 minutes behind the track of the pulse-storm CBE. Later in the event, two thunderstorm clusters entered western Tuscarora Valley; each took on a mini-bow shape as the moved east-northeast down the valley. The two bows merged forming what was likely the third SLBE of the night.
The archive II data in this event clearly showed that distinct echoes dominated the event. Additionally, distinct features appeared to form and sustain themselves once they entered a southwest-northeast oriented valley. Each of these valleys becomes more diffluent as one moved eastward down these valleys. It appears the unidirectional westerly shear and the valley orientation played a critical role in character and path of the distinct echoes of 13 August 1999.
To see a quick summary of the severe weather please read the local storm report (LSR). Not included in the LSR was the report of heavy rains, street flooding, and a few downed trees in northern Blair County. The times in the LSR do not necessarily reflect the times of actual damage due to reporting errors and other problems related to severe weather verification.
In addition to producing severe weather in the form of damaging winds, these storms produced locally heavy rainfalls. The rainfall was probably one of the most beneficial aspects of the event since the area was under severe drought conditions.
Method:
Archive II data were used to examine these storms from our RIDDS tapes and displayed using WATADS. Storm reports came from locally collected data. A list of all available radar images is available here.
RESULTS:
The squall-line like bow echo (SLBE) first appeared to the west of the State College County warning area (CWA) as seen in the 1954 UTC base reflectivity (BREF) and velocity images. The National Weather Service in Pittsburgh (KPIT) issued warnings on this line of storms as they moved over Indiana County. The strong inbound winds in the velocity image suggest this was a strong storm. KPIT called the State College Office (KCTP) when reports of severe weather were received. This SLBE moved generally eastward across the Allegheny plateau until reaching Happy Valley. Once in Happy Valley, the SLBE moved east northeastward down the valley over the next 3 hours.
By 2012 UTC, the composite reflectivity image showed that the northern portions of the SLBE had entered Clearfield County. The large pulse storm was still to the southwest over Indiana County was storm 16 in the BREF image. A cross section through this storm in both the velocity and reflectivity showed the depth of this storm. The outbound and inbound velocities captured a portion of the updraft and downdraft in the storm respectively.
By 2232 the SLBE was moving across Clearfield County and the pulse storm was about to enter Cambria County just southwest of Barnsboro. The supercell was still a vertically tall storm. The velocity cross section showed the updraft/downdraft couplets along the lone pulsing storm.
At 2137 the bow echo was over Centre and extreme southern Clinton Counties. A distinct circulation was present near Lockhaven in the SRM product. A more distinct bow shape evolved through 2157 UTC as the storm moved along the valleys of the region. The apex of the bow was near Jersey Shore in southwestern Lycoming County and extreme western Union County at this time. The WSR-88D showed a distinct circulation near Jersey shore in the SRM product at this time. This feature moved eastward toward Garden View by 2212 UTC. The zoomed in composite reflectivity data showed the tail of the bow near Mifflenburg. Ten minutes later, this bow was moving across eastern Union County toward Northumberland County as seen in the 2222 UTC composite reflectivity and velocity image. This bow probably produced all the damage reported that evening in northern Northumberland County.By 2137 area of slow moving heavy rain was moving in from the west behind this system. A second distinct bow echo would develop along this line producing damage in Union County. The distinct bow echo can be seen near Selinsgrove around 0008 UTC in the reflectivity and velocity data.
This storm had a long history more characteristic of a supercell storm. It could be traced back to before 1900 UTC. However, we pick the storm up around 1948 UTC, labeled storm #16 by the WSR-88D SCIT as shown in the reflectivity image at the southern flank of a line of thunderstorms along which the first SLBE of the day would develop. The storm slowly moved eastward over the nest 12 minutes and pulsed. By 2012 UTC the storm was pulsing upward a second time pulsing downward a second time by 2022 UTC. At 2032 UTC the storm was approaching the Cambria County line, just west of Barnsboro. The storm began to pulse up again around 2042 UTC. The well-defined upright updraft could be seen in the reflectivity cross section at this time near Barnsboro. Five minutes later, an elevated core was found above 21 kft in the storm, which rapidly descended in the next volume scan (2052 UTC). The storm continued to lower into the next volume scan. By 2112 UTC the storm was pretty much spent.
Velocity cross sections through the storm captured the downdraft/updraft couplet and the divergent downburst near Barnsboro and Prince Galitzen State Park. As shown the images from 2032, 2037, 2042, 2047, 2052, and 2057 UTC. By 2102 UTC, the storm was dominated by what appeared to be an outbound downdraft. The downburst divergence signature can be seen near 43 miles along the cross section at 2042 UTC and near 40 miles at 2047 UTC. In both of these images, strong outbound updraft can be seen on the RDA side of the downburst. In the 2037 UTC image, it is interesting to note the strong low to mid level convergence in the storm and the strong storm top divergence. These signatures may play a role in forecasting storm strength.
This storm took on a bow shape, which was more distinct in the base velocity and SRM data after 2057 UTC as it moved across Centre and Juniata Counties. In the reflectivity data, it had a weak bow shape but appeared to be a very multicellular feature. this storm rapidly took on a bow echo shape in both the reflectivity and velocity data.
The CBE began to take on a distinct shape upon entering the valley to the east appeared around 2300 UTC in Snyder County. It was labeled storm 17 by the WSR-88D SCIT as shown in the 2313 UTC reflectivity image. The corresponding velocity data showed some convergence just southwest of Selenisgrove at this time. By 2328 UTC, the storm (#17) had moved over Northumberland County with low-level convergence still present in the storm. The velocity cross section clearly showed the relative flow with this storm at about 40 miles out in the cross section. The larger SLBE can be seen farther west around 10 miles along the cross section. A strong rear inflow jet was already present with this feature near 7000 ft at about mile 5 along the cross section.
By 2343 UTC, the reflectivity data showed a more bow like shape to storm 17. Both the storm relative velocity and reflectivity cross sections through the storm showed a deep updraft core. A distinct mini-bow echo was clearly visible in the base velocity data at this time. Note the large SLBE to the west in the velocity image.This CBE move over Columbia County shortly after 0000 UTC and was labeled as storm 18 by the SCIT in the 0008 UTC reflectivity image. The distinct bow-echo shape was better defined in the 0008 UTC storm relative velocity image. For size comparison, the larger bow echo to the west is clearly visible in this image with the RFJ located near storm ID #15. The storm relative velocity cross section captured both features. The green inbounds give a good sense of the updraft in both bow echoes. However, the leading CBE has a more upright updraft than the steadier state SLBE.
In all the strong storms on this day, strong convergence at low and mid levels and strong storm top divergence appeared to be key signals in forecasting downburst and bow strengthening.
The isolated pulse storm, in addition to having a distinct set of the plan view velocity and reflectivity signatures; and vertical reflectivity of strong pulsing thunderstorm it also showed up as the high VIL storm of the day. The VIL product from 2047 UTC showed an area of 62 VIL over northwestern Cambria County. On the next volume scan (not shown) the VIL's dropped to 55 and steadily decreased with time. The storm did most of its damage shortly after 2047 UTC as it moved across Cambria County. The reflectivity cross sections at 2032 UTC and 2042 UTC sections showed the storm growing. Note the 65-dBZ core near 28 kft in the latter image and the weak echo region in the lighter reds from near 12 to 18 kft in the storm. The 2047 UTC cross section continued to show the elevated 65-dBZ core and a solid 60 dBZ core up to 35 kft in the storm, contributing to the high VIL value. In the next image, at 2057 dBZ note the near complete collapse of this storm. This lead to the downburst activity over Prince Gallitzen State Park, the falling trees and the one fatality. Note the strong downdraft near 7 kft in the velocity image at 2047 UTC about 14 miles from the radar.
The two weakly defined bow echoes, which formed the Tuscarora SLBE, can be seen in the 0047 UTC reflectivity and base velocity data. The leading bow is already in the valley, located northwest of Carisle. The WSR-88D found 3 circulation centers with this little feature and two with the bow to the west. Neither feature was very impressive at this time.
By 0127 UTC, both features can be seen moving down the valley, following a more east-northeast path due to the curvature of the valley. Again, the reflectivity data did not do the systems justice. However, the SRM product showed the two distinct features. Note the second bow echo was larger and was rapidly overtaking the first bow.
By 0143 UTC the reflectivity data shows one SLBE entering Dauphin County. The bow looked was more readily definable in the SRM product at this time. The distinct bow in the reflectivity data to the northeast was more impressive in appearance then the Tuscarora Valley SLBE. The bow continued to strengthen and grow in size once it exited the confined western portions of the valley as seen in the 0158 UTC SRM and reflectivity data. The WSR-88D indicated two mesocyclones along this feature.
By 0213 UTC, the bow began to develop a cyclonic bookend vortex along its northern flanks which could be seen in both the reflectivity and SRM data. The cyclonic bookend vortex was located north of the city of Lebanon along the Lebanon and Schuylkill County borders. The velocity data (not shown) displayed 37 kts of wind moving into the apex of the bow over northern Lebanon County with strong convergence along the gust front.
By 0223 UTC, the SRM data showed a distinct mesocyclone associated with the cyclonic bookend vortex over Schuylkill County and a well defined rear inflow jet (RIJ) over eastern Lebanon County. The reflectivity data finally showed the character of this distinct bow echo. Note the weak echo channel in Lebanon County along and just north of the RIJ. This and the cyclonic bookend vortex were two features with a good potential to produce severe weather. The base velocity data at this time showed up to 47 kts of wind near the RIJ.
By 2043 UTC the bow echo was over Schuylkill and Berks Counties. The cyclonic bookend vortex could be seen near Pottsville in both the reflectivity and SRM data. The strong RIJ could be seen to the south of this feature. With an astute eye, one can see the weak each channel in northern Berks County.
f. the needed rainfall
These storms produced some much need rainfall over the drought stricken region of the State. The heaviest rains fell a few hours either side of 2300 UTC as shown in the 0022 UTC three hour precipitation product (THP) and the 14/0153 UTC storm total precipitation product (STP). In the THP product, observe the impact on the rainfall of the supercell across northern Cambria County and the impact of the bow echo apex, which moved down the valley from State College into Union County. The two distinct echoes, which produced severe weather both produced heavy rainfall. The overall impact of these two features can be seen in the STP, however, the heaviest estimated rain fell over portions of Huntingdon, Mifflin, and Juniata Counties.
Several reports suggest that these rainfall estimates were reasonable but in no way exact. We should obtain some stage III integrated precipitation products to clearly show the rainfall associated with this event.
CONCLUSIONS:
