Isolated severe weather and cold air damming 9 November 2005 By Richard H. Grumm National Weather Service Office State College, PA 16801 1. INTRODUCTION Two lines of convection moved over the State of Pennsylvania and adjacent portions of New York and Ohio on 9 November 2005. Most of central and eastern Pennsylvania was in a cold air damming situation through both events. The first wave of convection, during, the late morning and early afternoon hours occurred with the warm front and warm advection. No severe weather was observed in Pennsylvania with these storms. Large hail and some wind was observed in western New York (Figure 1). The second wave of convection fired off late in the afternoon. The initial line of convection, ahead of the cold front, developed north of Lake Erie in Ontario and crossed the lake forming a line of thunderstorms in eastern Ohio. This line of storms eventually produced severe weather across portions of Ohio, Pennsylvania and New York (Figure 1). The easternmost report of damage was in Luzerne County, wind damage, during the afternoon elevated convection. Critical to forecasting this event was the low-level cold air and the shear profile. Despite the late season and late arrival of the cold frontal convection, severe weather was observed. The quick ability of the storms to produce rotation and the strong winds aloft may both have been contributing factors to the severe weather in northwest Pennsylvania. This paper will examine the event of 9 November 2005. 2. METHODS Most data shown here was obtained from AWIPS. The 7-day archive was used to replay the event and produce images. The data is available for use on the Weather Event Simulator too. Most of the data shown here is limited to satellite and radar data. Some short-term model forecasts and analysis fields are used to show the regions of cold air damming. 3. RESULTS a) Warm advection phase b) Cold frontal phase Figure 1 Storm Prediction Centers (SPC) reports of severe weather on 9 November 2005. Courtesy of the SPC.
western Ohio and the surge of warm air had moved into western New York, though the cold air had made little progress northeastward over Pennsylvania. Though not shown lifted indices and low-level θe showed the surface front was farther west than the 850 hpa frontal system. a) 1300 UTC The thunderstorm activity associated with the warm advection formed a line that is visible in the WV imagery. The thunderstorms over western New York did produce severe damage. The storms in Pennsylvania, with the exception of one event in northeastern Pennsylvania did not produce reported severe weather. This line of storms can be seen in the 1955 UTC water vapor image. At this time, a line of storms extended from New York southward into Pennsylvania. The convection near the cold and occluded fronts could be seen over Ontario at this time. It is interesting to note that no convection developed over Ohio, an area that was originally a forecast concern. a) 1815 UTC Figure 2 Water vapor imagery at a) 1300 and b) 1815 UTC with nearest hour 850 hpa winds and equivalent potential temperatures. Lines show approximate frontal systems in the regions explained in the text. Figure 2 shows the water vapor imagery at 09/1300 UTC. The red line approximates the 850 hpa warm frontal boundary estimated from the 850 hpa equivalent potential temperatures (θe). The initial thunderstorms were over Ontario and Lake Erie in the region of warm advection. By 1815 UTC, the 850 hpa cold front could be seen crossing The strong drying in the area of and to the west of the warm advection was a potential signal for destabilization in the affected area. This dry tongue (in loops which cannot be shown) moved from northwest to southeast and the convection remained just east of this area. This dry air is clearly visible in Figure 2b and is well into central Pennsylvania by 1955 UTC (Fig. 3). The relatively moist air over Ohio remained convection free.
Figure 4 Isochrones of approximate 50 dbz cores at 1816, 1835, 1903, 1914, 1930, 1947, 1958, and 1925 UTC. Red line shows approximate track of cyclonic rotation. This rotation center triggered several mesocyclone alerts and two TVS s early in it life cycle. Though not analyzed, the convection in Ontario was close to the triple point which can be visualized in Figure 3. Also not that mid-level drying was present in the environment in which this convection developed. Also not the strong (50-60KTS) winds at 850 hpa available for any developing storms to potentially bring to the surface. The summary of the track of the strongest storm of the warm advection phase, across central Pennsylvania is summarized in Figure 4. This figure shows the isochrones of the core of high reflectivity and the approximate path of the cyclonic rotation center associated with this system. Figure 5 shows the SRM and base velocity data at 1941 Figure 3 As in Figure 2 except 1955 UTC water vapor image.
Figure 5 KCCX velocity data valid at 1941 UTC. Upper panels show 0.5 and 1.5 degree storm relative and lower panels show base velocity data at the same elevations slices. UTC. Over Clinton County, the higher velocities and most organized circulation center associated with this elevated thunderstorm cluster are best viewed on the right hand side in the 1.5 degree elevations slice. The lower elevation was in the cool air. As the storm tracked across centeral Pennsylvania, an elevated TVS was detected at 1824Z> The storm moved over Clearfield County at 1835Z. Rotation in the storm throughout its life was best viewed at 1.5 slice. This storm tracked across Clearfield County and the northernmost part of Centre County. It had strong convergence over Clinton County and showed signs of rotation in the 1.5 degree data. The convergence was slightly cyclonic. Cyclonic rotation was visible over the Clinton/Lycoming border at 1952 UTC. Radar winds were on the order of 50-55KTS with the maximum winds of 74KTS observed over Lycoming County. These outbound winds, from the west southwest never made it to the surface based on known reports. This storm tracked into
Figure 6 GOES visual image and surface observations at 1956 UTC 9 November 2005. Columbia and Luzerne Counties persisting until at least 2157 UTC. As stated earlier this storm produced wind damage in Luzerne County. The impact of the cold air is clearly evident in Figure 6 which shows the GOES visual image and the surface observations. At Williamsport, note the east-southeast winds as the thunderstorm, which radar showed strong southwesterlies near the elevated gust front. This image and the Williamsport observation correspond to the last isochrone in Figure 4. Figure 6 also depicts the line of storms associated with the warm advection, the new convection along the approaching cold and occluded fronts, and the low clouds associated with the cold air damming. The latter can be discerned in the surface observations too. The storm at Williamsport (Figure 6) produced 0.5 inch to just under.75 inch hail in the KCTP CWA. Despite several volume scans with in excess of 70KTS of outbound winds, no wind damage was ever reported. This storm eventually produced wind damage in the KBGM CWA, as it moved deeper into the cold air. c) Cold frontal phase
Figure 6 showed the storms along the cold front and southern portion of the occluded front at 09/1956 UTC. By 09/2200 UTC KCCX radar (Fig. 7) showed that this line of storms had crossed Lake Erie. Individual storms, a broken line was evident. Velocity and SRM data (not shown) indicated some rotation in each enhanced cell along the line. In addition to the storms, the surface observations show cool temperatures and easterly winds over much of central and eastern Pennsylvania suggesting that the lowlevel cold air was still entrenched over the region. Figure 8 shows the evolution of the storms between 09/2309 and 09/2325 UTC. Three rotating storms are denoted by the letters A, B, and C. Storm B was warned for by both the Pittsburgh and State College offices. State College warned for Elk County. Not damage was ever associated with this storm. Storm C produced damage in the Pittsburgh County warning area. Storm A was warned for as it crossed Warren County. Earlier, a Special Weather Statement was issued for this storm that covered Warren and McKean Counties. This storm went on to produce damage in both extreme northwestern Elk and southwestern McKean Counties around 2330 UTC. Based on lifted indices and low-level theta-e data, it was clear that Storm A was likely close to the triple point and may have tapped a strong low-level Vorticity source close to the surface allowing it to rapidly spinup and maintain its identity. The storm was tracked across southern McKean County with some rotation until 10/0000 UTC. This storm persisted for well over an hour tracking across southern Warren and McKean Counties between approximately 09/2241 and 10/0000 UTC. A storm farther south produced large hail in Jefferson and Clearfield Counties. This storm too was warned for. Initial reports of 1 inch hail came in from Dubois. Later reports indicated hail of close to 2 inch diameter size. This storm can be seen in Figure 9 and its track is depicted by the white dashed line in the lower panel valid at 09/2352 UTC. Note the strong circulations in these storms at 09/2336 and 09/2352 UTC. Close to the triple point, these storms were able to persist. The observation at Dubois and State College (KUNV) indicated easterly winds and temperatures in the 50s. The low-level cold air likely limited the severe weather to large hail over Clearfield County. Farther north, the intrusion of warm air near the triple point likely allowed the micro-burst to reach the surface. 4. CONCLUSIONS On 9 November 2005, a strong cold front moved across the Mid-western United States entering Pennsylvania during the early evening hours. Cold air locked in at low-levels likely limited the severe weather over central Pennsylvania. The frontal boundary associated with the low-level cold air may have played a role in the elevated convection during the early afternoon hours. Clearly, two distinct waves of convection moved over the region. Initially, an area of enhanced warm advection produced convection over
Figure 7 KCCX radar valid at 09/2200 UTC with 2200 UTC surface observations. Note the line of storms east of Lake Erie. Pennsylvania and New York. The shallow cold air over Pennsylvania precluded the strong winds, in excess of 70 KTS as shown by radar, from reaching the surface. The only potential from severe weather was from hail reaching the surface and isolated areas of strong winds. Strong winds did reach the surface in Luzerne County and hail, slightly smaller than severe size hail, did reach the surface in Lycoming county. This convection was associated with an intrusion of mid-level dry air. West of the mountains, in western New York, these storms were able to produce severe weather. In Pennsylvania, the levated convection was a low probability type to produce wind damage at the surface due to decoupling. However, these storm can and sometimes do produce wind damage at the surface as evidenced by the wind damage reported in Luzerne County The second wave of convection occurred along the occluded and cold fronts. These storms rapidly showed rotation. Unlike the earlier storms, many of them occurred in the warm sector and near the triple point. Until the front encountered the low-level cool air, these storms were better able to produce severe weather which included high winds reaching the surface. The storm that traversed southern Warren, northernwestern Elk, and southern McKean counties was likely close to the triple point. This was evident by examining low-level θe and lifted indices (not shown) suggested the warm sector surged into these areas. The strong rotation in this storm, the
A A a) 2309 C b) 2315 c) 2325 UTC Figure 8 KCCX 0.5 degree reflectivity and SRM data valid at a) 2309, b) 2315, and 2325 UTC 9 November 2005. Labels are skewed as not to obscure the rotation storm to the southwest. Arrows and letters point to storms of interest. proximity of the triple point, and the high winds in the base velocity suggest a warning this storm was a high probability storm to warn on. The
a) 2336 a) 2352 Figure 9 As in Figure 8 except valid at a) 09/2236 and b) 09/2252 UTC. Arrows point to storms A and B from Figure 8 and the dashed line shows the storm which produced hail in Jefferson and Clearfield Counties. downburst in southwestern McKean County and downed trees in northwestern Elk County verify the significance of this storm. and numerous EMA directors and their staffs for relaying information on the storms in near real-time and post event information. 5. ACKNOWLEDGMENTS WCM Dave Ondrejik for storm survey activities in Elk and McKean Counties