Showers and thunderstorms, some with local heavy rainfall, are expected across Pennsylvania and northward to New England. These storms will elevate the potential for flash flooding the rest of this weekend. Hot temperatures are expected across the Desert Southwest and Texas. Monsoonal thunderstorms are expected across New Mexico and Arizona. Read More >
The July 8th, 2001, Mesoscale Convective
System in the Western Carolinas
Bryan P. McAvoy
Author's Note: The following report has not been subjected to the scientific peer review process.
On the evening of 8 July 2001, a classic MCS (Mesoscale Convective System) affected the North Carolina Mountains and a small part of the South Carolina Upstate with damaging winds of 60 to 80 mph. Locally, the winds may have been higher in some mountain valleys which served to funnel the strong winds. The MCS entered the northern mountains of North Carolina shortly before 4:45 p.m. (Fig. 1), and moved between 50 and 60 mph for the next two hours. In other words, in about one hour and 45 minutes, the MCS went from the northernmost tip of Avery County, to the North Carolina - South Carolina border.
Figure 1. KGSP radar base reflectivity on the 0.5 degree scan from 2049 UTC (4:49 p.m.) to 2255 UTC (6:55 p.m.). Click on image to view loop.
The MCS started in the late morning over western Indiana. The system tracked east southeast in excess of 60 mph. During this time it was a classic derecho system producing winds of 70 to 80 mph over a path length of about 300 miles. The Storm Prediction Center (SPC) storm reports from 8 July show where the MCS tracked (Fig. 2). However, there was actually a second system that formed later in the day, and took nearly the same track. In fact, some wind damage was reported after midnight in the North Carolina Mountains as this weaker MCS moved in. For the purposes of this review, however, we will focus on the first MCS.
Click here for a list of severe weather reports for this event.
Figure 2. Hail, damaging wind, and tornado reports received by the Storm Prediction Center for the 24 hour period ending 1200 UTC 9 July 2001.
2. Synoptic Features and Pre-Storm Environment
The reason for the convective system's intensity and rapid movement can be inferred from the 1200 UTC upper air sounding from Wilmington, Ohio (Fig. 3). This was a classic derecho sounding with very high Convective Available Potential Energy (CAPE), and strong, unidirectional wind flow. Modified for a temperature of 89 degrees Fahrenheit and a dewpoint of 74 deg F (rather conservative numbers based on some of the reports ahead of the MCS), the CAPE was 4700 J/kg, and wind speeds changed 40 knots from the surface to 3 km, which was very strong shear.
Figure 3. Skew-T, log P diagram of upper air sounding observed at Wilmington, Ohio (ILN), at 1200 UTC 8 July 2001. A table of selected convective parameters is shown below. Click on image to enlarge.
Interestingly, the numerical models were completely unaware of this system's existence as it formed in association with no minor upper-air troughs. Nor was it in an area of strong warm air advection. The dynamics were not as good by early evening once the MCS begin to move into an area of more northerly shear, as it rounded the periphery of a strong, warm ridge which was centered over the southern Plains. At this time, the MCS began to move on a more southerly course, approaching the Greenville - Spartanburg (GSP) County Warning Area (CWA). The upper air sounding from Blacksburg, Virginia (RNK), is the closest, non-convectively contaminated sounding available (Fig. 4). In fact, it probably represents very well the air mass that the MCS was in as it entered the GSP CWA in the evening of 8 July. At 2345 UTC, there was still a pronounced 700 mb wind maximum in the vicinity of the MCS. However, shear decreased in the mid-levels of the atmosphere. Despite the weakening shear aloft, the system maintained a strong cold pool and new convective development occurred at the leading edge of the outflow boundary all the way into South Carolina. By the time the MCS made it into South Carolina, it had weakened considerably. Available CAPE of 2000 J/kg and slightly lower dewpoints, coupled with weakening shear as the system was getting south of the mid level ridge axis, lead its gradual dissipation.
Figure 4. Skew-T, log P diagram of upper air sounding observed at Blacksburg, Virginia (RNK), at 0000 UTC 9 July 2001. A table of selected convective parameters is shown below. Click on image to enlarge.
3. Radar observations
A gap can be seen in the SPC damage reports just upstream of the GSP CWA as the strong derecho in southern Kentucky lost almost all of its associated leading edge cells a little after 2000 UTC (4 pm). However, new cells began to form over northeast Tennessee a little ahead of the outflow boundary associated with the convection over southern Kentucky. These cells were not very strong with a Vertically Integrated Liquid only up to around 45 (not shown). There was one strong cell which formed ahead of the line in Watauga County, North Carolina but it formed east of where the MCS tracked. In fact, a leading area of convective cells is often thought of as a location where an MCS will frequently bow out. This MCS tracked well west of these leading line cells. However, the new cells did exhibit some mid altitude radial convergence (MARC) as seen from the KGSP radar (Fig. 5). There was about 50 knots of convergence in the mid-levels of the leading edge of the line when it was still north of Mitchell County, North Carolina. It is a little difficult to see in Figure 5, but the bright red on the left most panel is about 35 knots of outbound velocity, and the darker green is 15 knots of inbound velocities. Studies have shown that a MARC signature of this strength is sometimes associated with damaging winds.
Figure 5. KGSP storm relative motion (left) and composite reflectivity (right) at 2059 UTC 8 July 2001. Values are given by color tables on the right side of each image. Click on image to enlarge. 4. Discussion
The active ham radio network in northeast Tennessee was probably the primary reason that the National Weather Service at GSP issued the warnings with substantial lead time. Otherwise, forecasters may have been tempted to wait and see if the line produced any damage as it was not clear if the rear inflow jet would translate to the ground in the mountains. The local storm reports from our office show just how much wind did translate down to the ground. Interestingly, once the MCS exited the mountains, damage reports all but stopped. A plot of wind speeds aloft as the line crossed WFO GSP shows that the system still had severe criteria winds a few thousand feet off the ground, but they were not making it to the surface. In fact, the 35 knot gust at GSP matched very well what the VAD wind profile showed at the surface (Fig. 6).
Figure 6. KGSP Velocity Azimuth Display wind profile ending 2345 UTC 8 July 2001. Click on image to enlarge.
Patrick Moore and Neil Dixon edited this web page so it would conform to the standard template