Event Summary
     National Weather Service, Raleigh NC

July 8, 2008 Excessive Lightning and Severe Weather Event
Updated 2008/10/08





Event Headlines -
...58 severe weather events were documented by the NWS Raleigh...
...33 Severe Thunderstorm, Tornado and Flash Flood Warnings were issued...
...Significant Flash Flooding was reported in Anson County ...
...Forecasters anticipated the potential for thunderstorms with excessive amounts of lightning. The 1120 AM Area Forecast Discussion noted that "several parameters including the high CAPE (both MUCAPE and -10 to -30C CAPE)... K-index of 33-36... and normalized CAPE of 0.25 all support very vigorous updrafts and rapid charge separation... thus the mention of frequent lightning has been added to the forecast through tonight"...
...There were 11,734 cloud to ground lightning strikes in the NWS Raleigh County Warning Area during the 24 hour period ending at 12 UTC on Wednesday, July 9, 2008...
...In Anson County there were 1,001 lighting strikes in the hour ending at 2300 UTC, or an average of around 16.7 strikes per minute...


Event Overview -

The synoptic set up for the July 8, 2008 severe weather event featured a sharp upper-level shortwave trough, evident at 250mb and 500mb approaching North Carolina from the west. By the morning of the 8th (1200 UTC), a closed cyclonic circulation was evident in water vapor imagery rotating through North Carolina (NC) and Virginia (VA). As the upper-level low rotated through the area, the shortwave trough at 500mb became negatively tilted. 500mb temperatures over central North Carolina were around -12 degrees Celsius, 2-4 degrees cooler than surroundings areas. At the surface, dew points rose into the upper 60s and lower 70s across much of the area during the afternoon. Instability quickly increased with surface based CAPE values greater than 3,000 J/kg over much of Central North Carolina and surface based lifted indices (LI) as high as -9 in the northeast portions of the County Warning Area (CWA). The 12 UTC upper air sounding from Greensboro featured a very unstable airmass with surface Convective Available Potential Energy (CAPE, a measure of atmospheric instability) values in the 2000 to 3000 j/kg range, precipitable water values of 1.68 inches, and 0-6 km bulk shear values around 26 kts.

Thunderstorms developed and then moved east along the North Carolina-Virginia border just after noon (16 UTC), with another area of thunderstorms just south of Raleigh at around 2pm (18 UTC). Eventually, more storms developed and merged into a line and moved through northeastern North Carolina. While these storms were occurring, a second round of convection was already getting organized well upstream of central North Carolina in eastern Tennessee at around 1pm (17 UTC). By 5pm (21 UTC), a line of storms had formed in southwestern North Carolina and was approaching the southwestern portion of the Raleigh CWA. The storms pushed through the Southern Piedmont and Sandhills around sunset (9pm on the 8th, or 01 UTC on the 9th). Finally, the line of storms weakened over southeastern North Carolina, with the outflow from the storms racing east out well ahead of the updrafts. The final episode of convective storms occurred around midnight (04 UTC on the 9th) in Anson and Richmond Counties in the southern Piedmont. This area had already received tremendous rainfall, up to 4-5 inches, from the previous convective episode, which lead to flash flooding overnight.

Combined, the three rounds of storms were responsible for widespread wind damage and a few severe hail reports from northern Georgia to southern Virginia. In addition, the number of cloud-to-ground lightning strikes observed with these storms was remarkably high. At one point, there were over 2,000 strikes in a one hour period across the Piedmont of southern Virginia, North Carolina, and northern South Carolina. Forecasters at the National Weather Service Forecast Office in Raleigh, NC, have undertaken a research project to anticipate the level of lightning activity, in hopes of providing the public with an improved awareness of events where the number of lightning strikes will be exceptionally high. Forecasters look for certain conditions that are needed for the production of lightning by completing a checklist each morning. More information on the lightning forecast process is discussed below.


Severe Weather Reports -
Text of severe weather reports across central North Carolina





Regional Radar Loop

A Java Loop of regional reflectivity imagery from 1538 UTC on July 8 through 0758 UTC on July 9, 2008 is available here. Note - this loop includes 68 frames.

The regional reflectivity image below is from 0058 UTC (858 PM EDT on July 8, 2008). A convective cluster was pushing through the Southern Piedmont and Sandhills at this time. Around this time or just a few minutes earlier, these storms produced a wind gust up to 58 MPH in southern Wake County and wind damage in Harnett and Johnston Counties.


Regional reflectivity image - click to load loop



KRAX Radar Loops

A Java Loop overview of the entire event with images from every hour between 1602 UTC July 8 through 0700 UTC July 9, 2008 is available here. Note - this loop includes 16 frames

A Java Loop overview of the entire event with images from every 15 minutes between 1602 UTC July 8 through 0700 UTC July 9, 2008 is available here. Note - this loop includes 60 frames

A Java Loop overview of the entire event with images from every volume scan between 1602 UTC July 8 through 0700 UTC July 9, 2008 is available here. Note - this loop includes 206 frames

The KRAX reflectivity image below is from 2254 UTC or 654 PM EDT on July 8, 2008 just after wind damage was reported in Anson, Stanly, Richmond, and Randolph Counties.





Extreme Lightning


Cloud to ground lightning strikes in 5 by 5 km grid boxes during the 24 hour period ending at 12 UTC on Wednesday, July 9, 2008 - click to enlarge There were 48,912 cloud to ground lightning strikes across the Carolinas and the southern Mid Atlantic region during the 24 hour period ending at 12 UTC on Wednesday, July 9, 2008. The image to the right was produced from the Graphical Forecast Editor in AWIPS (the NWS's computer platform for data analysis and forecasting) and shows the number of cloud to ground lightning strikes in 5 by 5 km grid boxes during this period. The warmer the colors, the greater the number of lightning strikes. Purple shading indicates fewer than 20 strikes per 5 x 5 km grid box, blue shading indicates between 30 and 60 strikes, yellow shading between 100 and 120 strikes, with more than 130 strikes in the orange and red shading.

In just the NWS Raleigh CWA there were 11,734 strikes detected. The lightning strikes were concentrated in a few locations primarily across the Southern Piedmont in Anson County; across eastern portions of the Sandhills in Harnett, Cumberland, and Lee Counties; along the VA-NC border in northern portions of Person, Granville, and Vance Counties; and along the Coastal Plain from Johnston County northward into Halifax County.

click to control, stop, or start the loop The looping image to the right (click here to control, stop, or start the loop) is from the AWIPS D2D application showing the cloud to ground lighting strikes in 15 minute intervals across central North Carolina from 1500 UTC on July 8 through 0900 UTC on July 9, 2008. Negative lightning strikes are shown with a "-" symbol and positive lightning strikes are shown with a "+" symbol. (Click here for more information on lightning from the NWS's Jetstream Weather School.)

The progression of organized convection during the afternoon and evening hours can be seen in the loop. The initial convective complex developed and intensified across southern Virginia and far northern North Carolina. The thunderstorms moved east across northern portions of Person, Granville, and Vance Counties. Another area of convection developed across the Coastal Plain in a line just west of Interstate 95 and then intensified as they moved into the northern and central Coastal Plain.

A third convective complex with exceptional amounts of lightning moved from the Southern Foothills into the Southern Piedmont during the early evening hours. These thunderstorms produced a significant amount of lightning across Anson County and then moved east into the Sandhills.

Finally, another thunderstorm cluster developed around midnight in the Southern Piedmont and produced over a thousand cloud to ground lightning strikes, very heavy rain and flash flooding in Anson County.


Anson County Storm

The greatest cloud to ground lightning activity during this event was located across Anson County where 2,212 strikes were detected during the 24 hour period ending at 1200 UTC on July 9. More impressively, there were 1,001 lighting strikes in the hour ending at 2300 UTC, or an average of around 16.7 strikes per minute during the hour across Anson county. The sampling area that approximately covers Anson County is across 41 5x5 km grid boxes, or 1,025 km². This corresponds to an average of 2.16 cloud to ground lightning strikes per km² during the 24 hour period ending at 1200 UTC on July 9, with 0.98 cloud to ground lightning strikes per km² in the hour ending at 2300 UTC.

Cloud to ground lightning strikes across Anson County and adjacent areas during a 15 minute period ending at 2300 UTC (700 PM EDT) on July 8, 2008 are shown below. Note that 603 lightning strikes were observed over this area during the 15 minute period with an average of 40 lightning strikes each minute.

This loop shows cloud to ground lightning strikes every 15 minutes across Anson County and adjacent areas on July 8, 2008.

Cloud to ground lightning strikes during 15 minute periods ending at 2300 UTC across Anson County and adjacent areas on July 8, 2008 - click to enlarge



Harnett, Cumberland, and Lee County Storm

A large number of cloud to ground lightning strikes during this event were located across Harnett, Cumberland, and Lee Counties, where 1,649 strikes were detected during the 24 hour period ending at 1200 UTC on July 9. More impressively, there were there were 1,110 lighting strikes in the hour ending at 0100 UTC or an average of around 18.5 strikes per minute during the hour. The sampling area for the Harnett, Cumberland, and Lee County data is across 179 5x5 km grid boxes or 4,475 km².

Cloud to ground lightning strikes across Harnett and Cumberland Counties and adjacent areas during a 15 minute period ending at 0015 UTC (815 PM EDT) on July 8, 2008 are shown below. Note that 425 lightning strikes were observed over this area during the 15 minute period with an average of 28 lightning strikes each minute.

This loop shows cloud to ground lightning strikes every 15 minutes across Harnett and Cumberland Counties and adjacent areas on July 8, 2008.

Cloud to ground lightning strikes during a 15 minute period ending at 0015 UTC across Harnett and Cumberland Counties and adjacent areas on July 8, 2008 - click to enlarge



Anticipating Extreme Lightning

Lightning is produced when there is a buildup of electrical charges within a cloud. In a typical thunderstorm structure, as the updraft increases, positive charges are vaulted into the upper portion of the storm, negative charges pool in the mid levels of the storm, and another pool of positive charges collects near the cloud base. Negative charges are sent down the storm in what is called a stepped leader, which then draws a stream of positive charge upward called a "positive streamer." As these two charges come together, the electric current forms.

This slow motion movie clip below from the BBC documentary The Power of the Planet Atmosphere shows the negative charges "snaking" their way down from the cloud bases. These negative charges are called "stepped leaders" and when one of them is close enough to the surface, it draws a stream of positive charge upward. As these two charges come together, the electric current forms and the lightning strike takes place.



Within a storm, two primary ingredients are critical for lightning formation: (1) the presence of graupel, or small balls of water-coated snow (collisional charging between graupel pellets and lighter ice crystals facilitates lightning production) in the layer between approximately -10 and -30 degrees Celsius, referred to as the "mixed phase" region of the storm, where ice coexists with supercooled water; and (2) strong instability in this same layer of the storm, to facilitate strong updraft velocities and rapid charge separation within the storm. While we cannot measure precisely the presence of graupel inside the storm and the velocity of the storm's updraft, several near-storm parameters can indicate a favorable environment for extreme amounts of lightning.

One such parameter that forecasters monitor to anticipate extreme amounts of lightning is the layer CAPE within the -10 to -30 degree Celsius layer (the "mixed phase" region of the storm). Strong instability within this layer has been shown to be favorable for rapid hail growth. In local studies of extreme lightning events, values over 200 J/kg have been associated with high-frequency lightning strike events.

Another parameter that helps forecasters measure the potential for vigorous updrafts which favor lightning is the normalized CAPE, or N-CAPE. N-CAPE is the CAPE measured from the LCL (lifted condensation level) to the EL (equilibrium level), divided by the depth of that layer, and it gives a measure of the "shape" of the CAPE. A wide or "fat" CAPE will equate to high N-CAPE and indicates the potential for strong updrafts, whereas a narrow or "skinny" CAPE will have a low N-CAPE and indicates weak updrafts. N-CAPE values above 0.1, and especially those above 0.2, indicate a "fat" CAPE and better chance for very strong updrafts.

Also important to consider is the potential for sufficient graupel in the "mixed phase" region of an updraft. The amount of moisture present in the hail growth layer (-10 to -30C) is difficult to quantify. This quantity is not readily available from numerical models for forecasters, so forecasters are forced to subjectively identify whether or not there is sufficient moisture available. However, past studies have shown that column total precipitable water may be a good indicator, even though it does not isolate the "mixed phase" layer. Upcoming refinements to the lightning forecast process may include evaluation of precipitable water.

Finally, forecasters also consult experimental output from two specialized model-based algorithms run by the Storm Prediction Center (SPC). One algorithm is based on the NAM model, while the other is based on the Short Range Ensemble Forecast (SREF). Both sets of output show the probability of 100 or more lightning strikes. These products, along with the other parameters mentioned above, are considered together by forecasters by means of the lightning checklist.


Lightning Checklist

In order to quickly assess the potential for high-frequency lightning activity, forecasters complete a daily checklist on each midnight shift, with the option to update the checklist as new model data and observations arrive during the day. The checklist is broken up into three categories of lightning activity - low, medium, or high. Corresponding thresholds for the parameters discussed above were chosen for each level based on previous research and experience. The parameters that are objectively evaluated are K-index, Most Unstable CAPE (MUCAPE), CAPE in the -10C to -30C layer, Normalized CAPE (N-CAPE), SPC-produced model-based probabilities, and persistence of the weather pattern (what was the activity level the day before?). Other parameters that are more subjectively evaluated include the moisture in the hail growth layer, the presence of synoptic forcing, and the location and orientation of high equivalent potential temperature air at 850mb. Each of these factors was noted in past studies as being correlated with excessive lightning events.

Lightning Checklist from July 8, 2008 - click to enlarge


During the early morning hours of the 8th, forecasters completed the checklist, and noted several indicators that would suggest that storms would produce an unusually high number of lightning strikes. The main parameters that suggested high potential included
  • K-index values above 30. (K-index measures mid-level lapse rates and moisture.)
  • CAPE in the -10C TO -30c layer. (Values of greater than 200 J/kg suggest strong updrafts, and the forecast layer CAPE on the 8th was over 900 J/kg.)
  • Normalized CAPE of 0.25.
  • Persistence. Forecasters noted that lightning activity was high the day before, and there would be little change to the synoptic pattern and air mass.

    Regarding the more subjective parameters, forecasters also noted that there would be a strong synoptic feature in the vicinity of central North Carolina to initiate convection. Specifically, an upper-level low would be over Central Virginia, with a strong vorticity maximum rotating through North Carolina with the shortwave mentioned in the event overview. The vorticity center and its differential positive vortcity advection (DPVA) would provide strong lift to help develop widespread thunderstorms. The final evaluation on the checklist is of any ridging of equivalent potential temperature at 850mb, which in this case was found to be fairly uniform across the entire CWA. Forecasters will often use extra space on the checklist to describe spatial differences in the values of parameters across the CWA, but given that no comments were made, and that the 850mb equivalent potential temperature would be essentially uniform, there was no indication that one area would be favored for high lightning activity versus another.

    As noted above, in just the NWS Raleigh CWA there were 11,734 strikes detected. The lightning strikes were concentrated in a few locations primarily across the Southern Piedmont in Anson County; across eastern portions of the Sandhills in Harnett, Cumberland, and Lee Counties; along the Virginia-North Carolina border in northern portions of Person, Granville, and Vance Counties; and along the Coastal Plain from Johnston County northward into Halifax County. Although the task of identifying individual storms that will realize their full lightning potential is difficult, focusing on parameters that lead to high lightning frequency can assist forecasters in targeting days where the environment is conducive to more frequent lightning strikes.

    Observed lightning output - click to enlarge


    Analysis Data

    Two of the main forecast parameters are readily available via the SPC mesoanalysis Page, Most Unstable CAPE and CAPE in the -10 to -30C layer. These plots can be used to show what the conditions were in and around Anson County at the time of the excessive lightning. At 23 UTC, Most Unstable CAPE values were highest, from 2000 to 3000 J/kg, from Anson County eastward into the Sandhills. This is a good indicator that the environment in that area was highly unstable and conducive to strong convection. Even more representative of the potential for very strong updrafts is the CAPE in the -10 to -30C layer, which was 250 to 300 J/kg in the Southern Piedmont, with a maximum analyzed very near Anson County. Finally, the K-index is not analyzed on the SPC mesoanalyis page, but can be computed using various mesoanalysis products and the equation (T850 - T500) + Td850 - (T700 - Td700), where T is the temperature and Td is the dewpoint temperature). Each value can be approximated from the upper air analyses (850mb, 700mb, 500mb). An evaluation of the K-index based on these images gives a value of 33C, which is exactly what was noted by forecasters 18 hours earlier in the lightning checklist.


  • Mesoscale Data

    Analyzed surface pressure and wind barbs from SPC at 21 UTC on Tuesday, July 8, 2008
    A weak surface trough extended across the Western Piedmont providing some weak convergence.

    SPC Analysis at 21 UTC on Tuesday, July 8, 2008



    Analyzed surface Theta-E (green) and Theta-E advection (purple) from SPC at 21 UTC on Tuesday, July 8, 2008
    The greatest theta-e values ranged in an axis from northeast South Carolina into southern North Carolina where theta-E values 352K or greater.

    SPC Analysis at 21 UTC on Tuesday, July 8, 2008



    Analyzed low level lapse rates in the 0-3 km layer (blue, green, and orange) from SPC at 21 UTC on Tuesday, July 8, 2008
    A lapse rate is the rate of temperature change with height and the image below is for the layer from the surface to around 10,000 feet. Note the surface based, low level lapse rates shown below range in the 7.0 to 7.5 deg C/km across much of central North Carolina. Values less than 6 degrees C/km represent "stable" conditions, while values near 9 degrees C/km are considered "absolutely unstable." An axis of steep low level lapse rates extended from north to south across central North Carolina.

    SPC Analysis at 21 UTC on Tuesday, July 8, 2008



    Analyzed surface based convective available potential energy (SBCAPE) (red) and surface based convective inhibition (blue lines - shaded) from SPC at 21 UTC on Tuesday, July 8, 2008
    SBCAPE values ranged between 2500 and around 3000 J/kg across eastern portions of the Southern Piedmont and the Sandhills with no significant convective inhibition (CIN).

    SPC Analysis at 21 UTC on Tuesday, July 8, 2008



    Analyzed mixed layer convective available potential energy (MLCAPE) (red) and mixed layer based convective inhibition (MLCIN) (blue lines - shaded) from SPC at 21 UTC on Tuesday, July 8, 2008
    MLAPE values ranged between 1500 and around 2000 J/kg in a north-south axis across central North Carolina with no significant convective inhibition (MLCIN).

    SPC Analysis at 21 UTC on Tuesday, July 8, 2008



    CAPE in the layer from -10 C to -30 C, 0-6-km shear vector, and the freezing level height from SPC at 21 UTC on Tuesday, July 8, 2008
    Large CAPE in the layer from -10 C to -30 C favors rapid hail growth. 0-6-km shear in excess of 30-40 knots supports supercells with persistent updrafts that contribute to large hail production. Finally, lower freezing level heights suggest a greater probability of hail reaching the surface prior to melting, though melting impacts small hail much more than very large hailstones.

    SPC Analysis at 21 UTC on Tuesday, July 8, 2008



    Normalized CAPE or (NCAPE) (red) from SPC at 21 UTC on Tuesday, July 8, 2008
    The NCAPE (Normalized CAPE) is CAPE that is divided by the depth of the buoyancy layer (units of m s**-2). Values near or less than 0.1 suggest a "tall, skinny" CAPE profile with relatively weak parcel accelerations, while values closer to 0.3 to 0.4 suggest a "fat" CAPE profile with large parcel accelerations possible.

    SPC Analysis at 21 UTC on Tuesday, July 8, 2008



    NWS Composite Reflectivity Imagery from 2130 UTC on Tuesday, July 8, 2008 (530 PM EDT).
    The composite reflectivity imagery is from the approximate time in which the analysis imagery above is valid.

    Composite Reflectivity Imagery from 2130 UTC on Tuesday, July 8, 2008



    Archived Text Data from the Severe Weather Event

    Select the desired product along with the date and click "Get Archive Data."
    Date and time should be selected based on issuance time in GMT (Greenwich Mean Time which equals EDT time + 4 hours).


    Product ID information for the most frequently used products...

    RDUAFDRAH - Area Forecast Discussion
    RDUZFPRAH - Zone Forecast Products
    RDUAFMRAH - Area Forecast Matrices
    RDUPFMRAH - Point Forecast Matrices
    RDUHWORAH - Hazardous Weather Outlook
    RDUNOWRAH - Short Term Forecast
    RDUSPSRAH - Special Weather Statement
    RDULSRRAH - Local Storm Reports (reports of severe weather)
    RDUSVRRAH - Severe Thunderstorm Warning
    RDUSVSRAH - Severe Weather Statement
    RDUTORRAH - Tornado Warning


     from 



    Selected Photographs of the Severe Weather Event

    Photos courtesy of John Hamilton, Mark Calaway, and Michael Moneypenny.
    (Click the image to enlarge)



    Large thunderstorm over Stanly and Montgomery Counties - photo courtesy of John Hamilton - Click to enlarge           Thunderstorm near sunset as seen from near Fuquay-Varina - photo courtesy of Mark Calaway - Click to enlarge           Thunderstorm near sunset as seen from near Fuquay-Varina - photo courtesy of Mark Calaway - Click to enlarge

    Thunderstorm near Fuquay-Varina - photo courtesy of Mark Calaway - Click to enlarge           Thunderstorm near Fuquay-Varina - photo courtesy of Mark Calaway - Click to enlarge           Thunderstorm near Fuquay-Varina - photo courtesy of Mark Calaway - Click to enlarge

    Large uprooted tree on Sheriff Watson Road in Lee County - photo courtesy of Michael Moneypenny - Click to enlarge           Huge pecan fallen on trailer on Sheriff Watson Road - photo courtesy of Michael Moneypenny - Click to enlarge           Ten inch trees down on Douglas Farm Road - photo courtesy of Michael Moneypenny - Click to enlarge


    Lessons Learned

    Forecasters anticipated the potential for thunderstorms with excessive amounts of lightning. The 1120 AM Area Forecast Discussion noted that "several parameters including the high CAPE (both MUCAPE and -10 to -30C CAPE)... K-index of 33-36... and normalized CAPE of 0.25 all support very vigorous updrafts and rapid charge separation... thus the mention of frequent lightning has been added to the forecast through tonight".

    The excessive lightning checklist was successfully used preceding this event to highlight the potential for excessive amounts of lightning. This high lightning potential was mentioned in the Zone Forecast Product and the Hazardous Weather Outlook.



    Acknowledgements

    Many of the images and graphics used in this review were provided by parties outside of WFO RAH. The surface analysis graphic was obtained from the Hydrometeorological Prediction Center. The upper air analysis images were obtained from the University of Wyoming. Lightning data was collected from AWIPS and the Graphical Forecast Editor. lightning data made available in GFE via the Ltg Procedure created by Timothy Barker. Objective analysis of the Graphical Forecast Editor lightning data was produced via the Lightning Tools Procedure created by Ryan Knutsvig. Slow motion lightning movie clip is from the BBC documentary The Power of the Planet Atmosphere. Photos courtesy of John Hamilton, Mark Calaway, and Michael Moneypenny.



    Case study team -
    Gail Hartfield
    Barrett Smith
    Lara Pagano
    Jonathan Blaes

    For questions regarding the web site, please contact Jonathan Blaes.


  • NWS Disclaimer.