May 20, 2008 Severe Weather Event

Event Summary
National Weather Service, Raleigh NC

May 20, 2008 Severe Weather Event
Updated 2008/07/12

Collage of images related to this event

Event Headlines -
...98 severe weather events were documented by the NWS Raleigh with several dozen more reports received containing similar or less significant damage...
...37 Severe Thunderstorm or Tornado Warnings were issued between 229 PM and 830 PM EDT...
...Several supercells produced large hail and wind damage along paths that stretched across several counties...
...The NSSL Hail Swath product clearly showed the track of several hail producing supercells that tracked across central North Carolina...

Event Overview -
A broad upper level trough with several embedded short waves was located across the eastern United States on Tuesday, May 20, 2008. At the same time a cold front was pushing south across Virginia into North Carolina. The atmosphere ahead of the front was moderately unstable with a vigorous wind field. Convection developed before midday across the southern Appalachians and Virginia, moved southeast and intensified in the form of a large line of convection with a few discrete cells. The thunderstorms became severe with numerous reports of large hail with some wind damage.

Event Details -
The upper air pattern across the United States on Monday and Tuesday, May 19th and 20th, 2008 featured a large ridge of high pressure across the Rocky Mountain region and a large trough over the eastern U.S.. A strong cyclonic jet was present in the large eastern U.S. trough. Several shortwave troughs were rotating around the deep mid level vortex centered over southeastern Canada.

A cold front moved southeastward across central North Carolina during the early morning hours of Monday, May 19, 2008. The front stalled across the Tennessee Valley and the southern Carolinas on Monday afternoon. A significant shortwave trough was diving south across the upper Mississippi Valley on Monday. A weak surface low developed along the stalled frontal system in the lower Tennessee Valley on Monday evening and the associated warm front across the Carolinas lifted northward allowing the return of moisture to begin across the Carolinas.

The surface low associated with a shortwave across the Ohio Valley moved eastward across Maryland and Virginia early on Tuesday. The trailing cold front slipped south across southern Virginia during the late morning and early afternoon hours on Tuesday.

A large area of mid and high level cloudiness across North Carolina that was enhanced orographically during the morning hours diminished significantly during the afternoon. The increasing amounts of sunshine allowed surface temperatures to warm into the mid to upper 80s. The atmosphere over central North Carolina became increasingly moist during the day as indicated by the NOAA IPW (Integrated Precipitable Water) data and surface dew points which climbed into the lower to mid 60s. The moisture can also be seen in the plume of higher precipitable water values that extended northeastward out of the western Gulf of Mexico.

An Elevated Mixed Layer (EML) originated across the Southern Plains region and then moved east near the Gulf Coast and into the Southeast U.S.. The EML can be seen in the moderately steep mid level lapse rates of 7 to 7.5 deg C/km over central and southern North Carolina around midday Tuesday afternoon. The low level moisture and the moderately steep lapse rates contributed to a moderately unstable air mass with mixed layer CAPE's approaching 1000-1500 J/KG during the early afternoon.

The wind field across the region was impressive with the 12Z May 20, 2008 RAOB from KGSO ( Skew-T | Text listing) indicating a 32 knot southwesterly wind (240 deg) at 925 mb, a 44 knot westerly wind (265 deg) at 850 mb, a 57 knot westerly wind (265 deg) at 700 mb, and a 61 knot westerly wind (270 deg) at 500 mb. The bulk shear values remained strong during the afternoon hours and supported organized convection with 0 to 6 km shear values in excess of 50 knots across North Carolina. The wind field was largely unidirectional which would support the evolution of convection into short line segments and bows with some splitting supercells.

The effective storm relative helicity values ranged between 150 and 200 m2/s2 which was sufficient for low level mesocyclones and isolated tornadoes, mainly with discrete supercells. But the primary threat across central North Carolina was wind damage and large hail.

The scattered showers and thunderstorms that moved across West Virginia and Virginia during the late morning hours began to intensify just after midday. The convection became more significant as convergence along the front, increasing instability, and the lift associated with the mid level short wave contributed to increasing the intensity and coverage of convection. A Severe Thunderstorm Watch was issued by the Storm Prediction Center at 110 PM for much of northern and central North Carolina through 800 PM.

The first Severe Thunderstorm Warning issued by NWS Raleigh was at 229 PM EDT as the line of storms dropped into northern North Carolina. Additional warnings were issued as the line dropped south and east with several reports of wind damage and hail. The thunderstorm at the southern end of the line in northern Davidson county developed into a long lived supercell just after 400 PM EDT. This storm produced damage as it moved east all the way to Wake County when it merged with a convective line segment. Just before 500 PM EDT the Storm Prediction Center issued an update on the Severe Thunderstorm Watch at 438 PM noting that isolated supercells were still possible, mainly south and east of the large convective line.

Another significant supercell developed near Charlotte around 400 PM EDT and moved east reaching Stanly County just before 500 PM EDT. This storm intensified and produced damage as it moved across southern North Carolina. A Tornado Warning was issued for southern Cumberland County at 704 PM EDT and a funnel cloud was observed by the Hope Mills police near downtown Hope Mills but no damage was reported. The convection continued to move south and east and merge into convective line segments and clusters before moving out of the RAH CWA by around 930 PM EDT.

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

Severe weather reports via SPC from May 20-9, 2008

Severe weather reports via SPC from May 20-9, 2008

Regional Radar Loop

A Java Loop of regional reflectivity imagery from 1558 UTC on May 20 through 0158 UTC on May 21, 2008 is available here. Note - this loop includes 39 frames.

The regional reflectivity image below is from 2138 UTC on May 20, 2008. At this time large hail was being reported in Wake 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 volume scan between 1702 UTC May 20 through 0128 UTC May 21, 2008 is available here. Note - this loop includes 112 frames

The KRAX reflectivity image below is from 2116 UTC or 516 PM EST on Tuesday, May 20, 2008 when large hail was being reported in Stanly and Wake Counties.

KRAX Radar Examples

KRAX 4 panel radar image from 2053Z on Tuesday, May 20, 2008 over Pittsboro in Chatham County. Around this time large trees were being blown over in Pittsboro with hail larger an inch in diameter falling in the area.

The radar image shows the Base Reflectivity product in the upper left, the Cell Based VIL (greater than 70 in white and greater than 100 in dark gray) in the upper right, the Enhanced Echo Tops (greater than 40 thousand feet in yellow) in the lower left, and the Maximum Expected Hail Size (greater than 3 inches in the gray pixels) in the lower right.

4 panel radar image from 2053Z on Tuesday, May 20, 2008 - click to enlarge

KRAX 4 panel radar image of velocity data from 2301Z on Tuesday, May 20, 2008 near Hope Mills in Cumberland County. Around this time large hail up to golf ball size was reported in southwestern Cumberland County and trees were blown down just southeast of Hope Mills. A funnel cloud was observed by the Hope Mills police near downtown Hope Mills but no damage was reported.

The radar image shows the Storm Relative Velocity imagery from 2301Z at 0.5, 0.9, 1.3, and 1.8 degrees. The 0.5 degree data shows 29 knots of flow away from the radar and 45 knots toward the radar in a gate to gate fashion at around 4,200 feet. The 0.9 degree data shows 38 knots of flow away from the radar and 53 knots toward the radar in nearly a gate to gate fashion at around 6,400 feet. The circulation can still be seen but it is much weaker in the 1.3 degree data at around 8,600 feet. The circulation is poorly defined further up at 1.8 degrees, or around 11,000 feet. Funnel clouds were observed but no tornado touchdown was confirmed.

4 panel radar image from 2301Z on Tuesday, May 20, 2008 - click to enlarge

KRAX 4 panel radar image of reflectivity data from 2301Z on Tuesday, May 20, 2008 zoomed in over Hope Mills in Cumberland County. Around this time large hail up to golf ball size was reported in southwestern Cumberland County and trees were blown down just southeast of Hope Mills. A funnel cloud was observed by the Hope Mills police near downtown Hope Mills and there was some wind damage reported southeast of Hope Mills but no tornado touchdown was confirmed.

The radar image shows the Base Reflectivity imagery from 2301Z at 0.5, 0.9, 1.3, and 1.8 degrees. The 0.5 degree data shows a broad hook shape signature that extends back toward Parkton in northeastern Scotland County.

KRAX 4 panel radar image from 2310Z on Tuesday, May 20, 2008 over Cumberland County. Around this time large hail up to golf ball size was reported in southwestern Cumberland County.

The radar image shows the Base Reflectivity product in the upper left, the Cell Based VIL (greater than 70 in white and greater than 100 in dark gray) in the upper right, the Enhanced Echo Tops (greater than 40 thousand feet in yellow) in the lower left, and the Maximum Expected Hail Size (greater than 2.5 inches in the purple pixels) in the lower right.

4 panel radar image from 2310Z on Tuesday, May 20, 2008 - click to enlarge

NSSL MESH Hail Swath Product

NOAA's National Severe Storms Laboratory (NSSL) has been developing techniques for getting popular WSR-88D cell-based hail information from the Hail Detection Algorithm (HDA) into formats that users can more effectively use. Some of the cell-based hail information has been incorporated into high-resolution gridded products that allow users to diagnose which portions of storms contain large hail. One such product is the "Hail Swath" product which accumulates hail size data over a period of time to provide hail swath maps, showing both maximum hail size by location, and hail damage potential (a combination of hail size and how long the hail has been falling).

More specifically, the "Hail Swath" product is a derivative of the MESH (Maximum Expected Size of Hail) output from the HDA. Reflectivity data from all of the CONUS NEXRAD radars are merged into a three dimensional (latitude/longitude/height) grid. A modified version of the NSSL HDA is then run on this grid producing a MESH grid at 60 second intervals. The maximum MESH value at each 1 square kilometer grid point is plotted over a chosen time period in order to create the areal �swaths� of MESH.

The NSSL Hail Swath product valid from the afternoon and evening of May 20, 2008 is shown along with storm reports of large hail (H) and wind damage (W) in the image above and/or to the right (click on the image for a larger view). You can clearly see the track of several hail producing supercells that moved across central North Carolina. The color scale on the image indicates the maximum expected hail size during the period with the light blue indicating hail estimated at around a half inch in diameter, the darker blue indicating hail greater then an inch in diameter, and the lighter green indicating hail potentially larger then an inch and a half.

More information on the MESH is available from the Verification of multi-sensor, multi-radar hail diagnosis techniques by Kiel L. Ortega, Travis M. Smith, and Gregory J. Stumpf.

CoCoRaHS Observer Network

CoCoRaHS is a grassroots volunteer network of weather observers of all ages and backgrounds working together to measure and map precipitation (rain, hail and snow) in their local communities. By using low-cost measurement tools, stressing training and education, and utilizing an interactive Web-site, CoCoRaHS is to provide the highest quality data for natural resource, education and research applications. The only requirements to join are an enthusiasm for watching and reporting weather conditions and a desire to learn more about how weather can effect and impact our lives. For more information, visit the CoCoRaHS web site at www.cocorahs.org.

The CoCoRaHS Web page provides the ability for CoCoRaHS observers to see their observations mapped out in "real time", as well as providing a wealth of information for our data users. The NWS Raleigh has been actively utilizing CoCoRaHs data for a variety of uses including real time precipitation and data analysis, severe weather operations, and post storm reviews.

CoCoRaHS observers are encouraged to submit critical weather observations such as large hail and very heavy rain to the CoCoRaHS web page as soon as it is safe to do so. This is in addition to the regular 24 hour precipitation report submitted each morning. These significant or critical weather observations are sent to the NWS and immediately alarmed so that forecasters can use the data as a part of the forecast process. A compilation of the text products sent to the NWS Raleigh office containing CoCoRAHS spotter reports for central North Carolina on May 20, 2008 illustrates the various types of CoCoRAHS spotter reports including heavy rain, hail, and even photos of hail. Forecasters also have the ability to search and view hail reports via the CoCoRAHS web site.

The maps below are from the CoCoRaHS web site and they show the CoCoRAHS hail reports and the 24 hour CoCoRAHS precipitation reports.

CoCoRAHS Hail Reports

24 hour CoCoRAHS Precipitation Reports

Mesoscale Data

Analyzed surface pressure and wind barbs from SPC at 20 UTC on Tuesday, May 20, 2008
A surface low was located off the Virginia coast with a trailing cold front extending southwest across northern North Carolina.

SPC Analysis at 20 UTC on Tuesday, May 20, 2008

SPC Analysis at 20 UTC on Tuesday, May 20, 2008

Analyzed surface temperatures (red/purple), dew points (brown/green), and wind barbs from SPC at 20 UTC on Tuesday, May 20, 2008
A cold front stretched from Southeast Virginia southwestward across northwestern North Carolina. Temperatures ahead of the cold front were in the lower to mid 80s with dew points in the lower 60s.

SPC Analysis at 20 UTC on Tuesday, May 20, 2008

Analyzed low level lapse rates in the 0-3 km layer (blue, green, and orange) from SPC at 20 UTC on Tuesday, May 20, 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 6.5 to 7.0 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." The lower portion of the atmosphere on May 20 was obviously very unstable.

SPC Analysis at 20 UTC on Tuesday, May 20, 2008

Analyzed mid level lapse rates in the 700-500 mb layer (blue, green, and orange) from SPC at 20 UTC on Tuesday, May 20, 2008
A lapse rate is the rate of temperature change with height and the image below is for the layer from around 10,000 to 18,000 feet. Note the mid level lapse rates shown below range around 6.5 deg C/km across much of central North Carolina which was only slightly unstable. The greatest mid level instability, associated with the elevated mixed layer was present across far southern North Carolina, South Carolina, and Georgia.

SPC Analysis at 20 UTC on Tuesday, May 20, 2008

Analyzed surface based convective available potential energy (SBCAPE) (red) and surface based convective inhibition (blue lines - shaded) from SPC at 20 UTC on Tuesday, May 20, 2008
SBCAPE values ranged between 1500 and around 2000 J/kg ahead of the cold front in central North Carolina with no significant convective inhibition (CIN).

SPC Analysis at 20 UTC on Tuesday, May 20, 2008

0-3 Km Storm Relative Helicity (blue) and storm motion (brown) from SPC at 20 UTC on Tuesday, May 20, 2008
The 0-3 Km SRH values ranged between 100 and 200 units across central North Carolina with the greatest values in southern North Carolina.

SPC Analysis at 20 UTC on Tuesday, May 20, 2008

CAPE in the layer from -10 C to -30 C, 0-6-km shear vector, and the freezing level height from SPC at 20 UTC on Tuesday, May 20, 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 20 UTC on Tuesday, May 20, 2008

Analyzed Lifting Condensation Level (red, blue, and green) from SPC at 20 UTC on Tuesday, May 20, 2008
The LCL height is the height at which a parcel becomes saturated when lifted dry adiabatically. The importance of LCL height is thought to relate to sub-cloud evaporation and the potential for outflow dominance. Low LCL heights imply less evaporational cooling from precipitation and less potential for a strong outflow that would likely inhibit low-level mesocyclone development. Thunderstorms that produce tornadoes generally have a lower LCL height with LCL heights less than 1,000 meters typically favorable for significant tornado development. The LCL values during this event were around 1,500 meters.

SPC Analysis at 20 UTC on Tuesday, May 20, 2008

NWS Composite Reflectivity Imagery from 0130 UTC on Thursday, May 20, 2008 (930 PM EST).
The composite reflectivity imagery is from the approximate time in which the analysis imagery above is valid.

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, David Stark, Graham Johnson, Billie Johnson, and emergency management officials in Anson and Chatham Counties.
(click on the image to enlarge.)

Photo of mammatus clouds in Cary at around 545 pm on May 20th. Photo courtesy of David Stark. - Click to enlarge

A large tree was blown over. Photo courtesy of Chatham County emergency management. - Click to enlarge

A large tree was blown over into a trailer. Photo courtesy of Chatham County emergency management. - Click to enlarge

A large tree was blown over onto a house. Photo courtesy of Chatham County emergency management. - Click to enlarge

Large tree fell onto a home. Photo courtesy of Anson County emergency management. - Click to enlarge

Hail accumulation in Raeford. Photo courtesy of Graham Johnson and Billie Johnson. - Click to enlarge

Half inch to 3 quarters of an inch hail fell in Raeford. Photo courtesy of Graham Johnson and Billie Johnson. - Click to enlarge

Hail that originally covered the ground was melting producing fog. Photo courtesy of Graham Johnson and Billie Johnson. - Click to enlarge

Warning Strategies

Segmentation of warning responsibility was done by line segments or clusters of storms, rather than specific areas. This worked well given the concept of polygon storm based warnings, however, if storms had not been so well organized (e.g. short-lived pulse storms), this method may not have worked.

The areal extent of the warning polygons was generous, especially later in the event, warning for entire counties due to the organization and coverage of the storms.

The storm coordinator position was crucial. At the height of the event there were 2 coordinators, one tracking warnings, SVS's and polygons working with the warnings forecasters and a second communicating with 911 centers, Skywarn and those issuing LSR's. The 2 coordinators also communicated making sure forecasters were away of severe weather reports and need for additional warnings and statements.

For the most intense supercells, enhanced wording was used in the SVS's to express the heightened threat (very large hail) with these storms.

The GRLevel2 and GRLevel3 software packages were utilized to monitor surrounding radars, and they were used to detect MARC signatures in one bowing segment. Street level road information was instrumental in communicating storm tracks and forecast impact areas to county warnings points.

Forecasters anticipated the event and several decisions early in the event resulted in a good utilization of resources. Initially, five of the southernmost counties were left out of the severe thunderstorm watch. Recognizing the speed at which the storms were moving and developing, and then including the entire CWA in the initial watch kept bookkeeping easier and kept us from having to use resources to issue another watch later. In addition, staffing was increased early in the event, a Skywarn controller was brought into the office, and AWIPS workstations were prepped for the active weather.

Including an N.C. State meteorology student as a part of the Joint Severe Operations was a key asset. The student helped with calls, logged storm reports, and assisted wherever needed.

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. NOAA IPW data provided by NOAA/GSD Ground-Based GPS Meteorology (GPS-MET) Real Time Water Vapor Interface. GOES satellite data was obtained from National Environmental Satellite, Data, and Information Service. SPC meso-analysis graphics provided by the Storm Prediction Center. Hail Swath product was obtained from the NSSL On-Demand website, and was displayed in Google Earth. CoCoRaHS precipitation and hail data provided by the CoCoRaHS web site. Photos courtesy of John Hamilton, David Stark, Graham Johnson, Billie Johnson, and emergency management officials in Anson and Chatham Counties.

Case study team -
Gail Hartfield
Brandon Vincent
Barrett Smith
Jeff Orrock
Darin Figurskey
Jonathan Blaes

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

NWS Disclaimer.