Event Summary
     National Weather Service,
     Raleigh NC

March 07, 2004 High Wind Event
Updated 2004/04/28

Additional in-depth looks at this event (preliminary)
A look at the factors affecting the thermal profile in this case (updated 2004/04/28)
     Michael Brennan Michael Brennan NC State University

A look at the potential contribution of a stratospheric intrusion in this case (added 2004/04/07)
     Gail Hartfield gail.hartfield@noaa.gov NWS Raleigh


A strong cold front accompanied by a very strong upper level disturbance raced across North Carolina during the evening hours of Sunday March 7th. Although some thunderstorms accompanied the frontal passage, thunderstorms were not generally responsible for the damaging winds that moved across the state. The large swath of high winds between 45 and 65 MPH that moved across the state between 6:00 PM on Sunday, March 7 through about 12:00 midnight on Monday, March 8 appear to be from downburst winds enhanced by an unstable atmosphere.

Wind gusts in excess of 50 MPH were reported at more than two dozen official NWS/FAA reporting stations. There were countless reports of trees blown down and more limited reports of property damage. Local media reported that as many as 200,000 customers across North Carolina lost power during the storm.

     Figure 1 - Map of maximum winds (in MPH) from locations across North Carolina


Damaging winds in central North Carolina are typically the result of severe thunderstorms, tornadoes, or the inland effects from tropical cyclones. While there were reports of thunder and lighting on Sunday evening (3/7/04) associated with a very strong upper level system ( see Figure 2 ), NWS meteorologists observed that most of the damaging wind events in central North Carolina were not associated with the thunderstorms themselves. Rather the damaging winds were generally reported near the leading edge of rain advancing eastward on radar ( see radar loop ).

     Figure 2 - 00Z 03/08/04 500 MB Upper Air Plot
    Synoptic scale forcing supporting strong vertical ascent can been seen at 500 mb on 00z 03/08/04. Note the 125 kt wind and height falls of 16 decameters in the northwest flow over Nashville TN.

So if thunderstorms and strong dynamics aloft did not offer a complete accounting for the damaging winds, what might? A preliminary evaluation points to an unlikely answer for this region of the country - dry down bursts. The term refers to high winds associated with little, if any precipitation. Observations of winds and rainfall across central North Carolina indicates that high wind reports did indeed occur with little or no rainfall (e.g., 61 knots of wind with a trace of rain at Fayetteville). More compelling evidence for a dry down burst environment is shown in the GSO RAOB at 00Z on 3/8/04 ( see Figure 6 ), just prior to the 0034Z report at GSO of a 48 knot down burst wind accompanied by only 0.01 inches of rain.

     Figure 3 - 00Z 03/08/04 ETA Initialization at GSO from BUFKIT
    Shows environmental conditions favorable for a dry down burst wind. Note the steep temperature lapse rates (instability) in the sub cloud layer capped by a moist layer above (potential for strong down bursts initiated by rapid evaporation of precipitation).

Many of the RAOB sounding characteristics associated with dry down bursts can be seen on the initialized ETA forecast sounding at GSO valid at 00Z 3/8/04. Note the steep temperature lapse rate in the dry sub cloud layer ( see Figure 3 ). The sharp decrease in temperature from the surface to around 650 MB provides a great deal of instability to generate updrafts which in turn produce precipitation aloft. The spatial separation between the temperature (red line) and dew point (green line) indicates an unusual deep and dry sub cloud layer capped by a moist layer aloft. As precipitation falls into the sub cloud dry air, it is quickly evaporated and/or sublimated. This process cools the air producing a downward directed buoyancy with negative momentum. In addition, higher winds found aloft in the precipitation generating area are maintained as the down drafts descend toward the surface.

     Figure 4 - 00Z 03/07/04 GSO RAOB
    GSO RAOB from 00Z 03/07/04.

     Figure 5 - 12Z 03/07/04 GSO RAOB
    GSO RAOB from 00Z 03/07/04.

As previously noted, the characteristics of the afternoon sounding at GSO at 00z 3/8/04 ( see Figure 6 ). were very favorable for supporting dry down bursts winds. But early Sunday morning (12Z 3/7/04), meteorologists saw a striking different GSO sounding( see Figure 5 ). Note the quite stable and rather moist layer from the surface to around 800 mb and the warm inversion above centered near 700 mb. By referencing the GSO sounding observed on the previous afternoon (Saturday 00Z 3/7/04), ( see Figure 5 ), we see the inversion aloft was one associated with the passage of a dry cold front during the day on Saturday. With atmospheric mixing from daytime heating, we would expect to see a dry sub cloud adiabatic layer beneath the 700 mb inversion Sunday afternoon much as appeared on Saturday afternoon at 00z 3/8/04. Instead, the Sunday afternoon sounding shows a remarkably deep dry sub cloud adiabatic layer. The depth of the dry adiabatic layer so favorable for dry down bursts winds was largely produced by the strong dynamics (160 m height falls) and cold air advection associated with the vigorous upstream trough approaching North Carolina in the northwest flow aloft. Thus the upper dynamics were essential for producing the remarkably deep dry sub cloud adiabatic layer favorable for dry down burst winds.

     Figure 6 - 00Z 03/08/04 GSO RAOB
    GSO RAOB from 00Z 03/07/04, shows environmental conditions favorable for a dry down burst wind.

     Figure 7 - Wind data from the roof top at Jordan Hall on the campus of NC State University.

The cooling effects and wind gusts associated with the down burst gust front can be seen from the surface observations plotted at 02Z on 3/8/04 ( see Figure 9 ). Note temperatures are in the 40s accompanied by wind gusts in the 35 to 45 mph range on the cool side of the gust front.

     Figure 8 - Surface METAR plot from 00Z 03/08/04

     Figure 9 - Surface METAR plot from 02Z 03/08/04

More in-depth looks at this event (preliminary... more to come!)


Dr. Michael Kaplan (NC State University) for his insights into the evolution of the deep dry sub cloud adiabatic sounding.
Dr. Allan Riordan (NC State University) for supplying the wind data atop Jordan Hall on NCSU campus.

Case Study Team

Brandon Locklear
Michael Brennan
Gail Hartfield
Trisha Palmer
Kermit Keeter
Jonathan Blaes

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