Event Summary
     National Weather Service, Raleigh NC

February 22, 2007 Fire Weather Event
Updated 2007/08/15

Event Headlines

...On February 22, 2007, the combination of strong westerly winds, temperatures soaring into the 70s, very low relative humidity's, and increasingly dry fuel moistures lead to a critical fire weather situation with adverse fire behavior over central North Carolina...

Event Overview

On February 22, 2007, the combination of strong westerly winds, temperatures soaring into the 70s, very low relative humidity's, and increasingly dry fuel moistures lead to a critical fire weather situation with adverse fire behavior over central North Carolina. There were dozens of fires reported across North Carolina on February 22, 2007 with at least two fires in central North Carolina that spread out of control and burned several residential structures.

A wildfire in Hoke County spread from a rural area into a residential neighborhood in Raeford and destroyed several homes. Another fire occurred in north Raleigh, in the middle of a sprawling urbanized area, and its initiation was attributed to “improperly disposed smoking materials” in some shrubbery. The fire quickly spread from the bushes into the town homes. It was so windy, warm, and dry that the fire jumped from one townhouse row to another.

The image below from the krax WSR-88D Doppler radar shows the smoke plumes from the larger fires across central North Carolina including the north Raleigh, Raeford and Harnett County fires.

krax WSR-88D Doppler radar imagery showing smoke plumes from the fires in north Raleigh, Raeford and Harnett County fires

Synoptic Overview

On Wednesday evening, February 21, 2007 an intense shortwave trough was located over the far northern Plains. Aided by a 130 kt jet at 250 MB, the shortwave had moved southeast and was located over the Great Lakes region at 12Z on Thursday, February 22. Winds behind the trough at 500 MB were in excess of 100 kts while a large area of winds in excess of 50 kts was noted behind the trough at 850 MB. The associated cold front reached the Ohio and Tennessee Valley by 12Z Thursday. The bulk of the clouds associated with the cold front were confined to the southern Great Lakes and Appalachians.

A dry surface cold front pushed from the
Ohio Valley region into the Appalachian Mountains during the morning hours, then surged off the Mountains through central North Carolina during the early afternoon. The westerly surface winds increased during the late morning hours and became strong and gusty by early afternoon. The increase in winds was due to the strong surface gradient and the deep mixing of the lower portions of the atmosphere up to 750-800 MB which allowed higher momentum air to reach the surface.

The strong westerly winds, aided by down slope subsidence drying and warming off the Appalachians, helped spread very warm and dry air eastward across central North Carolina during the afternoon. The time section plot at Raleigh-Durham shows the temperatures soared into the 70s, winds gusted in excess of 30 MPH, and relative humidity values crashed into the 8 to 15 percent range. A time section plot at the Fort Bragg RAWS station shows a similar pattern with a tremendous drop in relative humidity, warm temperatures, and wind gusts around 40 MPH.

At the time of the event, central North Carolina was experiencing abnormally dry conditions. During the 6 week period prior to the event, rainfall totals were averaging less than 50 percent of normal rainfall and no rain had been recorded in over a week (rainfall estimates from January 1-15, 2007  |   January 16-31, 2007  |   February 1-15, 2007). The 6 week period had also been significantly colder than normal with many subfreezing nights. This led to a significant drying of the fuel moisture levels.

Factors Contributing to Extreme Fire Behavior

Various studies have produced a list of weather factors that result in extreme fire behavior by influencing burning conditions at the time of fire initiation or that impact ongoing fires. The critical weather factors contributing to extreme fire behavior and large fire growth include:
  • Dew point Depression of 10 degrees C at the 850 MB level.
  • Temperature lapse rate of 7 degrees C at the 950 - 850 MB levels.
  • Position and/or location of frontal passages possessing strong winds, and of intense upper level troughs at 500 MB level.
  • Strong winds above the nocturnal inversion.
  • Low level jets just before and just after cold front passage.
  • Sea breeze fronts.
  • Pre-frontal and post-frontal (cold) passage associated with Byram (reverse wind speed and low level jet) & Brotak (surface at 9 kts and upper air at 34 kts wind profiles.
  • Haines Index of 4.

The Red Flag Program

Specific details about NWS Raleigh's fire weather program can be found in our Fire Weather Operating Plan. The Fire Weather Operating Plan specifically defines the critical weather parameters for collaboration of a Red Flag Warning. These parameters were agreed upon by the National Weather Service Offices that produce fire weather forecasts in NC, and the local user agencies, which include the U.S. Forest Service and the N.C. Forest Service. The critical weather parameters for collaboration of a Red Flag Warning include:
  • Sustained winds of 20 MPH or greater and/or gusts of 30 MPH or greater
  • Minimum relative humidity values of 25 percent or less.
When the weather parameters (as determined by the NWS) are forecast to reach the critical thresholds, the NWS contacts the N.C. Forest Service (NCFS). The NCFS is responsible for monitoring fuel moisture levels. When the fuel moisture values are expected to reach critically dry levels (as determined by the NCFS) and the critical weather parameters are expected to be met, a Red Flag Warning or a Fire Weather Watch is issued. The User Agency (NCFS) makes the final decision on whether or not to issue the Red Flag Warning or Fire Weather Watch after all coordination has taken place.

There are times in which the NCFS has preferred to have a Fire Danger Statement issued in lieu of a Red Flag Warning, even when the critical fire weather parameters are forecast to be met. This may be in part due to marginal fuel moisture levels, time of year, and other issues within the agency.

Forecast Process

During the midnight shift (February 21-22, 2007) NWS forecasters were still concerned with potentially strong winds the following day. The concern grew a few hours later when the analysis of the 12Z surface and upper air data revealed that fire weather parameters would become critical enough to reach collaboration thresholds for a Red Flag Warning.

NWS Raleigh issued a Special Weather Statement on the mid shift which highlighted strengthening winds up to 20 and 30 MPH with gusts to 40 MPH during the afternoon. Fire Danger was not initially a concern on the mid shift as the modeled relative humidity forecasts were forecast to remain above 30 percent. In fact, the low levels of the atmosphere were so moist that dense fog had developed. The midnight shift forecasters recognized the drying potential of the strong west winds and lowered the relative humidity forecasts. However, the relative humidity was expected to remain above the critical threshold of 25 percent. Winds were forecast to be at or above critical threshold for adverse fire concern.

As noted before, the increasing fire danger potential was recognized during the morning data analysis. The 50 kt 850 MB wind max oriented from northwest to southeast and extending from near Chicago to near Cincinnati was of most concern. This wind maximum was rapidly approaching the central Appalachians and was expected to cross the Mountains of North Carolina. and move over the Piedmont of N.C. during the afternoon. A dry cold front was approaching the western slopes of the Appalachians at this time. Winds were already gusting to between 35 and 50 MPH across the northern Mountains of NC at 12Z. It was expected that the 500 MB shortwave would swing north of central NC allowing the 850 MB wind max to surge across central North Carolina. This would allow significant wind from aloft to mix to the surface producing strong and gusty winds up to advisory criteria, especially across the Piedmont. The visible satellite image and surface map shown to the right (click on it to enlarge) is from 14Z on February 22 shows a band of clouds associated with the front crossing the southern Appalachians with west to northwest winds increasing in far western N.C. and TN.

The Wind Advisory and Red Flag Warning fire weather parameters were collaborated with neighboring NWS offices and with the NCFS. The Area Forecast Discussion contained critical information concerning the expected strong and gusty winds, warming temperatures, and drying relative humidity. A Wind Advisory was issued with the late morning update at 1129 AM for the Northwest Piedmont including Greensboro. The Special Weather Statement was reissued as a Fire Danger Statement with enhanced wording specifically addressing the critical fire weather parameters of strong wind, low relative humidity, and warm temperatures.

Even though the critical fire weather parameters were expected to be met, the NCFS requested that the NWS issue a Fire Danger Statement instead of a Red Flag Warning. This product was well publicized by the local media outlets, the fire community, and on NOAA Weather All Hazards Radio. The decision not to issue the RFW was due in part to the NCFS interpretation of the current and forecast 10 hour fuel moisture values. The observed 10 hour fuel moisture values from February 21, 2007 were only marginally close to the critical threshold values for concern. The image to the left (click on it to enlarge) depicts the observed 10 hour fuel moisture values for February 21, 2007. At the time of the coordination, the most recent 10 hour fuel moisture values were from the previous day. Note that observations reported 10 hour fuel moisture values of 11-15 percent. Only Duke Forest and Caswell Game Land reported values that were deemed dry enough to be of concern, 7-9 percent respectively. Ten hour fuel moisture values of 6 percent or less are generally considered extremely critical.

As the winds increased, relative humidity values continued to crash (click on the relative humidity plot for Duke Forest, shown to the right to enlarge), and temperatures warmed from late morning into the early afternoon, additional coordinating with the NCFS led to the issuance of a Red Flag Warning for all of central North Carolina. This RFW was issued at 220 PM. The peak winds reached between 40 and 49 mph throughout most of central North Carolina. This combined with temperatures reaching between 72 and 75 along with a humidity that crashed to between 8 and 12 percent which ultimately contributed to several devastating wild fires in central North Carolina during the afternoon of February 22, 2007.

Remotely Sensing Fires

Remote sensing instruments, specifically satellite and radar, provide imagery which can be quite useful for fire detection. Smoke plumes from fires can be viewed with weather surveillance radars and visible channel satellite imagery while the shortwave infrared (3.9 micron) satellite channel can literally sense heat associated with fires.

Particles from a smoke plume associated with a fire in North Raleigh can be seen in the radar imagery below. The images are from the krax WSR-88D Doppler radar at 2118Z on February 22, 2007. The radar was able to pick up on the higher reflectivity associated with the smoke plume from the fire.

The image below is a base reflectivity product at 1.4 degrees from the 2118Z volume scan. By using the 1.4 degree slice, the image excludes a majority of the ground clutter while still showing the reflectivity associated with the smoke plume. You can see the origin of the smoke plume just north of the Raleigh beltline.

The image below is a 3-D cross section with a best fit along the smoke plume from the 2118Z volume scan. The image is oriented such that the viewer is looking northwest from a position aloft just southeast of Smithfield. The cross section provides the viewer with some view of the vertical extent of the smoke plume.

The image below is a 3-D cross section from the 2118Z volume scan with the same axis as the image above. The image is zoomed in somewhat and the viewing angle has been lowered so that the plume is being viewed from near the surface.

Meteorologists often use longwave (10.7 micron) infrared satellite imagery in weather forecasting. But the properties of the 3.9 micron channel, however, make it valuable for detecting hot spots associated with fires. Blackbody radiance in the 3.9 micron channel increases more rapidly with temperature than the radiance in the 10.7 micron channel. Therefore, the 3.9 micron channel is more sensitive to subpixel hot spots than the 10.7 micron channel, and is resultantly better suited for fire detection.

The image below is the 3.9 micron channel satellite imagery from 2031Z over central North Carolina. Two dark colored pixels can be seen in the imagery. They indicate pixel temperatures of 25 deg C or more. The dark colored areas correspond very well to the location of fires at the time of the satellite image.

A web page has been developed that allows visitors to interactively view data from the trained satellite analysts in the Satellite Analysis Branch (SAB), within the Satellite Services Division (SSD), to manually integrate data from various automated fire detection algorithms with GOES and polar (Advanced Very High Resolution Radiometer (AVHRR), Moderate Resolution Imaging Spectroradiometer Fire Algorithm (MODIS) and Defense Meteorological Satellite Program/Operational Linescan System (DMSP/OLS)) images. The result is a quality controlled display of the locations of fires and significant smoke plumes detected by meteorological satellites.

You can view this data at... http://www.firedetect.noaa.gov/

Archived Text Data from the Fire 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 EST time + 5 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
RDULSRRAH - Local Storm Reports
RDUSPSRAH - Special Weather Statement
RDUSPSRAH - Fire Danger Statement
RDUFWFRAH - Fire Weather Forecast
RDUFWMRAH - Fire Weather Forecast


Lessons Learned

  • Situational awareness (SA) is the key to successful Warnings, Watches, Advisories, and forecasts. It is imperative that surface and upper air data be thoroughly analyzed and compared to model initializations and forecasts. A comprehensive SA program, developed over many years, has been in place at WFO Raleigh. This includes hand analysis of surface and upper air data, comparison of this data analysis to the latest model initialization and forecasts, and pattern recognition of critical severe weather patterns.
  • Had the day shift not detected among many key features that typically lead to the critical fire weather concerns in central North Carolina, including the 50 kt wind maximum diving southeast over IL pointing toward NC in the 12Z/22 February 2007 850 MB analysis, the dry surface cold front coming across the mountains, and the increasing wind gusts up to 45 kts in the northern Mountains, the Wind Advisory, enhanced Fire Danger Statement, and subsequent Red Flag Warnings may have been missed.
  • A Red Flag Warning could have been issued earlier had the main user agency (in our case the NCFS) given us the go ahead. This would have increased the lead time of the warning and possibly helped the local and state fire and emergency officials have a better knowledge of the potential adverse fire behavior situation. However, the SPS (Fire Danger Statement) has been well received by the fire user agencies.
  • In the future, the NWS should consider making the Red Flag Warning a more visible product...at least equal to the SPS (Fire Danger Statement).
  • There are several remote sensing tools that can be useful in observing and locating fires.

Case Study Team

Phil Badgett
Michael Strickler
Jonathan Blaes

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

  • NWS Disclaimer.