Colorado Firecamp - wildfire training wildland firefighter training Wildfire Blog Engine Boss Apprenticeship Location and Facility About Colorado Firecamp Frequently Asked Questions
Colorado Firecamp - wildland firefighter training

South Canyon Fire

William Teie Report to OSHA — 1994

6 Minutes for Safety — 2009

Fire Behavior Report, 1998

Cover & Dedication

Executive Summary & About the Authors

Preface & Contents


Fire Behavior Overview

Fire Environment

Fire Chronology

Fire Behavior Discussion



Appendix A

Appendix B

Appendix C

Report of the South Canyon Fire Accident Investigation Team, August 17, 1994

USFS shield logoFire Behavior Associated with the 1994 South Canyon Fire on Storm King Mountain, Colorado

Fire Environment


Long Term—The South Canyon Fire Investigation report states that “weather significantly contributed to the blowup of the fire.” Below normal precipitation (as compared to the 30 year average) during the winter and spring of 1994 had pushed western Colorado into a severe drought (South Canyon Report). Precipitation levels at Glenwood Springs from October 1, 1993, through July 6, 1994, were 58 percent of normal. Accompanying the below-normal precipitation were much warmer than normal temperatures through May and June. This pattern persisted into July. Fires in western Colorado during the previous weeks had exhibited rapid spread rates and long range spotting, both characteristic of drier than average conditions.

July 5, 1994—On this day, weak high pressure aloft and a hot, dry air mass covered western Colorado. Upper level winds were light from the southwest through 14,000 feet and then increased to 30 miles per hour at 16,000 feet. At Grand Junction, CO, the lower atmosphere was dry and unstable as indicated by a Haines Index of 6 (Haines 1988). That morning a strong cold front developed in western Idaho. The front was associated with an unseasonably cold upper level low pressure system, centered over northern Oregon. By late afternoon the cold front extended from eastern Idaho into central Nevada.

We are not aware of any weather observations taken by firefighters working on the fire. The closest remote automatic weather station (RAWS) was southwest of the airport at Rifle, CO, approximately 10 miles west of the fire site. We assume that the measurements from Rifle are representative of the conditions over the South Canyon Fire.

Between 0100 and 0600 on July 5 the relative humidity at the Rifle RAWS ranged from 10 to 40 percent. Relative humidity remained near 11 percent until after midnight (fig. 11). Winds at the Rifle RAWS were light and variable through the morning and late afternoon of July 5, 1994 (fig. 12). By 1800 on July 5, 1994, winds over the South Canyon Fire area were blowing generally from the south at 10 to 15 miles per hour with gusts to 20 miles per hour (South Canyon Report; Bell 1994).

July 6, 1994—At 0600 this day, the cold front extended from central Wyoming across northwest Colorado into southwestern Utah (fig. 13). Relative humidity at Rifle reached a high of 29 percent at about 0700 as compared to 40 percent the previous date (fig. 11). The lower relative humidity during the early hours of July 6 enhanced burning during the night. This is supported by witness statements that the fire was more active through the night on July 5 than on previous nights.

The front passed over Grand Junction around 1300 and caused the 10 to 15 mile per hour winds to increase to approximately 30 miles per hour (South Canyon Report). The front passed over the Rifle RAWS station between 1400 and 1500, as indicated by increased winds out of the west-southwest. They reached their peak at about 1600 with gusts to 45 miles per hour. The relative humidity at Rifle reached a minimum of 8 percent at 1700 on July 6. The winds at Rifle changed to the northwest at 1900. They gradually decreased through the evening and were blowing less than 6 miles per hour by 2300 (fig. 12).

Figure 11—Temperature and relative humidity measurements from the Rifle, CO, Remote Automatic Weather Station for period midnight on July 5, 1994, through 2300 on July 6, 1994.

Figure 12—Wind speeds from the Rifle, CO, Remote Automatic Weather Station for period midnight on July 5, 1994, through 2300 on July 6, 1994.

Figure 13—Surface air pressure charts showing progression of cold front on July 6, 1994 (taken from original Accident Investigation Report).

Firefighters on the fire and at Canyon Creek Estates, a residential subdivision about 2 miles west of the fire, noted an increase in winds between 1400 and 1430 (Scholz 1995). We surmise these were prefrontal winds associated with the approaching cold front. The Accident Investigation Team meteorologist estimated that the cold front passed over the South Canyon Fire at approximately 1520 (South Canyon Report). Witness statements characterized the winds as continuing to increase in speed, reaching their peak strength shortly after 1600. The original investigation report estimated southerly winds in the West Drainage of 20 to 35 miles per hour and westerly winds blowing across the ridges of 45 miles per hour at this time. Statements made by firefighters and the helicopter pilot suggest that it is likely that gusts over 50 miles per hour occurred in the chimneys and saddles along the ridges. Smokejumpers near the Lunch Spot at 1606 estimated upslope winds of 30 to 45 miles per hour (Petrilli 1995). Winds over the fire area remained strong until 2000 (South Canyon Report).

Topography and Wind—There is inconsistent testimony in witness statements regarding winds in the area of the fire. While the general flow was from the west, the direction and intensity of the winds varied significantly from one location to another (South Canyon Report; Shepard 1995). For example, firefighters working on the Main Ridge stated that as they cleared away brush from the fireline they could throw it into the air and the wind would carry it over the side of the ridge. At nearly the same time firefighters working on the northeast portion of the West Flank Fire Line (a location topographically sheltered from the wind) and at an exposed rocky outcropping near H-1 stated they experienced almost no wind at all. Though inconsistent with each other, these observations are consistent with the character of surface winds in complex mountainous terrain, particularly when influenced by strong, synoptic-scale weather patterns.

We feel topography was particularly influential on the surface winds the afternoon of July 6. West of the fire site the Colorado River flows through a relatively open and broad canyon. The low level westerly winds accompanying the cold front passed through this canyon. The canyon narrows dramatically about 2 miles west of the mouth of the West Drainage then opens to the east near Glenwood Springs (fig. 7 and 14). As the wind flowed from the broad canyon into the narrow Colorado River Gorge it increased in speed (fig. 15).

Several factors combined to create unusually strong upcanyon flow in the bottom of the West Drainage. First was the topography near the mouth of the West Drainage. The orientation of the ridge running from the ignition point, west to the Colorado River, acted as a “scoop” redirecting a portion of the westerly flow in the Colorado River Gorge northward up the bottom of the West Drainage (fig. 4 and 5). Second, normal daytime heating of the upper slopes in the West Drainage induced general upcanyon flow. This upcanyon flow resulted in a low pressure region at the entrance to the West Drainage, causing even more air to flow from the gorge into the drainage. Third, the increasingly strong westerly winds associated with the cold front evacuated air from the top of the West Drainage (fig. 16). This resulted in pressure-induced flow from the Colorado Gorge into the West Drainage. Some photographs and video footage taken during the blowup show smoke movement (Bell 1994) and clearly illustrate the complexity and turbulence of the surface winds shortly after 1610. High resolution meteorological modeling (see appendix C) supports the mesoscale wind analysis by the Accident Investigation Team meteorologist.

Figure 14—Aerial oblique photograph looking north over the Main Ridge and Storm King Mountain from the south side of the Colorado River. Note the widening of the river gorge as it opens into Glenwood Springs. While not shown in this picture, the Colorado River Gorge again widens to the west (off left side of photograph) of the fire site.
J. Kautz, U.S. Forest Service, Missoula, MT.

Figure 15—Schematic diagram of venturi effect that caused increased local winds where the West Drainage meets the Colorado River Gorge (not to scale).

Figure 16—The general air flow over the fire area at the time of the crown fire runs south of the Double Draws (about 1555). The free air westerlies (light blue) evacuate the afternoon upslope flow (green) out of the upper end of the West Drainage. This creates an area of divergence in the north end of the West Drainage. The topography and orientation of the mouth of the West Drainage redirected a portion of the strong winds in the Colorado River Gorge (dark blue) up the West Drainage (red). A shear zone is created where the westerly winds interact with the flow up the West Drainage. Turbulent eddies created by the shear zone may have enhanced burning near the Double Draw and along the Lunch Spot Ridge and West Bench. The eddies transported burning embers north up the West Drainage.
J. Kautz, U.S. Forest Service, Missoula, MT.

<<< continue reading—Fire Behavior at South Canyon Fire, Fire Chronology >>>


©2014 Colorado Firecamp, Inc. home scheduleblogENGBfacilityabout usFAQ's