# Encyclopedia of Wildfires and Wildland-Urban Interface (WUI) Fires

Living Edition
| Editors: Samuel L. Manzello

# Canyons

• D. X. Viegas
• J. Raposo
Living reference work entry
DOI: https://doi.org/10.1007/978-3-319-51727-8_118-1

## Definition

Canyons are terrain features composed by two slopes that intersect along an inclined line forming a gulley. Due to their concave shape, convective processes induced by the presence of fire in canyons are greatly enhanced, and their interaction with the fire can become extremely dangerous.

## Introduction

Forest fire spread in canyons is a subject of great practical and theoretical importance because it is associated with a large number of fatal accidents in the past, namely, to accidents with multiple fatalities. This is due to the rapid evolution of the fire spread in the canyon that surprises the persons that are involved in its extinction. Erroneously this change of fire behavior is to a modification of the ambient conditions, namely, to wind. Therefore the analysis of fire in canyons is important for the better understanding of the dynamic nature of forest fires and of the relevance of the fire induced convection.

A canyon is a terrain feature that is quite common in mountainous regions; it is characterized by the existence of at least the following elements: (i) a slope or face A that has an inclination θ1 in relation to the horizontal surface and (ii) a slope or face B that has an inclination θ2 in relation to the horizontal and intersects face A as a straight line that has an inclination angle α with the horizontal plane. In Fig. 1 a schematic view of a canyon is shown. In this figure the canyon faces A and B are represented as plane surfaces. Other elements of the canyon shown in Fig. 1 are the following:
• The dihedral angle γ between the two faces A and B

• The angle δ1 between face A and the reference axis OX

• The angle δ2 between face B and the reference axis OX

When θ12 we have a simple or symmetrical canyon; otherwise the canyon will be nonsymmetrical.

The intersection line between the faces is designated as “water line” because in natural canyons this line is usually the path of a water stream.

We may have composite canyons with more than two faces and several water lines. In nature the faces of the canyons are not plane surfaces; the curvature of each face has to be considered in such cases (Fig. 2).

Fire spread in canyons was studied systematically in laboratory conditions by Viegas and Pita (2004) at the Forest Fire Research Laboratory of the University of Coimbra, using a canyon table of 3 × 3 m2 with variable geometry. The simple canyon with symmetrical fire ignition case was studied systematically using infrared cameras to analyze the spread of the fire and to determine the local rate of spread (ROS) in relevant directions of the canyon. Later these studies were extended to a larger table 6 × 8 m2 using different fuel beds to determine the incubation time of different fuels. A large-scale field experiment in a canyon of more than 250 m of length and slope angles of the order of 30°, covered with shrub vegetation, was performed. The same behavior was observed at all scales for different fuels, confirming the assessment that the development of the fire is not dictated by the dimensions of the canyon but rather by its shape and geometry. The analysis of real fires at even larger scales reinforced this conclusion.

From this set of experiments and field observations, a mathematical semiempirical model to describe the variation of ROS of the head of the fire in a canyon in the course of time was developed (Viegas 2005). This model was later extended to several types of fuels (Viegas 2006).

## Properties of Fire Spread in Canyons

If the faces of the canyon are covered by flammable material such as natural vegetation (litter, herbaceous, shrubs, or trees) and if there is a fire ignition at some point inside the canyon, the fire will spread in a dynamic form reaching potentially very high values of the rate of spread (ROS) and of the fire line intensity that make this fire very difficult and dangerous to be suppressed.

The dynamic process of fire acceleration is produced by the strong convection induced by the presence of the fire that feeds the combustion zone with more oxygen from the ambient air. The concave shape of the canyon enhances this mechanism of air entrainment by the combustion zone and produces a gradual increase of the size of the flames and of the ROS of the head fire in the course of time. After a certain period of time that we designate as incubation time, the fire acceleration is very important, and the ROS may reach very high values making it impossible to run ahead of the head fire or even to attack it directly with standard equipment. The process of fire acceleration is designated as a fire eruption because it happens gradually and can produce a smoke plume that resembles that of volcano eruptions.

The duration of the incubation time depends on the type of fuel and of the steepness on the faces of the canyon (Fig. 3). For herbaceous fuels (HB) the incubation time is of the order of 1 min; for dead litter (LT) it is of the order of few minutes; and for shrubs (SR) the incubation time is between 10 and 30 min. These are just indicative values as they can be reduced when the values of δ1, δ2, and α are larger than 30°.

The process of fire eruption in a canyon may occur regardless of the existence of an ambient wind. If the wind is blowing along the water line and upslope, then the process of fire acceleration will be enhanced, and the incubation time will be shorter. If the wind blows downslope, the fire growth will be retarded, but its eruption may occur in the same way and eventually overwhelm the ambient wind (Viegas and Simeoni 2011; Liu and Xie 2013).

In some cases the sudden increase of the ROS of a fire in a canyon is mistakenly attributed to a change of ambient conditions, namely, to a change of wind velocity or direction. This may occur but it is not a necessary condition for a fire to erupt in a canyon.

The simpler case is when the ignition is at some point of the water line of a symmetrical canyon. If the ignition occurs in one face but not at the waterline, the fire will spread initially in that face, but when the back or lower flank of the fire reaches the waterline, it will start to spread on the other face, and the interaction of the two fires produces a very fast growth of the joint fire on both slopes.

## Fire Safety in Canyons

As the head of the fire spreads very rapidly and with an increasing ROS, a fire in a canyon should never be attacked on its head. The fireline intensity can easily overcome the fire suppression capacity of terrestrial forces, and with the increasing values of the ROS, it becomes very difficult to run faster than the fire spreading upslope. If it is not possible to suppress the fire in its initial stage, during the period of incubation, then it should be attacked only from the back or from the flanks.

Canyons are associated to many fatal accidents in the entire World. A large number of accidents with multiple fatalities in which groups of persons died occurred in canyons. We give some examples of accidents that were associated to fires in canyons.

### Mann Gulch

This accident occurred on the 5th of August 1949, and 13 elite firefighters lost their lives in a canyon above the Missouri River. A group of 15 smokejumpers were dropped on the top of the canyon to suppress a relatively small fire that had started by lightning on the ridge of the right side face of the canyon (Rothermel 1993). After landing they collected their equipment part of which was damaged. They were joined by a ranger that was already scouting the fire and started to walk along the water line to reach the burning area from below. In the meantime a set of firebrands fell on the bottom of the canyon and were not perceived by the fire fighters due to the curvature of the water line. After walking for some time, they noticed the fire that was starting to develop from these spots, below them, and decided to run away from it. The slope in which they were walking, trying to reach its ridge, was very steep, and although it did not have much vegetation, it was very difficult for the group of men to outrun the accelerating fire. One firefighter burned a plot of vegetation ahead of him and took refuge in this space and survived, but all the others, with the exception of two, were caught by the fire and perished.

### Storm King

This accident occurred on the 6th of July 1994 in Storm King Mountain, Glenwood Springs, Colorado, in the USA. Fourteen elite firefighters were killed in this accident that had many similarities with the Mann Gulch fire accident (Butler et al. 1998). A group of 49 firefighters were carried mainly by helicopter to the ridge above a large canyon over the Colorado River to attack a fire that was also ignited by lightning on the 2nd of July. On the afternoon of the 6th, a group of 16 firefighters were working on the descending fire front that seemed to spread without creating safety concerns. At around 16 h, a gust of wind promoted the projection of several firebrands to the bottom of the canyon below the position of the fire fighters. When they realized the presence of this threat, they decided to run away from the fire and tried to reach the ridge of the canyon. Once again they were walking on a very steep slope with a large distance to travel before reaching the ridge line. Only 3 were able to make it, but the other 13 were caught by the fire and died. Two members of a heli-tank group that were near the ridge line above the canyon were also trapped by the fire.

### Freixo

This accident occurred on the 7th of August 2003 in Freixo de Espada à Cinta in Northeast of Portugal. Two persons were killed in this accident (Viegas 2004). It started with a fire ignited by accident in a farm during agricultural works on a large slope overlooking the Douro River in the border between Portugal and Spain. Wind was blowing downslope, and the firefighters that came to attack this fire were managing to keep it contained near the bottom of the slope. The fire was almost controlled with the exception of a small section that was reaching the bottom of a canyon inserted in the main slope. When the fire reached this canyon, it erupted, and the intensity of the fire was so large that it burned in around 20 min the entire slope. Two persons that had gone to check their properties in the slope were at a place where they could not see the bottom of the canyon. They were caught by surprise and could not run away from the fire and were killed. A meteorological station that was located at the ridge of the canyon registered the wind velocity and direction during the entire process. It was found that the wind was blowing downslope all the time since the fire started, but when the fire erupted, it rotated 180° and started to blow upslope. The wind velocity reached 96 km/h and the air temperature reached 60° (10 min average), showing that even in a case in which the wind was blowing downslope, the power of the fire overwhelmed the ambient wind and produced the violent eruption in the canyon.

### Famalicão da Serra

This accident occurred in Portugal on the 9th July 2006, and it caused the death of six firefighters (Viegas et al. 2009). This group of firefighters were working with several others trying to stop a fire that was spreading on a face of a very large canyon. They were working on the upper flank of a wind-driven fire inside a pine stand, but when the head of the fire began to burn the crown of the trees, everyone had to run away from the head of the fire. This group went along a road toward the center of the canyon trying to reach the other face of the canyon and get away from the fire. When the back of the fire reached the water line, it started to spread with great intensity on the other face of the canyon and trapped the group of firefighters in the two fire fronts that were spreading in each face.

### Kornati

This accident occurred in 30th of August 2007 in the Island of Kornati in Croatia. Twelve firefighters were killed in this case (Stipanicev and Viegas 2009). Kornati Island is a small rocky island that is part of a Natural Park and has no permanent inhabitants being visited mainly by tourists. The terrain is very irregular and is covered by very scant vegetation that consists of herbaceous and some very small shrubs. A fire was started by carelessness near the shore of the island, but with the very strong wind that was blowing, it spread very rapidly along the very narrow land.

## References

1. Butler BW, Bartlette RA, Bradshaw LS, Cohen JD, Andrews PL, Putnam T, Mangan RJ (1998) Fire behaviour associated with the 1994 South Canyon Fire on Storm King Mountain, Colorado, Research Paper RMRS-RP-9. USDA Forest Service, Rocky Mountain Research Station, Fort Collins
2. Liu NA, Xie XD (2013) Extreme fire behaviors in large forest fires: current status and challenges. In: Bradley D, Makhviladze G, Molkov V, Sunderland P, Tamanini F (eds) Proceedings of the seventh international seminar on fire and explosion hazards. Research Publishing, Providence, pp 16–26.
3. Rothermel RC (1993) Mann gulch fire: a race that couldn’t be won. USDA Intermountain Research Station, Gen. Tech. Rep. INT-299Google Scholar
4. Stipanicev D, Viegas DX (2009) The accident of Kornati (Croatia). In: DX Viegas (ed), Recent forest fire related accidents in Europe. European Commission, Joint Research Centre, Institute for Environment and Sustainability, pp 26–53Google Scholar
5. Viegas DX (2004) Surrounded by fire. Editorial Minerva, Coimbra, 283 pp. (in Portuguese)Google Scholar
6. Viegas DX (2005) A mathematical model for forest fires blowup. Combust Sci Technol 177:27–51.
7. Viegas DX (2006) Parametric study of an eruptive fire behaviour model. Int J Wildland Fire 15:169–177.
8. Viegas DX, Pita LP (2004) Fire spread in canyons. Int J Wildland Fire 13:253–274.
9. Viegas DX, Simeoni A (2011) Eruptive behaviour of forest fires. Fire Technol 47:303–320.
10. Viegas DX, Pita L, Rossa C, Ribeiro L (2009) The accident of Famalicão (Portugal). In: DX Viegas (ed), Recent forest fire related accidents in Europe. European Commission, Joint Research Centre, Institute for Environment and Sustainability, pp 18–25Google Scholar