Embed Size (px)
Citation preview
Fire Behaviour in Black Spruce Forest Fuels following Mulch Fuel treatments:
A case study at Red Earth Creek, Alberta
September 2016 – Technical report no. 42
Steven Hvenegaard, FPInnovations, Wildfire Operations
Dave Schroeder, Alberta Agriculture and Forestry
Dan Thompson, Natural Resources Canada, Canadian Forest Service
First Name, Name, Title, Titre, Program Name
First Name, Name, Title, Titre, Program Name
First Name, Name, Title, Titre, Program Name
fpinnovations.ca
FPInnovations is a not-for-profit world-
leading R&D institute that specializes in
the creation of scientific solutions in
support of the Canadian forest sector’s
global competitiveness and responds to
the priority needs of its industry members
and government partners. It is ideally
positioned to perform research, innovate,
and deliver state-of-the-art solutions for
every area of the sector’s value chain,
from forest operations to consumer and
industrial products. FPInnovations’ staff
numbers more than 525. Its R&D
laboratories are located in Québec City,
Montréal and Vancouver, and it has
technology transfer offices across
Canada. For more information
about FPInnovations, visit:
www.fpinnovations.ca.
Follow us on:
301010385: Effectiveness of mulching as
a forest fuel treatment.
Technical Report no. 42
REVIEWERS
Jen Beverly, Faculty Service Officer
Professional education, Wildland fire,
Department of Renewable Resources,
University of Alberta
Kelsy Gibos, Wildfire Management Specialist
Edson Wildfire Management Area
Alberta Agriculture and Forestry
CONTACT
Steven Hvenegaard
Wildfire Operations
780-740-3310
© 2016 FPInnovations. All rights reserved. Unauthorized copying or redistribution prohibited.
Disclosure for Commercial Application: If you require assistance to implement these research findings, please contact FPInnovations at [email protected].
FPInnovations Page 3
Table of contents
ABSTRACT ........................................................................................................................................... 5
INTRODUCTION ................................................................................................................................... 6
STUDY SITE ......................................................................................................................................... 7
Mulch Treatment Techniques ............................................................................................................. 7
Fuel Environment ............................................................................................................................... 8
Fuel Moisture ................................................................................................................................... 11
Historical Weather ............................................................................................................................ 11
METHODS ........................................................................................................................................... 12
Weather Data Collection .................................................................................................................. 12
Fire Behaviour Data Collection and Processing ................................................................................ 12
RESULTS ............................................................................................................................................ 13
Ignition ............................................................................................................................................. 13
Weather ........................................................................................................................................... 13
Fire Behaviour .................................................................................................................................. 13
Fire behaviour in the strip mulch treatment ................................................................................... 16
Fire behaviour in the mulch thinning treatment ............................................................................. 17
Ember transfer and spot fire development .................................................................................... 18
DISCUSSION ...................................................................................................................................... 20
Fireline Intensity ............................................................................................................................... 20
Effects of strip mulching treatment ................................................................................................... 21
Effects of mulch thinned treatment ................................................................................................... 21
Comparison with other black spruce wildfire/fuel management interactions...................................... 22
Fuel Management Implications ......................................................................................................... 23
Enhanced suppression ..................................................................................................................... 24
CONCLUSION ..................................................................................................................................... 25
NEXT STEPS ...................................................................................................................................... 26
ACKNOWLEDGEMENTS .................................................................................................................... 26
REFERENCES .................................................................................................................................... 27
FPInnovations Page 4
List of figures
Figure 1. Mulch fuel treatments under study at Red Earth Creek FireSmart research area. ................... 7
Figure 2. Block 1 treatment area with camera locations and initial ignition line shown. .......................... 8
Figure 3. Mulch thinning treatment (left) and strip mulch treatment (right). ............................................. 8
Figure 4. Sparse regrowth and mulched fuel components in the surface fuel layer. ............................. 11
Figure 5. Estimated fire progression (time after ignition) and spread rates through the treated stands. 14
Figure 6. ‘Developing’ active crown fire in the natural stand one minute after ignition. ......................... 15
Figure 7. Fire breaching mulch strips with spot fire development in advance of fire front. .................... 16
Figure 8. Intermittent crown fire in the strip mulch treatment. ............................................................... 17
Figure 9. Relatively low fire intensity along the west flank of the thinned treatment. ............................. 18
Figure 10. Spot fire initiation along the north boundary in the early stages of fire development. ........... 19
Figure 11. Vertical fire spread along residual stems with standard pruning height. .............................. 20
Figure 12. Influence of high intensity fire in strip mulch on fire behaviour in mulch thinned plot. .......... 21
Figure 13. Reduced depth of burn and fuel consumption in hummocks of sphagnum moss................. 23
List of tables
Table 1. Pre-treatment and post-treatment stand inventory (standard deviation in parentheses). .......... 9
Table 2. Dead woody debris and mulch loading in treatment areas by activity phase. ......................... 10
Table 3. Composition (%) of ground fuel components for each fuel stand. ........................................... 10
Table 4. Fuel moisture content (%) ...................................................................................................... 11
Table 5. Total fuel consumption (kg/m2) based on mean values. ......................................................... 12
Table 6. Weather data and hourlya FWI values during fire passage through the treated area. ............. 13
Table 7. Forecast FBP values for C-2 (Boreal Spruce) fuel type .......................................................... 15
Table 8. Calculated head fire intensity for varying spread rates at different times of fire passage. ....... 16
Table 9. Comparison of fire behaviour resulting in black spruce stands treated through fuel removal. . 22
FPInnovations Page 5
ABSTRACT
Mulching of forest fuels is a vegetation management practice commonly applied in the wildland urban
interface to mitigate the risk of wildfire. This experimental fire at the Red Earth Creek FireSmart
research area was designed and conducted to document how two different mulch fuel treatments
modified fire behaviour when challenged by an approaching crown fire.
The two fuel treatments applied at the Red Earth Creek study site included mulch thinning and strip
mulching. We ignited a strip of natural black spruce forest upwind of these treatments to observe and
document changes in fire behaviour as crown fire encountered the mulch treatments.
The mulch thinning treatment modified crown fire to high-intensity surface fire. With a greater volume of
crown fuels available, the strip mulch treatment maintained intermittent crown fire behaviour with
crown-to-crown fire spread across mulched strips. In both treatments, fine chipped debris and
feathermoss in the surface layer and aerial lichen in standing stems provided an abundance of ignition
receptors for spot fire propagation. With varying wind speeds throughout the experimental fire, the rate
of spread in the treated areas ranged between 14 and 29 m/min. Calculated head fire intensity with
wind variations ranged between 13,440 and 27,840 kW/m.
The Red Earth Creek experimental fire was conducted under conditions of severe fuel moisture deficit
with temperature/relative humidity crossover and moderate wind conditions. The hourly ISI (15.7) at the
time of ignition was at the 98th percentile. The historical 90th percentile BUI value was 64 while the BUI
on May 14 (50.9) was at the 83rd percentile. These factors contributed to volatile fire behaviour in the
ignition zone in the natural stand, which tested the potential limits of a fuel treatment of this type. This
case study of fire behaviour in mulched fuel treatments at extreme fuel and weather conditions does not
address the effectiveness of fuel treatments in other fuel types under different weather conditions.
FPInnovations Page 6
INTRODUCTION
Forest fuels engineering is one of the primary wildfire mitigation strategies advocated by FireSmart™
Canada1 and applied by partnering wildfire management agencies and industry operators. Fuel
treatments have been extensively applied in and around communities in the wildland-urban interface,
through a broad range of fuel modification techniques. A primary objective of fuel treatments is to
modify fire behaviour to a ‘less difficult, disruptive, and destructive’ state (Reinhardt et al. 2008) which
can allow for safer, more effective fire suppression operations (Moghaddas and Craggs 2007).
Black spruce is one of the most prevalent fuel types surrounding communities in central and northern
Alberta, as well as other parts of boreal Canada. The densely stocked black spruce forest stands in the
Red Earth Creek FireSmart research area exhibit typical crown fuel properties of black spruce: high
crown bulk density and low crown base height, which contribute to crown fire initiation (Van Wagner
1977). These fuel characteristics, combined with low fuel moisture contents and strong winds, create
ideal conditions for high-intensity, rapidly-spreading catastrophic wildfire (Flat Top Complex Wildfire
Review Committee 2012).
Mulch fuel treatments use various types of equipment to masticate forest vegetation resulting in a
reduction in crown bulk density and the conversion of canopy and ladder fuels to a more compacted
and less available fuel source in the surface layer (Battaglia et al. 2010). Mulch thinning and strip mulch
treatments create a more open surface fuel environment with both negative and positive impacts. Due
to increased exposure to sun and wind flow, the chipped debris and other surface fuels in the open
areas of the treatments dry more quickly than fine fuels in enclosed stands (Schiks and Wotton 2015).
From a control perspective, the open thinned areas of the treatments allow more effective penetration
of water/suppressant through canopy fuels to surface fuels (Hsieh in progress). Additionally, fine fuels
at the surface of openings respond more quickly to water and suppressant application. Open areas of
the treatments that have been wetted by sprinkler systems or aerial water delivery should reduce the
potential for ignition and sustained burning, providing a potential barrier to fire spread.
Experimental crown fires have been conducted to challenge fuels treatments in other forest fuel types
(Schroeder 2010, Mooney 2013) to evaluate the efficacy of these treatments in moderating fire
behaviour. Mechanical (shearblading) fuel treatments in black spruce fuels (Butler et al. 2013) have
been shown to reduce fire intensity. However, documentation of crown fire challenging mulch fuel
treatments in black spruce fuels is limited. Fire and fuels managers would like to evaluate the
effectiveness of mulch fuel treatments in reducing fire intensity and rate of spread and, ultimately, their
ability to mitigate wildfire risk to communities surrounding these hazardous fuels.
Alberta Agriculture and Forestry (AAF) Wildfire Management Branch fuels managers designed the Red
Earth Creek FireSmart research area with the objective of conducting research that will lead to a better
understanding of mulch fuel treatments and how these changes in the black spruce fuel environment
affect fire behaviour. On May 14, 2015, Slave Lake Forest Area personnel conducted an experimental
fire at this site; FPInnovations and research partners collected data to document changes in fire
behaviour.
1 https://www.firesmartcanada.ca/
FPInnovations Page 7
STUDY SITE
The town of Red Earth Creek is 167 km north of Slave Lake, Alberta along Highway 88. The Red Earth
Creek FireSmart research area is located 5.5 km east of Red Earth Creek from Highway 88. The
research area was designed to treat seven separate blocks to be used for experimental fires. Block 1
(Figure 1) was treated in the winter of 2013/14. The fuel treatments were located and oriented directly
adjacent to natural fuel stands capable of sustaining active crown fire that would challenge the
treatment area. In central Alberta, southeast winds are frequently experienced during periods of high
fire hazard in the spring season.
Figure 1. Mulch fuel treatments under study at Red Earth Creek FireSmart research area.
Mulch Treatment Techniques The mulch treatments applied at the research area are fuel treatment techniques2 commonly applied in
the Slave Lake Forest Area. Block 1 (2.36 ha) was treated using a strip mulch treatment (0.65 ha) in the
east half and a mulch thinning treatment (0.71 ha) in the west half (Figure 2). The natural stand of black
spruce southeast of the treated area was 1.35 ha.
2 http://wildfire.fpinnovations.ca/119/Slave%20Lake%20Summary_v5.pdf
FPInnovations Page 8
Figure 2. Block 1 treatment area with camera locations and initial ignition line shown.
Fuel Environment The overstory species in the treated and natural stands was 100% black spruce. The mulch thinning
treatment resulted in stem spacing of 5 to 7 m with a reduction in stand density from 1,400 stems/ha to
400 stems/ha (stems greater than 9 cm DBH). The strip mulching treatment resulted in mulch strips and
strips of standing stems (residual strips) approximately 4 m wide (Figure 3). The stem density in this
treatment area was reduced from 1,450 stems/ha to 650 stems/ha.
Figure 3. Mulch thinning treatment (left) and strip mulch treatment (right).
Courte
sy o
f Alb
erta
Agric
ultu
re a
nd F
ore
stry
26
25
27
23
24 22
28 28
FPInnovations Page 9
The Alberta Wildland Fuels Inventory Program (AWFIP) crews conducted fuel sampling activities to
measure pre-treatment and post-fire fuel inventories. Due to time constraints, a thorough post-
treatment sampling could not be conducted according to AWFIP standard procedures. Post-treatment
canopy fuel load and canopy bulk density were estimated by using pre-treatment mean height and
lower crown base height, in conjunction with post-burn measurements of stand density and diameter at
breast height (DBH). AWFIP personnel processed inventory data to determine changes in stand density
and canopy fuel loading (Table 1).
Table 1. Pre-treatment and post-treatment stand inventory (standard deviation in parentheses).
Pre-treatment inventory
Density (stems/ha) Mean Height
(m) Live Crown
Base Height (m) Canopy
1 Fuel
Load (kg/m2)
Canopy Bulk Density (kg/m
3)
Overall Treatment
Area (Block 1)
Overstory2 1,260 (550) 10.3 (1.6) 4.1 (1.1) 1.2 (0.4) 0.2 (0.1)
Understory3
4,000 (2,200)
Post-treatment inventory
Mulch Thinning
4004
10.3 (1.6)
4.1 (1.1) 0.3 (0.1) 5 0.05 (0.02)
5
Strip Mulching
650 10.3 (1.6) 4.1 (1.1) Residual strips = same as pre-
treatment; mulch strips = 0
6
Notes:
1. Canopy fuels includes needles and live twigs <0.5 cm diameter.
2. Overstory stems are those greater than 9 cm. in diameter.
3. Understory stems are those less than 9 cm in diameter and greater than 1.3 metres in height.
4. All understory stems were processed in mulch thinning treatment.
5. Post-treatment estimate is based on pre-treatment mean height and LCBH and post-treatment DBH and density.
6. Post-treatment canopy fuel load in the residual strips was unchanged from pre-treatment value; overall canopy fuel
loading in the strip mulch treatment area was reduced by 44%.
The intensity of these treatments is also reflected in the resultant post-treatment dead woody debris
loading. Pre-burn sampling using line transects in the treated areas yielded post-treatment debris
loading values. Destructive sampling of the mulch fuels yielded an average mulch fuel loading of 2.67
kg/m2. Mulch debris was distributed unevenly throughout the treatment areas, with the depth averaged
at 2.3 and 2.1 cm deep for the strip mulch and mulch thinning treatments, respectively. Table 2
indicates post-treatment and post-fire changes in surface fuel loading and surface fuel consumption.
FPInnovations Page 10
Table 2. Dead woody debris and mulch loading in treatment areas by activity phase.
Phase
Weight (kg/m2)
Size class (SC in cm)
SC1
(0 – 0.49)
SC2
(0.50 – 0.99)
SC3
(1.00 – 2.99)
SC4
(3.00 – 4.99)
SC5
(5.00 – 6.99)
Coarse
Woody
Debris
(>7.0 cm)
Mulch Sum of mean values
Litter Duff
Pre-treatment
0.0032 (0.0016)
0.0032 (0.002)
0.0065 (0.0051)
0.011 (0.0085)
0.0212 (0.0181)
0.7 (0.3)
NA 0.75 0.54
(0.21) 10.14 (6.52)
Post-treatment
0.0059 (0.0035)
0.0054 (0.0041)
0.017 (0.0106)
0.0111 (0.0097)
0.0147 (0.0148)
Thin 1.1
(0.4) 2.67 (3.34)
3.82 Strip
1.1 (0.3)
Post-Fire
a 0.0005
(0.0004) 0.0011
(0.0006) 0.0073
(0.0029) 0.0304
(0.0159) 0.048 (0.03)
Thin 0.6
(0.4) 0.23
(0.23) 0.92 b
10.11 (2.86)
Strip 1.3
(0.3) 0.23
(0.23) 1.67 b
8.72 (2.28)
Fuel consumed Thin 2.9 b 0.03
Strip 2.15 b 1.42 aDead woody debris in some size classes increased post-fire due to fallen trees/broken branches
bLitter was included with mulch for post-treatment and post-fire samples.
In both treatments, there was little mixing of the duff layer and mulch debris. The predominant ground
fuel component was feathermoss with lesser amounts of Sphagnum fuscum and litter (Table 3).
Table 3. Composition (%) of ground fuel components for each fuel stand.
Fuel Component Natural Stand Mulch Thinning Strip mulching
Dead Feathermoss 0 65 8
Dead Sphagnum.
fuscum 0 27 0
Live Feathermoss 77 6 87
Live S. fuscum 21 2 5
Litter 2 0 0
Courtesy of McMaster University Ecohydrology Group
There was a minimal amount of vegetative regrowth in the treated area, with short Labrador tea stems
resprouted sparsely amongst the moss and chipped debris surface layer (Figure 4).
FPInnovations Page 11
Figure 4. Sparse regrowth and mulched fuel components in the surface fuel layer.
Fuel Moisture Fuel moisture samples were collected three hours prior to ignition, and wet weighed on-site. Moisture
content of fuel classes is shown in Table 4.
Table 4. Fuel moisture content (%)
Fuel type Gravimetric moisture content
Litter/fine woody debris 7.7%
Mulch 7.5%
Conifer foliage 82%
Historical Weather Using historical weather data (1981 to present) for the Red Earth lookout (LO) weather station (15 km
to the north), we determined that the 90th percentile ISI value was 8.5. At the time of ignition, the hourly
ISI (15.7) was at the 98th percentile. The historical 90th percentile BUI value was 64 while the BUI on
May 14 (50.9) was at the 83rd percentile.
Photo
court
esy o
f M
cM
aste
r E
cohydro
logy G
roup
FPInnovations Page 12
METHODS
Weather Data Collection On-site temperature and relative humidity were recorded at 1300h on May 14 using a Vaisala handheld
sensor. Over the duration of the fire passage through the treatment, we recorded and averaged the
wind speed from a portable weather station 0.6 km south of Block 1. We applied these weather values
and May 13 baseline FWI values from Red Earth Creek LO weather station in REDapp3 to calculate
hourly FFMC, ISI and FWI values at the time of ignition. Precipitation amounts registered at Red Earth
LO in the week prior to the experimental fire were compared to data from the Environment Canada
weather station at Red Earth Creek airport (4 km west of the research site) to ensure FWI values from
the Red Earth LO were representative of the experimental fire site.
Fire Behaviour Data Collection and Processing We placed 20 data loggers in a grid pattern throughout the two treatment areas to record temperature
and time of fire passage. The data loggers were placed 20 m apart along five lines parallel to the
anticipated wind direction. Data from the dataloggers was downloaded and processed to calculate rate
of spread along each of the five lines at four different intervals.
We input rate of spread and fuel consumption data in Byram’s (1959) equation for fire intensity
adapted for use in the Canadian Forest Fire Behavior Prediction System (Forestry Canada Fire Danger
Group 1992) as:
FI = Hwr
Where:
FI = fire intensity (kW/m),
H = fuel low heat of consumption (kJ/kg); 18,600 kJ/kg is the accepted value for fuel low heat of
combustion when rate of spread is input as meters/second. In order to use metres/min for rate of
spread, this value is converted to 300 kJ/kg.
w = weight of fuel consumed during the active flaming zone (kg/m2) and
r = rate of spread (m/min).
We input fuel consumption values (Table 5) to this formula to calculate head fire intensity for the two
treatments. Duff was not included in the head fire fuel consumption, as it was assumed to burn primarily
after the head fire passage.
Table 5. Total fuel consumption (kg/m2) based on mean values.
Treatment type Surface Canopy Total
Mulch thinning 2.90 0.30 3.20
Strip mulching 2.15 1.20 3.35
3 REDapp is a fire management decision support tool. http://redapp.org/
FPInnovations Page 13
We captured video from six in-fire cameras located in the fire area, and one unprotected camera
mounted on an elevated platform near the fire (Figure 2). A videographer filmed the fire from an aerial
platform in the operations helicopter. Additionally, we captured video with a remote-controlled camera
mounted on an 80 foot portable tower 250 m north of Block 1. We used video footage to analyze spot
fire initiation and fire spread rate and intensity through the treatment areas.
RESULTS
Ignition Ignition was achieved using a helitorch equipped with an aerial-ignition tracking system4 to record the
GPS track and timing of ignition patterns. The initial ignition line was set in the natural fuel stand of
black spruce approximately 5 to 8 m back from the treated areas (Figure 2). The ignition line was
approximately 70 m and was centred at the boundary between the thinned and strip treatments. The
torch tracker data logger recorded the start time for the initial ignition strip as 14:17:14 and stop time as
14:18:08. At 14:19, two shorter secondary ignition strips were lit in the natural fuel stand behind the
initial ignition line.
Weather During the ignition operations--and as the fire entered the treated area--the wind speed average was 14
km/h with gusts up to 22 km/h. Wind speed temporarily increased to 19 km/h with fluctuation of 3-6
km/h due to gusts during the fire passage through the treated stands. A maximum wind speed of 25.6
km/h was recorded as the fire progressed through the treated stands. Table 6 shows how the FWI
values changed in response to the increase in wind speed.
Table 6. Weather data and hourlya FWI values during fire passage through the treated area.
Temperature (Celsius)
Relative Humidity
(%)
Wind Speed (km/h)
Wind Direction (degrees)
Hourly FFMC
Hourly ISI
DMC DC BUI Hourly
FWI
25 15 14 146 94.2 15.7 32.8 281.9 50.9 30.6
25 15 19 146 94.2 18.8 32.8 281.9 50.9 34.6
a Adjusted according to Lawson et al.(1996)
Fire Behaviour Thirty seconds after ignition was initiated, surface fire in the ignition zone spread to the crown fuels.
The crown fuels became fully involved and fire spread through the natural black spruce fuels to the
treated areas within 1 minute and 25 seconds of ignition start. The general SSE wind directed the main
head of fire along the boundary between the two treatment plots (Figure 5). This boundary was the
central axis of the elliptical fire growth pattern.
4 http://wildfire.fpinnovations.ca/140/ProjectReportDesigningAerialIgnitionTrackingSystem_vFINAL.pdf
FPInnovations Page 14
The head fire rate of spread along its central axis ranged from 14 to 29 m/min (Figure 5). An
intermittent crown fire occurred throughout the strip mulch and an intense surface fire passed through
the mulch thinning treatment in which all of the standing trees candled. Ember transfer was prolific with
many spot fires developing in advance of the head fire, both on the surface and in the canopy.
The head of the fire crossed the mulch treatment perimeter 4 minutes and 50 seconds after ignition
start. The fire travelled through the treated area in 3 minutes and 48 seconds at an overall ROS of 21.7
m/min. A combination of flank and backing fires occurred in both the thinned and strip mulch treatments
with average rates of spread shown in Figure 5. Eventually the entire treatment area was burned over.
Figure 5. Estimated fire progression (time after ignition) and spread rates through the treated stands.
Due to the short span of fuels between the ignition line and the treated blocks, fire in the natural stand
did not achieve steady state fire intensity or equilibrium rate of spread (Figure 6). The fire behaviour
was therefore classified as a ‘developing’ active crown phase (Alexander and deGroot 1988). This
ignition zone is not considered a ‘control’ where observations and documentation of fire behaviour
could be considered representative of outputs indicated by the Fire Behaviour Prediction (FBP) system
(Forestry Canada Fire Danger Group 1992). Table 7 shows the FBP fire behaviour outputs calculated
for the black spruce fuels (FBP C-2 fuel type) at both observed wind speeds from Table 6.
FPInnovations Page 15
Figure 6. ‘Developing’ active crown fire in the natural stand one minute after ignition.
Table 7. Forecast FBP values for C-2 (Boreal Spruce) fuel type
Weather Station
Wind Speed (km/h)
Rate of Spread (m/min)
Surface Fuel Consumption
(kg/m2)
Head Fire Intensity using standard FBP
crown fuel load
1 (kW/m)
Head Fire Intensity using local crown fuel
load2 (kW/m)
Fire Type
Red Earth LO with on-site
1300 readings 14 23.5 2.22 21,256 24,076
Continuous crown fire
On-site wind speed change
19 31.5 2.22 28,470 32,250 Continuous crown fire
Notes: 1 0.80 kg/m
2 as stated in Forestry Canada Fire Danger Group (1992)
2 1.2 kg/m
2 (Table 1)
Fireline intensity calculations, based on Byram’s equation, for the varying rates of spread throughout
the fire are shown in Table 8. The values used for weight of fuel consumed for the strip mulching and
mulch thinning treatments are 3.35 and 3.2 kg/m2, respectively. The head fire intensity values
calculated in Table 8 are based on complete fuel consumption in the active flaming zone during head
fire passage. It is unlikely that there was 100% consumption of surface fuel in the active flame front; this
concept is proposed in the Fireline Intensity portion of the Discussion section.
FPInnovations Page 16
Table 8. Calculated head fire intensity for varying spread rates at different times of fire passage.
Rate of Spread
(m/min)
Head Fire Intensity (kW/m)
Strip Mulching
Mulch Thinning
Head fire
14 14,070 13,440
19 19,095 18,240
29 29,145 27,840
20 20,100 19,200
22
(overall ROS) 22,110 21,120
Flank fire 2.9 2,784
4.4 4,422
Fire behaviour in the strip mulch treatment
Ember transfer, spot fire growth and crown to crown fire spread all contributed to fire growth in the strip
mulch treatment. Spot fires in advance of the head fire grew quickly irrespective of fuel bed, and
supported the head fire rate of spread and intensity. Figure 7 illustrates spot fires developing on the
surface and within crown fuels.
Figure 7. Fire breaching mulch strips with spot fire development in advance of fire front.
Fire behaviour in the mulch strips appears to be at the threshold of an intermittent crown fire, but not
quite an active crown fire (Figure 8), as the fire briefly stalled while crossing the mulch strips before
igniting crown fuels on the other side. Stocks and Hartley (1995) described a similar fire (Experimental
fire#5/75) as an intermittent/fully developed crown fire.
FPInnovations Page 17
Figure 8. Intermittent crown fire in the strip mulch treatment.
Fire behaviour in the mulch thinning treatment
Fire transitioned quickly from intermittent/developing active crown fire in the natural stand to intense
surface fire in the mulch thinning treatment. After this transition, the fire behaviour varied across the
different sections of the treatment. Intensity and rate of spread were greater adjacent to the strip mulch
(Figure 7) as compared to the west flank with increasing fire intensity and accelerating rate of spread as
the fire progressed through the treatment.
Along the west flank of the fire in the thinning treatment, a backing (flanking) fire exhibited relatively low
spread rate and intensity. In the early stages of fire growth, convective indraft generated by the high-
intensity fire in the strip mulching treatment pulled the fire toward the central axis and limited forward
growth along the west flank (Figure 9).
FPInnovations Page 18
Figure 9. Relatively low fire intensity along the west flank of the thinned treatment.
Five minutes and 45 seconds later, as the head of the fire front advanced through the treatment area,
fire spread along the west flank changed dramatically with increased spread rate and a more forward
spread direction.
Ember transfer and spot fire development
In both treatments, short- and long-range ember transfer with almost instantaneous spot fire ignition
contributed to rapid fire growth. However, forward spread of spot fires was slowed by the strong indraft
generated by the high intensity of the main fire front.
Two minutes and 20 seconds after ignition start, spot fires developed in the strip mulching treatment at
the third line of rate of spread dataloggers (40 m from the natural fuel stand). Short-range spot fire
development was also captured on video within 2 minutes of ignition at a distance of 8 metres from the
north treatment boundary. Two minutes and 40 seconds after ignition commenced, a spot fire
developed near the north boundary of the inter-tree spacing treatment (Figure 10). Assuming that
crown fire in the natural black spruce stand generated the embers, the spotting distance was
approximately 70 to 80 m.
As spot fires joined together or the fire front burned over the spot fires, the treated area was covered by
a continuous blanket of surface fire with a flame height up to 1.5 m. in the mulch thinning treatment.
Flaming residence time was estimated at four minutes.
FPInnovations Page 19
Figure 10. Spot fire initiation along the north boundary in the early stages of fire development.
Residual trees within the thinned stand candled vigorously and canopy fuels created a good ignition
receptor for airborne embers. Very low foliar moisture content and abundant lichen in the black spruce
crowns (Johnston et al. 2015) contributed to ease of ignition in the canopy fuels. Several black spruce
trees with a live crown base height (LCBH) of at least 2 m candled along the bark of the stems initiating
combustion of canopy fuels (Figure 11).
FPInnovations Page 20
Figure 11. Vertical fire spread along residual stems with standard pruning height.
DISCUSSION
Fireline Intensity The calculated fireline intensities for the treated stands (from Table 8) seem much higher than other
test fires in standing timber when making visual comparisons (Alexander and De Groot 1988, Stocks
and Hartley 1995). Fuel consumption is similar among studies; the main difference is the higher spread
rate we observed. For example, Stocks and Hartley fire #8/76 had a spread rate of 16.2 m/min, fuel
consumption of 3.44 kg/m2 and fire intensity of 16,718 kW/m in mature pine and is clearly an active
crown fire. Alexander and De Groot fire Unit 6 documents fire behaviour, also in mature pine, with
spread rate of 6.1 m/min, fuel consumption of 3.92 kg/m2, and frontal intensity of 7460 kW/m, described
as a ‘developing’ active crown fire. We were able to assess rate of spread through the use of data
loggers and in-fire cameras and, therefore, feel confident that the documented spread rates are
accurate. The question remains: was the assumption of complete mulch consumption during head fire
passage valid? Alexander (1982) states that determining precise fuel consumption values in the field is
virtually impossible and suggests adjustments based on experience may need to be made.
Recent lab studies on the rate of mulch bed flaming and smouldering combustion by Thompson et al
(2016) showed that flaming consumption of 2 kg/m2 of mulch at 9% moisture content takes on the order
of four minutes. Though we observed a flaming residence time of four minutes, some of that time was
occupied by active flaming after the passage of the head fire. While active crown fires in natural forest
stands have residence times of 45-60 seconds (Stocks et al., 2004), if we extend the duration of the
active head fire to two minutes, an estimate of 50% mulch consumption in the active flaming phase is
made. Based on 50% of mulch being consumed in the active flaming zone during head fire passage,
revised calculations resulted in head fire intensity reduced by 37%.
FPInnovations Page 21
Effects of strip mulching treatment The observed rate of spread at the head of the strip mulch treatment was similar to the forecast value
for the C-2 fuel type. However, we observed intermittent crown fire within residual strips, suggesting
the treatment had a small effect on fire intensity. Intermittent crown fire in the residual strips easily
breached the mulched strips and sustained crown-to-crown fire spread. Hence, the strips did not
mitigate the fire spread rate, but marginally reduced fire intensity. Widening the strips might reduce the
intensity; however, in-fire imagery indicates that the residual strips would still generate crown fire and
produce additional ember transfer. The rate of spread documented in the thinned stand suggests that
wider mulched strips would not slow down the fire.
Effects of mulch thinned treatment Fire behaviour in the mulch thinning treatment was observed to be influenced by fire in the strip mulch
treatment. Fire spread rate within 20–30 m of the strips seemed to be accelerated by the higher-
intensity fire in the adjacent strip mulch treatment. In-stand video suggests spot fires generated by
ember transfer from the strip mulch and natural stand, along with strong indrafts from the head and
flanks of the strip mulch fire, would have “pulled” the fire along in the adjacent thinned stand (Figure
12). Intensity within this fast-moving portion of the fire was at times very high. The prevalence of fire
whirls is an indicator of high intensity and may result in short bursts greater than 10,000 kW/m (Hirsch
1996).
Figure 12. Influence of high intensity fire in strip mulch on fire behaviour in mulch thinned plot.
As a result, we cannot draw any conclusions about wind-driven fire behaviour in the thinned treatment
because of the effect of high-intensity fire in the strip mulch treatment. The potential for interaction was
considered when the treatment blocks were designed. It was expected that the strip mulch would not
FPInnovations Page 22
support such an intense fire and therefore interaction would have been minimal. The observed fire
behaviour still provides valuable insight into fire behaviour in mulch thinned black spruce.
The interaction effect showed that planning fuel treatments should not omit portions of a treatment area
that contain volatile fuels. This should be obvious to fuels practitioners, but could be a factor if land
ownership precludes ability to do fuel management. And, planners need to be aware that high-intensity
surface fires will occur with low humidity even at moderate wind speed in open mulch treatments, due
to fast rate of spread and available fuel. The intensity appears lower compared to a crown fire in a
natural stand, but rate of spread may be comparable. Without pre-positioned resources or rapid
response, suppression would have been problematic in our case study. This test fire demonstrates the
importance of having appropriately-sized treatments that consider suppression tactics when planning
fuel management.
Comparison with other black spruce wildfire/fuel management interactions Two other case studies of wildfire interaction with black spruce fuel treatments indicate that the fuel
removal approach can result in far greater reduction in fire intensity, albeit at lower indices. Table 9
provides a comparison.
Table 9. Comparison of fire behaviour resulting in black spruce stands treated through fuel removal.
Study Treatment Temp
(C) Rh (%)
Wind speed (km/h)
Residual Density
(stems/ha) Fire behaviour
Government of Saskatchewan, 2014, Lagoon Lake case study
Thin, prune, pile burn piles
20 25 19 1000
Crown fire entered treatment. Surface fire (air attack and ground crews actioned fire. Intermittent crown fire where surface fuel treatment had not been completed (debris pile burning). Air attack had limited effect in natural stand, indicated success in treated area. Allowed ground crews to action fire.
Butler et al., 2013.
Thin, prune, remove debris
(hauling)
21 47 3 G 20 1290-2359
Crown fire (38,990 kW/m) changed to surface fire and burned 30m into treated stand. Intensity was 160 kW/m and ROS was < 0.5 m/min. Authors note compaction in the treated stands may have affected fire behaviour in the treated stands, but did not in the control.
FPInnovations Page 23
Fuel Management Implications This case study demonstrates the effect of changing fuel structure as a fuel management treatment.
Mulching alone does not remove fuel; rather, the intent is to limit crown fire activity by displacing fuel to
the ground. In other words, the total fuel load was not changed after treatment, nor was it different
between the treatments. However, the amount of available fuel is increased. The mulch increased the
potential volatility of surface fuels by adding to the fuel volume through conversion of stems and large
branches to fine mulched fuels.
The Red Earth Creek fuel treatments were treated in the winter of 2013/14 and had one full season for
regeneration of surface vegetation and decomposition of chipped debris. Neither of these occurred to
an extent that would reduce the availability of surface fuels for ignition and consumption. The chipped
debris and feathermoss in the surface fuel layer were desiccated by very low relative humidity.
Additionally, there was limited migration of moisture from deeper peat layers caused by an ice layer at
the 7 cm depth below the duff surface. All of these factors contributed collectively to a dry surface fuel
layer that was very receptive to ember transfer and supportive of volatile fire behaviour. Partial removal
of mulch and other fine fuels would have decreased the surface fire intensity and rate of spread, but a
quantifiable relationship between fuel removed and resultant fire behaviour has not been developed.
A confounding factor was the abundant and extremely dry feathermoss within both treated areas.
Feathermoss was less dominant in the natural stand, sharing with Sphagnum moss (Table 3),
especially in the southeast corner of the natural fuel stand. Sphagnum moss has strong moisture
retention ability and fire behaviour was observed to be subdued in this part of the stand, as evidenced
by the unburned clumps of Sphagnum moss (Figure 13).
Figure 13. Reduced depth of burn and fuel consumption in hummocks of sphagnum moss.
FPInnovations Page 24
Logically, increasing Sphagnum moss cover while reducing feathermoss would be desirable as a part
of a FireSmart fuel treatment. Thinning may help in this cause as Sphagnum mosses are adapted to
open sunlight and are displaced by feathermoss as a black spruce stand matures and a denser canopy
shades out the forest floor (Benscoter and Vitt 2008, Kettridge et al. 2013). However, it is not yet
known how effectively Sphagnum moss propagation could be applied, especially at an operational
scale. The aforementioned mulch reduction burn testing might also benefit Sphagnum moss
propagation. These questions will be addressed in the next iteration of this research project.
Candling of residual stems and ignition of canopy fuels can be problematic even in aggressively thinned
and pruned fuel stands. In-stand video from this fire suggests that an intense surface fire can overcome
the live crown base height of 2 m typically prescribed in FireSmart guidelines. In addition, abundant
lichen in the canopy fuels, common in black spruce, are easily ignited and create additional airborne
embers as well as being capable of igniting from embers mid-canopy rather than from below.
Observations of other treated blocks at the Red Earth Creek site in July 2015 indicated prolific
regeneration of Labrador tea and broadleaf shrubs. As the surface fuel environment in fuel treatments
continues to evolve with regeneration and chip decomposition, these surface fuels may moderate fire
intensity (Kreye et al 2016). The extent to which vegetation succession and mulch decay interact to
moderate ignition potential and fire behaviour potential is not well documented and further monitoring
will contribute to a better understanding of the evolution of fuel treatments.
Enhanced suppression A reasonable target for FireSmart fuel management treatments is to limit fire behaviour to fire intensity
that will not exceed 4000 kW/m, a value that is determined to be the limit for direct ground attack. Fires
with higher intensity can be suppressed with the aid of airtankers and rotary wing aircraft. Our fireline
intensity calculations indicate that direct suppression would have had limited success in both treatment
types (Hirsch et al 1998). However, discussion with Slave Lake staff who conducted the experimental
fire indicated that suppression likely would have been successful in the thinned stand (ground crews
with charged hoses and aircraft support). Given the observed spot fire development in the thinning
treatment, aerial suppression may have been successful in slowing the fire spread and reducing the fire
intensity to a level that would have allowed fire crews to approach and attack the fires safely and
effectively. A case study of fire behaviour in a thinned black spruce stand in Saskatchewan indicated
that with aerial support, ground crews were able to successfully suppress fire in a treated stand
(Government of Saskatchewan 2014).
In an actual wildfire encounter with a fuel treatment in the WUI, the approaching fire front would likely
not be allowed to burn freely. Suppression in the natural fuel stands prior to impingement combined
with additional treatment of fuels within the treated areas would likely result in reduced rate of spread,
fire intensity, and ember transfer with improved overall effectiveness of the fuel treatment.
FPInnovations Page 25
CONCLUSION
This experimental fire represents one of few documented case studies of fire behaviour in black spruce
fuels that have had fuel management treatments. And of those case studies, the Red Earth
experimental fire was conducted at the highest fire danger indices.
Two other case studies indicated that fuel management, especially removal, along with suppression
resources can result in a desired outcome of stopping the head fire and limiting aerial ember transfer.
The Red Earth Creek case study demonstrated that at high fire danger indices suppression might be
challenged by the high spread rates we observed.
The strip mulching treatment in a black spruce stand had a limited effect on rate of spread and fireline
intensity when compared to the observed fire behaviour and calculated outputs for the natural black
spruce fuel stand (FBP C-2 fuel type). Calculated spread rate and head fire intensity were lower than
the C-2 outputs, but still supported a crown fire that would have been difficult to suppress.
Fire behaviour in the thinned stand was observed to be influenced by the adjacent strip mulch, and as a
result we cannot draw a conclusion about the effectiveness of this treatment in a wind-driven fire. The
effect of the interaction did show that black spruce stands thinned by mulching can support a fast-
moving fire that can exceed fireline intensity levels under which direct attack is deemed to be
successful.
This experimental fire demonstrated that fuel management in black spruce stands needs to consider
the effect of converting aerial fuel to the surface, and the composition of the ground cover. A notable
effect of mulching was to convert crown fuels including stems that do not burn in wildfires into available
surface fuel. This additional available fuel contributed to fireline intensities that may have been difficult
to control with direct ground attack. In addition, abundant feathermoss that is common to closed
canopy black spruce stands was extremely dry and contributed to high spread rates and intensity.
Any residual conifers within a FireSmart treatment have the potential to candle and spread aerial
embers, no matter what the live crown base height. Black spruce is especially vulnerable because it
can host abundant aerial lichens that are ideal ignition receptors for aerial embers. Displacing aerial
fuels by converting them to surface mulch will increase surface fire intensity and increase candling
vigour, which will result in more downwind aerial ember transport.
We are optimistic that the fuel management tactics used at Red Earth Creek FireSmart research area
can be modified to improve their effectiveness at reducing intensity and spread rate. Other tactics that
are being considered are listed below.
FPInnovations Page 26
NEXT STEPS
Experimental fires are planned to better understand fuel consumption and spread rates in open
mulched stands. Further experimental fires are being planned to evaluate thinning (with and without
post-mulch surface fuel treatment) and without the potential for interaction among plots.
Application of water, retardants and water-enhancing products will be studied to better understand the
effectiveness and longevity of these products in reducing ignition probability and potential fire
behaviour.
Research has been planned to evaluate the potential to increase Sphagnum moss abundance within
FireSmart treatments.
ACKNOWLEDGEMENTS
Many thanks to Alberta Agriculture and Forestry, partnering research agencies and all support
resources who contributed to the success of this project. Slave Lake Forest Area staff initiated this
project and provided resources and personnel to plan the project and execute the experimental fire.
McMaster University, University of Toronto and Canadian Forest Service conducted research projects
at the Red Earth Creek experimental fire and results from these projects will be presented through
separate publications at a later time.
FPInnovations Page 27
REFERENCES
Agee, J.K., & Skinner, C.N. (2005). Basic principles of forest fuel reduction treatments. Forest Ecology
and Management, 211(1), 83-96.
Alexander, M.E. (1982). Calculating and interpreting forest fire intensities. Canadian Journal of Botany,
60(4), 349-357.
Alexander, M.E., & De Groot, W.J. (1988). Fire behavior in jack pine stands as related to the Canadian
Forest Fire Weather Index System. Canadian Forest Service, Northern Forestry Centre, Edmonton, AB.
Poster with text.
Battaglia, M.A., Rocca, M.E., Rhoades, C.C., & Ryan, M.G. (2010). Surface fuel loadings within
mulching treatments in Colorado coniferous forests. Forest Ecology and Management, 260(9), 1557-
1566.
Benscoter, B.W., & Vitt, D.H. (2008). Spatial patterns and temporal trajectories of the bog ground layer
along a post-fire chronosequence. Ecosystems 11: 1054-1064.
Butler, B.W., Ottmar, R.D., Rupp, T.S., Jandt, R., Miller, E., Howard, K., Schmoll, R., Theisen, S.,
Vihnanek, R.E., & Jimenez, D. (2013). Quantifying the effect of fuel reduction treatments on fire
behavior in boreal forests. Can. J. For. Res. 43: 97-102.
Byram, G.M. (1959). Combustion of forest fuels. Pages 61-89 in K.P. Davis, ed. Forest fire: Control and
use. McGraw-Hill Book Co., New York, NY
Finney, M.A., & Cohen, J.D. (2003). Expectation and evaluation of fuel management objectives. In
USDA Forest Service Proceedings RMRS-P-29 (353-366).
Flat Top Complex Wildfire Review Committee. (2012). Flat top complex: Final report from the flat top
Complex Wildfire Review Committee. Alberta Ministry of Environment and Sustainable Resource
Development. Edmonton, Alberta. 55 pp. + Appendices. Available online at:
http://wildfire.alberta.ca/wildfire-prevention-enforcement/wildfire-reviews/documents/FlatTopComplex-
WildfireReviewCommittee-A-May18-2012.pdf
Forestry Canada Fire Danger Group. (1992). Development and structure of the Canadian Forest Fire
Behavior Prediction system. Forestry Canada, Ottawa. Inf. Rep. ST-X-3
Government of Saskatchewan. (2014). 14LA-lagoon fire: How fuel treatment areas affect wildland
urban interface fires. Available online at:
http://www.environment.gov.sk.ca/adx/aspx/adxGetMedia.aspx?DocID=4b81d714-58a5-4aca-a131-
c395590b101f
Hirsch, K.G. (1996). Canadian forest fire behaviour Prediction (FBP) System: user’s guide. Natural
Resources Canada. Canadian Forest Service. Northwest Region. Northern Forestry Centre. Edmonton,
Alberta. Special Report. 7.
FPInnovations Page 28
Hirsch, K.G., Corey, P.N., & Martell, D.L. (1998). Using expert judgment to model initial attack fire crew
effectiveness. Forest Science, 44(4), 539-549.
Hsieh, R. (2016). Comparing the canopy penetration of airtanker drops between forest fuel treatments
and untreated stands. FPInnovations Wildfire Operations Research Project report in progress.
Johnston, D.C., Turetsky, M.R., Benscoter, B.W., & Wotton, B.M. (2015). Fuel load, structure, and
potential fire behaviour in black spruce bogs. Canadian Journal of Forest Research, 45, 888-899.
Kettridge, N., Thompson, D.K., Bombonato, L., Turetsky, M.R., Benscoter, B.W., & Waddington, J.M.
(2013). The ecohydrology of forested peatlands: Simulating the effects of tree shading on moss
evaporation and species composition. Journal of Geophysical Research, 118, 1-14.
DOI:10.1002/jgrg.20043
Kreye, J.K., Varner, J.M., Kane, J.M., Knapp, E.E., & Reed, W.P.( 2016). The impact of aging on
laboratory fire behaviour in masticated shrub fuelbeds of California and Oregon, USA. International
Journal of Wildland Fire, in press.
Lawson, B.D., Armitage, O.B., & Hoskins, W.D. (1996). Diurnal variation in the Fine Fuel Moisture
Code: Tables and computer source code. Forest Resource Development Agreement Rep. 245,
Canadian Forest Service, Pacific Forestry Centre and British Columbia Ministry of Forests Research
Branch, Victoria, British Columbia.
Moghaddas, J.J., & Craggs, L. (2007). A fuel treatment reduces fire severity and increases suppression
efficiency in a mixed conifer forest. International Journal of Wildland Fire, 16, 673-678.
doi:10.1071/WF06066
Mooney, C. (2013). Can light stand thinning weaken an advancing crown fire? A case study in Black
Spruce (Picea mariana). FPInnovations Wildfire Operations Research report. Available online at:
http://wildfire.fpinnovations.ca/25/Report_LightThinning_CaseStudy1_v3.pdf
Reinhardt, E.D., Keane, R.E., Calkin, D.E., & Cohen, J.D. (2008). Objectives and considerations for
wildland fuel treatment in forested ecosystems of the interior western United States. Forest Ecology
and Management, 256(12), 1997-2006.
Schiks, T.J. & Wotton, B.M. (2015). Modifying the Canadian fine fuel moisture code for masticated
surface fuels. International Journal of Wildland Fire, 24(1), 79-91.
Schroeder, D. (2010). Fire behaviour in thinned jack pine: two case studies of FireSmart treatments in
Canada’s Northwest Territories. FPInnovations Advantage report 12(7).
Stocks, B.J., & Hartley, G.R. (1995). Fire behaviour in three jack pine fuel complexes (Poster with text).
Natural Resources Canada, Canadian Forest Service, Great Lakes Forestry Centre, Sault St. Marie,
Ontario.
FPInnovations Page 29
Stocks, B.J., Alexander, M.E., Wotton, B.M., Stefner, C.N., Flannigan, M.D., Taylor, S.W., Lavoie, N.,
Mason, J.A., Hartley, G.R., Maffey, M.E., & Dalrymple, G.N. (2004). Crown fire behaviour in a northern
jack pine black spruce forest. Canadian Journal of Forest Research, 34(8), 1548-1560.
Thompson, D.K., Schiks, T.J., & Wotton, B.M. (2016). Fuel size impacts on carbon residuals and
combustion dynamics in masticated woody debris. Forest Ecology and Management, 369, 59-65.
Wagner, C.V. (1977). Conditions for the start and spread of crown fire. Canadian Journal of Forest
Research, 7(1), pp.23-34.
Head Office
Pointe-Claire
570, Saint-Jean Blvd
Pointe-Claire, QC
Canada H9R 3J9
T 514 630-4100
Vancouver
2665 East Mall
Vancouver, BC.
Canada V6T 1Z4
T 604 224-3221
Québec
319, rue Franquet
Québec, QC
Canada G1P 4R4
T 418 659-2647
© 2016 FPInnovations. All rights reserved. Copying and redistribution prohibited.
® FPInnovations, its marks and logos are trademarks of FPInnovations
