THOMPSON RIVERS UNIVERSITY

TRENDS IN HOURLY WEATHER AND FIRE WEATHER INDICES ACROSS
BIOGEOCLIMATIC ZONES IN BRITISH COLUMBIA UNDER A CHANGING
CLIMATE
by
KATE ANTOINETTE HELEN BEZOOYEN

A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR
THE DEGREE OF
MASTER OF SCIENCE IN ENVIRONMENTAL SCIENCE
KAMLOOPS, BRITISH COLUMBIA
March 2025

Mike Flannigan (PhD), Research Chair and Thesis Supervisor, Faculty of Science
Jillian Harvey (PhD), Research Chair and Committee Member, Faculty of Science
John Karakatsoulis (PhD), Associate Teaching Professor and Committee Member, Faculty of
Science

© Kate Bezooyen, 2025

ii

ABSTRACT
The increased occurrence of hot, dry, and windy weather is resulting in worsening fire
conditions with five of the worst fire seasons on record in British Columbia occurring since
2017. Two critical daily periods have implications for fire growth: peak burn and overnight
conditions. This study aims to provide insights into changing conditions during peak burn
and overnight weather by investigating trends in hourly weather variables (temperature,
relative humidity, and windspeed) and associated fire weather indices (Fine Fuel Moisture
Code, Initial Spread Index, and Fire Weather Index) across different Biogeoclimatic
Ecosystem Classification (BEC) zones in British Columbia. Historical weather data from
June, July, August, and September spanning two periods, 1990-2005 and 2006-2021 was
analyzed at an hourly scale. The analysis revealed the greatest magnitude of change occurred
in the afternoon hours for all zones and months in the study. Spread event conditions are
occurring more frequently and BEC zones affected by the rain shadow effect are seeing more
significant results indicating warming and drying. This analysis of hourly weather highlights
the implications of observed trends on expected changes in fire behaviour with a spatial and
temporal lens and can be used to inform wildfire management practices across British
Columbia.

Key words: hourly weather, BEC zones, climate change, wildfire management

iii

ACKNOWLEDGMENTS

“Our greatest weakness lies in giving up. The most certain way to succeed is always to try
just one more time.” ―Attributed to Thomas Edison

This degree spanned several years and a whole lot of life. Throughout this journey, I have
been fortunate to receive the support, guidance, and kindness of many individuals, for whom
I am deeply grateful.
First, to my supervisor, Dr. Flannigan—your expertise, guidance, and understanding have
meant the world to me. This project would not have been possible without your support.
To my committee members, Dr. Harvey and Dr. Karakatsoulis—thank you for your
invaluable insights. Without a doubt, your perspectives have strengthened this project.
To my external examiner, Dr. Bourbonnais—thank you for your constructive feedback, which
has further refined this thesis.
Funding for this research was made possible through the NSERC Discovery Grant, and R
troubleshooting support was provided through Dr. Castellanos-Acuña.
To my friends, family, and lab mates—your unwavering support and encouragement over the
years have meant everything. I am especially grateful to KS and DD; I couldn’t have asked
for better friendships during my graduate experience. Finally, to RR and, of course, Bean—
thank you for being there through the highest highs and the lowest lows. You have been both
my strongest support system and my best distraction.

iv

TABLE OF CONTENTS
ABSTRACT.............................................................................................................................. ii
ACKNOWLEDGMENTS ....................................................................................................... iii
LIST OF TABLES .................................................................................................................... v
LIST OF FIGURES ................................................................................................................ vii
LIST OF ACRONYMS............................................................................................................ xi
CHAPTER 1 - INTRODUCTION ............................................................................................ 1
CHAPTER 2 - METHODS....................................................................................................... 7
Study Area ............................................................................................................................. 7
Weather data .......................................................................................................................... 9
Data Cleaning ...................................................................................................................... 10
Analysis ................................................................................................................................11
Mean, Median, and 95th percentile hourly comparison ....................................................11
Interpretation of mean, median and 95th percentile results............................................. 12
Spread Day Thresholds .................................................................................................... 12
CHAPTER 3 - RESULTS ....................................................................................................... 14
Mean, Median, and 95th percentile hourly comparison ....................................................... 14
Spread Day Thresholds ....................................................................................................... 29
CHAPTER 4 – DISCUSSION ................................................................................................ 48
Overall Findings .................................................................................................................. 48
Regional Findings ............................................................................................................... 48
Historical Fire Context ........................................................................................................ 50
Timing ................................................................................................................................. 52
Monthly ........................................................................................................................... 53
Hourly .............................................................................................................................. 53
Limitations & Future Research ........................................................................................... 54
CHAPTER 5 – CONCLUSION ............................................................................................. 57
REFERENCES ....................................................................................................................... 58
APPENDIX A – RESULTS OF MEAN, MEDIAN AND 95th PERCENTILE ANALYSIS –
1990-2005 COMPARED TO 2006-2021 ................................................................................ 63
APPENDIX B – MAGNITUDE OF CHANGE IN RESULTS OF MEAN, MEDIAN AND
95th PERCENTILE ANALYSIS – 1990-2005 COMPARED TO 2006-2021......................... 82

v

LIST OF TABLES
Table 1: Threshold values from the literature for Fine Fuel Moisture Code (FFMC), Initial
Spread Index (ISI) and Fire Weather Index (FWI) ................................................................... 5
Table 2: Mean climate characteristics for the Biogeoclimatic Ecosystem Classification (BEC)
zones used in the study from 1961-1990. Characteristics shown include area (millions of ha),
mean annual temperature (°C, MAT), mean warmest month temperature (°C, MWMT), mean
coldest month temperature (°C, MCMT), differential between MWMT and MCMT (°C, TD),
mean annual precipitation (mm, MAP), mean summer precipitation (mm, MSP, May to
September)(Centre for Forest Conservation Genetics, n.d.). .................................................... 9
Table 3: Summary of weather station elevation data grouped by BEC zones. ......................... 9
Table 4: Data ranges for each variable used in the study from the raw dataset compared to
data cleaned to both ±3 z, and ±4 z ..........................................................................................11
Table 5: Summary of significant findings of the 1728 statistical tests by BEC zone of the total
number, and percentage of significant findings for both a cooling and warming trend. Bold
values indicate the zone with the highest percentage in each category. ................................. 15
Table 6: Number of hours with significant change (Mean, Median, and 95th percentile)
towards a cooling and wetting trend between 1990-2005 and 2006-2021 by BEC zone,
month, and weather variable (Temperature, Relative Humidity, Wind Speed, Fine Fuel
Moisture Code, Initial Spread Index, and Fire Weather Index). The number of hours are
colour coded on a gradient from light grey to dark blue from 0 to 24 hours to highlight fields
with a higher count of significant findings with a cooling trend. ........................................... 18
Table 7: Number of hours with significant change (Mean, Median, and 95th percentile)
towards a warming and drying trend between 1990-2005 and 2006-2021 by BEC zone,
month, and weather variable (Temperature, Relative Humidity, Wind Speed, Fine Fuel
Moisture Code, Initial Spread Index, and Fire Weather Index). The number of hours are
colour coded on a gradient from green to red from 0 to 24 hours to highlight fields with a
higher count of significant findings with a warming trend. .................................................... 19
Table 8: Maximum difference between 1990-2005 and 2006-2021 for the Mean, Median, and
95th percentile values for each variable (Temperature, Relative Humidity, Wind Speed, Fine
Fuel Moisture Code, Initial Spread Index, and Fire Weather Index) by BEC zone in June. The
associated hour where the max occurs is shown, and the mean difference across all
significant findings. The top 10% of each difference is in bold. NA indicates no significant
findings. .................................................................................................................................. 20
Table 9: Maximum difference between 1990-2005 and 2006-2021 for the Mean, Median, and
95th percentile values for each variable (Temperature, Relative Humidity, Wind Speed, Fine
Fuel Moisture Code, Initial Spread Index, and Fire Weather Index) by BEC zone in July. The
associated hour where the max occurs is shown, and the mean difference across all

vi

significant findings. The top 10% of each difference is in bold. NA indicates no significant
findings. .................................................................................................................................. 21
Table 10: Maximum difference between 1990-2005 and 2006-2021 for the Mean, Median,
and 95th percentile values for each variable (Temperature, Relative Humidity, Wind Speed,
Fine Fuel Moisture Code, Initial Spread Index, and Fire Weather Index) by BEC zone in
August. The associated hour where the max occurs is shown, and the mean difference across
all significant findings. The top 10% of each difference is in bold. ....................................... 22
Table 11: Maximum difference between 1990-2005 and 2006-2021 for the Mean, Median,
and 95th percentile values for each variable (Temperature, Relative Humidity, Wind Speed,
Fine Fuel Moisture Code, Initial Spread Index, and Fire Weather Index) by BEC zone in
September. The associated hour where the max occurs is shown, and the mean difference
across all significant findings. The top 10% of each difference is in bold. NA indicates no
significant findings.................................................................................................................. 23

vii

LIST OF FIGURES
Figure 1:Weather parameter inputs and components of the Fire Weather Index System (Van
Wagner, 1987). .......................................................................................................................... 3
Figure 2: Map of British Columbia showing the Biogeoclimatic Ecosystem Classification
(BEC) zones used in this analysis: Boreal White and Black Spruce Zone (BWBS), Coastal
Western Hemlock (CWH), Interior Cedar Hemlock (ICH), Interior Douglas-Fir (IDF)
Montane Spruce (MS), and Sub Boreal Spruce (SBS). Locations of weather stations are
indicated by the circles.............................................................................................................. 8
Figure 3: The count of significant findings (α<0.05) with a warming trend per hour for a 24hour period analyzed by BEC zone between 1990-2005 and 2006-2021. Variables in the
analysis were the means of temperature, relative humidity, wind speed, FFMC, ISI, and FWI.
Monthly data (June, July, August, and September) was analyzed individually and is being
presented in aggregate............................................................................................................. 24
Figure 4: The count of significant findings (α<0.05) with a warming trend per hour for a 24hour period analyzed by BEC zone between 1990-2005 and 2006-2021. Variables in the
analysis were the medians of temperature, relative humidity, wind speed, FFMC, ISI, and
FWI. Monthly data (June, July, August, and September) was analyzed individually and is
being presented in aggregate. .................................................................................................. 25
Figure 5: The count of significant findings (α<0.05) with a warming trend per hour for a 24hour period analyzed by BEC zone between 1990-2005 and 2006-2021. Variables in the
analysis were the 95th percentiles of temperature, relative humidity, wind speed, FFMC, ISI,
and FWI. Monthly data (June, July, August, and September) was analyzed individually and
is being presented in aggregate. .............................................................................................. 26
Figure 6: The aggregated count per hour of significant findings (α<0.05) with a warming
trend when the maximum difference occurred by BEC zone. Analyses included a
comparison of mean, median, and 95th percentile temperature for six variables (relative
humidity, wind speed, FFMC, ISI, and FWI) for each month (June, July, August, September)
over two time periods (1990-2005 and 2006-2021) for a total of 72 comparisons completed
for each hourly timestep.......................................................................................................... 27
Figure 7: The count of significant findings (α<0.05) with a warming trend, grouped by
periods, for each BEC ZONE. Analyses included comparisons of hourly temperature, relative
humidity, wind speed, FFMC, ISI, and FWI for each month (June, July, August, and
September) over two time periods (1990-2005 and 2006-2021). Counts are summarized by
mean, median, 95th percentile, and hour of maximum difference. ......................................... 28
Figure 8: BWBS BEC Zone: Percentage of observed FFMC values exceeding a threshold of
≥91 across four daily periods: 1) 0000h-0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 18002300h by month for data grouped from 1990-2005 and 2006-2021. The secondary Y-axis

viii

shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the
more recent dataset (2006-2021). ........................................................................................... 30
Figure 9: CWH BEC Zone: Percentage of observed FFMC values exceeding a threshold of
≥91 across four daily periods: 1) 0000h-0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 18002300h by month for data grouped from 1990-2005 and 2006-2021. The secondary Y-axis
shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the
more recent dataset (2006-2021). ........................................................................................... 31
Figure 10: ICH BEC Zone: Percentage of observed FFMC values exceeding a threshold of
≥91 across four daily periods: 1) 0000h-0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 18002300h by month for data grouped from 1990-2005 and 2006-2021. The secondary Y-axis
shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the
more recent dataset (2006-2021). ........................................................................................... 32
Figure 11: IDF BEC Zone: Percentage of observed FFMC values exceeding a threshold of
≥91 across four daily periods: 1) 0000h-0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 18002300h by month for data grouped from 1990-2005 and 2006-2021. The secondary Y-axis
shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the
more recent dataset (2006-2021). ........................................................................................... 33
Figure 12: MS BEC Zone: Percentage of observed FFMC values exceeding a threshold of
≥91 across four daily periods: 1) 0000h-0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 18002300h by month for data grouped from 1990-2005 and 2006-2021. The secondary Y-axis
shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the
more recent dataset (2006-2021). ........................................................................................... 34
Figure 13: SBS BEC Zone: Percentage of observed FFMC values exceeding a threshold of
≥91 across four daily periods: 1) 0000h-0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 18002300h by month for data grouped from 1990-2005 and 2006-2021. The secondary Y-axis
shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the
more recent dataset (2006-2021). ........................................................................................... 35
Figure 14: BWBS BEC Zone: Percentage of observed ISI values exceeding a threshold of
≥8.7 across four daily periods: 1) 0000h-0500h, 2) 0600-1100h, 3) 1200h-1700h and 4)
1800-2300h by month for data grouped from 1990-2005 and 2006-2021. The secondary Yaxis shows the percentage change of the FFMC values from the earlier dataset (1990-2005)
to the more recent dataset (2006-2021). ................................................................................. 36
Figure 15: CWH BEC Zone: Percentage of observed ISI values exceeding a threshold of
≥8.7 across four daily periods: 1) 0000h-0500h, 2) 0600-1100h, 3) 1200h-1700h and 4)
1800-2300h by month for data grouped from 1990-2005 and 2006-2021. The secondary Yaxis shows the percentage change of the FFMC values from the earlier dataset (1990-2005)
to the more recent dataset (2006-2021). ................................................................................. 37
Figure 16: ICH BEC Zone: Percentage of observed ISI values exceeding a threshold of ≥8.7
across four daily periods: 1) 0000h-0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 18002300h by month for data grouped from 1990-2005 and 2006-2021. The secondary Y-axis

ix

shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the
more recent dataset (2006-2021). ........................................................................................... 38
Figure 17: IDF BEC Zone: Percentage of observed ISI values exceeding a threshold of ≥8.7
across four daily periods: 1) 0000h-0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 18002300h by month for data grouped from 1990-2005 and 2006-2021. The secondary Y-axis
shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the
more recent dataset (2006-2021). ........................................................................................... 39
Figure 18: MS BEC Zone: Percentage of observed ISI values exceeding a threshold of ≥8.7
across four daily periods: 1) 0000h-0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 18002300h by month for data grouped from 1990-2005 and 2006-2021. The secondary Y-axis
shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the
more recent dataset (2006-2021). ........................................................................................... 40
Figure 19: SBS BEC Zone: Percentage of observed ISI values exceeding a threshold of ≥8.7
across four daily periods: 1) 0000h-0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 18002300h by month for data grouped from 1990-2005 and 2006-2021. The secondary Y-axis
shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the
more recent dataset (2006-2021). ........................................................................................... 41
Figure 20: BWBS BEC Zone: Percentage of observed FWI values exceeding a threshold of
≥19 across four daily periods: 1) 0000h-0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 18002300h by month for data grouped from 1990-2005 and 2006-2021. The secondary Y-axis
shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the
more recent dataset (2006-2021). ........................................................................................... 42
Figure 21: CWH BEC Zone: Percentage of observed FWI values exceeding a threshold of
≥19 across four daily periods: 1) 0000h-0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 18002300h by month for data grouped from 1990-2005 and 2006-2021. The secondary Y-axis
shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the
more recent dataset (2006-2021). ........................................................................................... 43
Figure 22: ICH BEC Zone: Percentage of observed FWI values exceeding a threshold of ≥19
across four daily periods: 1) 0000h-0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 18002300h by month for data grouped from 1990-2005 and 2006-2021. The secondary Y-axis
shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the
more recent dataset (2006-2021). ........................................................................................... 44
Figure 23: IDF BEC Zone: Percentage of observed FWI values exceeding a threshold of ≥19
across four daily periods: 1) 0000h-0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 18002300h by month for data grouped from 1990-2005 and 2006-2021. The secondary Y-axis
shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the
more recent dataset (2006-2021). ........................................................................................... 45
Figure 24: MS BEC Zone: Percentage of observed FWI values exceeding a threshold of ≥19
across four daily periods: 1) 0000h-0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 18002300h by month for data grouped from 1990-2005 and 2006-2021. The secondary Y-axis

x

shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the
more recent dataset (2006-2021). ........................................................................................... 46
Figure 25: SBS BEC Zone: Percentage of observed FWI values exceeding a threshold of
≥19 across four daily periods: 1) 0000h-0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 18002300h by month for data grouped from 1990-2005 and 2006-2021. The secondary Y-axis
shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the
more recent dataset (2006-2021). ........................................................................................... 47

xi

LIST OF ACRONYMS
AT – Alpine Tundra
BEC -Biogeoclimatic Ecosystem Classification
BCWS – British Columbia Wildfire Service
BG - Bunchgrass
BUI – Build-up Index
BWBS – Black and White Boreal Spruce
CDF – Coastal Douglas Fir
CWH – Coastal Western Hemlock
DC – Drought Code
DMC – Duff Moisture Code
ESSF – Englemann Spruce Subalpine Fir
FFMC – Fine Fuel Moisture Content
FTS – Forest Technology Systems
FWI – Fire Weather Index
ICH – Interior Cedar Hemlock
IDF – Interior Douglas Fir
IPCC – Intergovernmental Panel on Climate Change
ISI – Initial Spread Index
MAP – Mean Annual Precipitation
MAT – Mean Annual Temperature
MCMT – Mean Coldest Month Temperature
MH – Mountain Hemlock
MODIS – Moderate Resolution Imaging Spectroradiometer
MS – Montane Spruce
MSP – Mean Summer Precipitation
MWMT – Mean Warmest Month Temperature
PP – Ponderosa Pine
RH – Relative Humidity
SBPS – Sub-boreal Pine-Spruce
SBS – Sub-boreal Spruce
SWB – Spruce Willow Birch
TD – Differential between MWMT and MCMT
TEMP - Temperature
WS – Wind Speed

1

CHAPTER 1 - INTRODUCTION
Climate change & fire
Human-caused greenhouse gas emissions have resulted in global surface temperatures
increasing by 1.1 °C as of 2020 (Intergovernmental Panel on Climate Change, 2023). All
regions, globally, have experienced climate extremes and are expected to see increased
occurrence and adverse effects associated with natural disasters as global warming intensifies
hazards (Intergovernmental Panel on Climate Change, 2023). Climate change is leading to
longer, hotter, and drier fire seasons with increasing area burned (Flannigan et al. 2009).
A notable amount of research has been conducted on the implications of climate
change on wildland fires. Flannigan et al. (2009) conducted a meta-analysis examining the
effects of changing climate on global wildland fire across 47 studies published between 1991
and 2009. They found that fire activity, including fire weather, area burned, fire occurrence,
fire season, fire intensity, and fire severity were associated with future climate change
emission scenarios (Flannigan et al. 2009). Their synthesis highlighted a consistent pattern of
results suggesting an overwhelmingly positive relationship between climate change and
wildland fire globally, with only one study finding a negative relationship (specific to
Brandenburg, Germany) and 10 studies with trends showing mixed results, with both positive
and negative trends. Further research into future climate change scenarios has continued to
show agreement on a trend toward extreme fire weather and aggressive fire behaviour
(Abatzoglou et al., 2019; Ellis et al., 2021; Flannigan et al., 2009; Guindon et al., 2021; Jain
et al., 2021; Jolly et al., 2015; Meyn et al., 2010; Parisien et al., 2023; Wang et al., 2015;
Wotton et al., 2017).
Wildfire fundamentals
The physical process by which combustion occurs is a well-understood phenomenon;
however, the behaviour of fire varies depending on the characteristics of the fire
environment. The fire environment is comprised of fuel, weather, and topography
(Countryman, 1972). The interactions among these three components create a dynamic
environment for fire by either accelerating or dampening the process of heat transfer (Van

2

Wagner, 1983; Agee, 1996; Sullivan, 2017), which results in a mosaic of potential fire
intensities across the landscape (Van Wagner, 1983).
Weather has a dominant influence on fire behaviour due to a complex series of
processes that both directly and indirectly affect fire (Flannigan et al. 2009). Temperature,
relative humidity, wind and precipitation affect fuel moisture (Van Wagner, 1983; Sullivan,
2017) and fuel moisture determines whether wildfires can ignite and sustain combustion.
Temperature has a strong influence on the capacity of air to hold moisture. For every 1°C
increase in temperature, precipitation needs to increase by over 15% to counteract the fuel
moisture loss in fine fuels, 10% in medium fuels, and 5% for deep or heavy fuels (Flannigan
et al., 2016). Winds affect the drying rate of fine fuels and precipitation directly adds
moisture, with increasing thresholds required to impact deeper fuel layers (Van Wagner,
1987). Relative humidity influences the moisture content of dead fuels. Low relative
humidity has been linked with the development of extreme fire behaviour (Werth et al.,
2016).
In addition to fuel moisture impacts, wind is also key factor in determining the rate
and direction of fire spread. Winds supply oxygen to the fire which increases combustion.
Winds also tilt the angle of flame closer to fuels ahead of the flaming front, accelerating fire
spread through pre-heating and winds facilitate lofting and transfer of embers ahead of the
flaming front into available fuels. Ember loft is increased in areas of atmospheric instability,
which is associated with strong updrafts, gusty surface winds, dust devils, fire whirls,
pyrocumulonimbus development, strong downdrafts, and extreme fire behaviour (Werth et
al., 2016). When hot, dry, windy weather conditions coincide with continuous dry fuels
across the landscape, fires experience rapid growth, expelling substantial quantities of heat
energy (Sullivan, 2017) and challenging suppression efforts.
FWI System Overview
The influence of weather on fire growth and spread is of critical importance to
understand and forecast fire behaviour potential. Thus, the Fire Weather Index (FWI) System
was developed by federal government fire researchers and has been in use since the 1970s
(Van Wagner, 1987; Wotton, 2009). The FWI System uses measurements of the key weather
variables: temperature, 10-wind speed, relative humidity, and 24-h precipitation recorded

3

daily at noon LST (Lawson & Armitage, 2008). The weather variables are used to calculate
six fire weather indices representative of peak burning conditions at 1600 h (Figure 1)
(Forestry Canada Fire Danger Group, 1992; Taylor & Alexander, 2003; Lawson & Armitage,
2008). There are three primary indices of fuel moisture content, the Fine Fuel Moisture Code
(FFMC), Duff Moisture Code (DMC), and Drought Code (DC).

Precipitation
Relative Humidity
Wind Speed
Temperature

Precipitation
Relative Humidity
Temperature

Fine Fuel Moisture
Code

Duff Moisture
Code

Drought Code

FFMC

DMC

DC

Wind Speed

Precipitation
Temperature

Initial Spread
Index

Build-up Index

ISI

BUI

Fire Weather Index

FWI
Figure 1:Weather parameter inputs and components of
the Fire Weather Index System (Van Wagner, 1987).

The fuel moisture codes are differentiated by fuel size class and quantity, water
holding capacity, and the time it takes for fuels to dry or wet to meet equilibrium moisture
content with the environment (Van Wagner, 1987). The Fine Fuel Moisture Code (FFMC)
represents fuel moisture in litter, small twigs, and other fine fuels. Precipitation (>0.5mm in
the last 24 hours), increased relative humidity and cooler temperatures can cause these fuels
to draw in moisture from the environment. The drying or wetting rate will increase or
decrease depending on the atmospheric conditions, but drying in this layer is approximately
12 hours at 25°C, 30% relative humidity and 10km/h wind speeds (Wotton, 2009). The
FFMC value indicates fuel receptivity to ignition and fire spread, which is dependent on the

4

availability of fine dry fuels. The Duff Moisture Code (DMC) represents the layer of fuels
beneath litter, up to 7cm, which is comprised of decomposing organic material, and has a
drying time of approximately 10 days at 25°C and 30% relative humidity. Cumulative 24hour precipitation exceeding 1.5mm is required to effectively add moisture back to this fuel
layer (Van Wagner, 1987). Finally, the third moisture index, the Drought Code (DC)
represents fuel moisture in the deep organic layer and large-diameter fuels on the forest floor.
These fuels take approximately 50 days to dry under 25°C conditions and require cumulative
24-hour precipitation to exceed 2.8mm to effectively increase moisture content (Van Wagner,
1987).
The fuel moisture codes, FFMC, DMC and DC, are then used to derive two
intermediate indices, which represent the potential rate of fire spread, the Initial Spread Index
(ISI) which considers the FFMC and wind speed, and the potential fuel consumption,
represented by the Build-Up Index (BUI) and derived from a combination of the DMC and
DC. Finally, an overall index, the Fire Weather Index (FWI), is calculated from the ISI and
BUI to represent the overall potential level of fire intensity calculated from the indicators of
possible fuel consumption and rate of spread.
Fire Growth Thresholds
Thresholds of the FWI system can be used by management agencies to understand
potential fire danger due to current and forecast conditions. Most area burned occurs during
periods of extreme fire weather (Rothermel et al., 1994). A ‘spread event’ is defined as an
event where a wildfire exhibits substantial growth in a relatively short amount of time (Podur
& Wotton, 2011). FWI system values associated with extreme fire weather can be used to
understand conditions where fires are likely to exhibit ease of ignition, high intensity, and
rapid growth rates (Table 1). An FFMC of 91 and above is associated with vigorous fire
spread (Taylor & Alexander, 2006; Wotton, 2009). An ISI threshold of 8.7 has been shown
as a key indicator of spread event potential (Podur & Wotton, 2011; Wang et al., 2014).
Although there is no threshold for BUI, which is less effective as a spread event indicator
(Wang et al., 2023), the overall FWI takes into consideration the ISI and BUI and a threshold
value of 19 has been linked to active fire spread (Podur & Wotton, 2011). Wang et al. (2023)
found FFMC was the most robust measure of spread events across large geographic areas but

5

also recommend ISI and FWI as the top three indicators of choice. Their study addressed
spread events by season spring (March, April, May), summer (June, July, August) and fall
(September, October, November) across Canada and used ecozones as the geographic study
area, providing a more nuanced approach to spread event thresholds regionally across the
country.

Table 1: Threshold values from the literature for Fine Fuel Moisture Code (FFMC), Initial
Spread Index (ISI) and Fire Weather Index (FWI)
FWI System Indices
FFMC
ISI
FWI

Spread Day Threshold
91
8.7
19

Reference
Taylor and Alexander (2006); Wotton (2009)
Podur and Wotton (2011); Wang et al. (2014)
Podur and Wotton (2011)

Fire Management and the Diurnal Cycle
A typical operational firefighting period is between 1000h and 1900h. This ensures
resources (crews, equipment, aircraft) are available during the peak burning conditions of a
day to respond to new fire starts and work to prevent wildfire escapes. The diurnal weather
patterns of temperature and relative humidity lead to peak fire activity occurring between
1600h and 1800h when fuel moisture is at its lowest point. The increased occurrence of hot,
dry, and windy weather is resulting in worsening fire conditions. Five of the worst fire
seasons on record in British Columbia have occurred since 2017 (Government of British
Columbia, 2025). Extreme fire weather results in fire behaviour that exceeds the capacity
and effectiveness of suppression personnel and equipment. Under periods of extreme fire
growth, resources are removed from active duty due to safety concerns.
Conversely, the least active time for fire activity is overnight and early morning when
relative humidity is at its highest leading to fuel moisture recovery. However, in recent years,
more observations of overnight fire growth have occurred (Balch et al., 2022; Luo et al.,
2024). These studies have shown the number of flammable overnight burning hours has
increased globally from 1979-2020 (Balch et al., 2022), overnight wildfire intensity has

6

increased from 2003-2020 (Balch et al., 2022) and that overnight burning events were
associated with large fires from 2017-2020 (Luo et al., 2024) Globally, daily minimum
temperatures were found to rise at a faster rate than daily max temperatures, leading to a
narrowing of the diurnal temperature range (Easterling et al. 1997).
Many of the assumptions operational personnel have on daily fire behaviour are being
challenged due to recent extreme fire weather conditions. Many studies agree that fire
weather and fire behaviour are expected to become more extreme under a changing climate
(Abatzoglou et al., 2019; Flannigan et al., 2009; Jain et al., 2021; Jolly et al., 2015; Meyn et
al., 2010; Parisien et al., 2023; Wang et al., 2015; Wotton et al., 2017). However, the nuances
of changing weather haven’t yet been explored at the diurnal scale. This study aims to
understand how changing weather patterns are impacting regions in British Columbia by
analyzing trends in hourly weather across the full 24 hours of observed historical weather
data and hourly fire weather indices.

Research question:
This thesis investigates the following research question:
What trends (geographically, and diurnally) in hourly weather (temperature, relative
humidity, and wind speed), fire weather indices (FFMC, ISI, and FWI), and spread event
thresholds can be detected from a historical weather dataset spanning two periods (19902005 and 2006-2021) across biogeoclimatic ecosystem classification (BEC) zones in British
Columbia?

7

CHAPTER 2 - METHODS
Study Area
This study focuses on understanding weather and fire weather variation across the
province of British Columbia. The province is geographically diverse leading to variable
climatic conditions across the land base. Orographic influences of the coastal and rocky
mountain ranges affect moisture patterns from West to East off the Pacific Ocean. Latitudes
ranging from 48°N to 60°N span a variety of ecosystems from semi-arid deserts along the
southern border of British Columbia to boreal forests in the north and altitudes spanning sea
level to alpine tundra.
Due to the diversity of ecosystems within British Columbia, it is important to take a
detailed approach to analysis and avoid aggregating data from complex geographical areas
(Meyn et al., 2010). British Columbia’s unique ecosystems are classified into 14 zones, with
each zone representing generally homogenous patterns of vegetation and climate. These
zones are characterized by the Biogeoclimatic Ecosystem Classification system (BEC;
Meidinger & Pojar, 1991). This study uses BEC zones as analysis units. BEC zones
accurately reflect unique groupings of homogenous macroclimate characteristics, thus they
are an effective unit for examining the implications of climate change (DeLong et al., 2010).
However, because BEC zones are ecologically different, the fire effects experienced under a
changing climate will be realized differently depending on regional conditions. Furthermore,
BEC zones are classified into subzones and variants to further identify unique ecosystems
within the broad classes. Fire behaviour and ecological impacts will further vary within
these subzones and with localized physiographic features like slope and aspect.
Of the 14 BEC zones, eight zones were omitted from analysis due to insufficient data
(<20 weather stations) and included the Alpine Tundra zone (AT: 0 stations), Spruce-WillowBirch (SWB: 0 stations), Sub-Boreal Pine Spruce (SBPS: 3 stations), Mountain Hemlock
(MH: 4 stations), Englemann Spruce Subalpine Fir (ESSF: 13 stations), Bunchgrass (BG: 2
stations), Ponderosa Pine (PP: 4 stations), and Coastal Douglas Fir (CDF: 4 stations).
BEC zones with sufficient weather stations for analysis (Figure 2) included the Boreal
White and Black Spruce Zone (BWBS: 29 stations), Coastal Western Hemlock (CWH: 49

8

stations), Interior Cedar Hemlock (ICH: 53 stations), Interior-Douglas Fir (IDF:33 stations),
Montane Spruce (MS: 23 stations), and Sub Boreal Spruce (SBS: 44 stations) BEC zones.
Mean climate characteristics for each zone are summarized in Table 2, and a summary of
weather station elevation data (mean, minimum, maximum) are in Table 3.

Figure 2: Map of British Columbia showing the Biogeoclimatic Ecosystem Classification
(BEC) zones used in this analysis: Boreal White and Black Spruce Zone (BWBS), Coastal
Western Hemlock (CWH), Interior Cedar Hemlock (ICH), Interior Douglas-Fir (IDF)
Montane Spruce (MS), and Sub Boreal Spruce (SBS). Locations of weather stations are
indicated by the circles.

9

Table 2: Mean climate characteristics for the Biogeoclimatic Ecosystem Classification (BEC)
zones used in the study from 1961-1990. Characteristics shown include area (millions of ha),
mean annual temperature (°C, MAT), mean warmest month temperature (°C, MWMT), mean
coldest month temperature (°C, MCMT), differential between MWMT and MCMT (°C, TD),
mean annual precipitation (mm, MAP), mean summer precipitation (mm, MSP, May to
September)(Centre for Forest Conservation Genetics, n.d.).
Zone

Area

MAT

MWMT

MCMT

TD

MAP

MSP

BWBS
CWH
ICH
IDF
MS
SBS

15.8
11
5.24
4.14
2.74
10.31

0.4
6.5
3.3
4.2
1.5
2.3

14.0
14.6
14.6
15.4
12.1
13.7

-14.0
-1.0
-8.6
-7.5
-9.3
-10.2

27.9
15.7
23.1
22.9
21.4
23.9

638
2256
779
503
692
652

333
494
302
203
239
286

Table 3: Summary of weather station elevation data grouped by BEC zones.

Zone
BWBS
CWH
ICH
IDF
MS
SBS

Number of
Stations
29
49
53
33
23
44

Elevation (m)
Minimum
382
21
191
330
861
600

Maximum
1126
869
1608
1220
1683
1580

Mean
774
310
907
890
1408
897

Weather data
Historical records of hourly weather data, including temperature, relative humidity,
wind speed, and calculated FWI indices (FFMC, ISI, and FWI) from a total of 231 stations
were downloaded from the British Columbia Wildfire Service DataMart and compiled into
.csv files. Weather records were grouped by BEC zone, and month (June, July, August,
September), and separated into two time periods for analysis (1990-2005 and 2006-2021),
with each period grouping observations by hour for direct hour-to-hour comparisons between
the two periods.
Precipitation was not included in this study of hourly weather due to the spatial
variability across the landscape. Unlike observations of temperature, relative humidity and

10

windspeed which are measurable across the landscape at any given time, precipitation can be
very patchy across geographic areas. Indirectly, precipitation is accounted for in the study
through the calculated FWI indices, but as a stand-alone variable it was excluded from
analysis.
The weather stations used by the British Columbia Wildfire Service are made by
Forest Technology Systems (FTS) and transmit station data to satellites at hourly intervals.
Temperature records represent the temperature (°C) at hourly intervals, whereas recordings of
relative humidity (%) and windspeed (km/h) represent the 10-minute average before the
hourly transmission (FTS Inc. 2015). In the event of a station failure producing a missing
value, the data is interpolated from nearby stations using an inverse distance weighting
methodology. Interpolated temperature is corrected using a lapse rate of -6.5°C/km to
account for elevation differences and relative humidity values are adjusted using an
elevation-adjusted temperature grid in combination with the gridded mixing ratio of water
vapour to dry air. The completed and continuous record of hourly weather observations are
then used to calculate the FWI indices following the methodology outlined in Van Wagner
and Pickett (1985).
Data Cleaning
The dataset generated for this study consisted of hourly weather observations from
231 weather stations across British Columbia from 1990-2021. The raw weather
observations were reviewed for outliers using the z-score method for data quality control.
The z-score method is a standardized assessment to identify observations that are unusually
far from the mean. Typically, studies using the z-score method will classify outliers when
there is a z-score of ±2 or ±3. Because this study is particularly interested in capturing
observations at the extreme ends, it was found that a z-score of ±4 was more appropriate to
exclude what appeared to be true outliers from the observations, without excessively
trimming true extremes (Table 4). Data trimmed using a z-score ±3 overly constrained the
range of data, likely excluding true observations. The FFMC scale ranges from 0-101, a few
values in the original dataset exceeded this, likely due to minor calculation or rounding
errors, cleaning the data reduced the number of occurrences leaving only one value above
101 in the final dataset.

11

Table 4: Data ranges for each variable used in the study from the raw dataset compared to
data cleaned to both ±3 z, and ±4 z
Variable

Data Range (min-max)
Raw

Cleaned (±3 z)

Cleaned (±4 z)

Temperature

-50-60

-8-34

-12-42

Relative Humidity

0-100

0-100

0-100

Wind Speed

0-65

0-21

0-26

FWI

0-160

0-38

0-48

FFMC

0-105

0-102

0-102

ISI

0-115

0-12

0-15

Analysis
Mean, Median, and 95th percentile hourly comparison
Hourly weather variables (temperature, relative humidity, windspeed) and Fire
Weather Index variables (FFMC, ISI, FWI) were compared between two time periods (19902005, 2006-2021) by month (June, July, August, and September) and by BEC zone (BWBS,
CWH, ICH, IDF, MS, SBS). Comparisons were conducted at an hour-by-hour scale, where
data from each time period (1990-2005, 2006-2021) was pooled to represent overall
conditions occurring at each hourly timestep through the diurnal cycle. Therefore, statistical
tests looked directly at how weather at X hour of the day during period 1, compared to the
same X hour during period 2.
Weather data from individual stations was aggregated and summarized at the BEC
zone level to understand changes across each region, rather than at each station location.
Hourly mean, median, and 95th percentiles of each of the three weather variables and three
FWI variables from the two time periods were compared. All analysis was completed using
R Studio (R Core Team, 2022). Base R was used to conduct the Welch 2-sample t-test to
compare means, and the WRS2 statistical analysis package (Mair and Wilcox, 2020) was
used to conduct a quantile analysis of variance (Qanova) on the median and 95th percentile
values. The WRS2 package is more robust to outliers compared to a traditional ANOVA. All
tests used a significance value of α<0.05.

12

Interpretation of mean, median and 95th percentile results
This study has a focus on informing decisions made at an operational wildfire level.
Several studies have confirmed climate change is showing trends towards hotter, drier and
windier conditions (Abatzoglou et al., 2019; Ellis et al., 2021; Flannigan et al., 2009;
Guindon et al., 2021; Jain et al., 2021; Jolly et al., 2015; Meyn et al., 2010; Parisien et al.,
2023; Wang et al., 2015; Wotton et al., 2017). This study looked across six variables to detect
a change between two time periods at the hourly level. Results were interpreted to broadly
infer a trend in the fire environment towards conditions reflecting an environment with either
higher fuel moisture or lower fuel moisture between the two time periods.
Each variable was assessed independently for the direction of change (positive or
negative) between the 1990-2005 and 2006-2021 datasets for each of the three tests run
(mean, median and 95th percentile). The six variables chosen for analysis in the study all
relate to fuel moisture. If conditions are showing an increase in temperature and wind speed,
or a decrease to relative humidity, fuels will dry more quickly and the fire environment
would be more hazardous, as indicated by the fire weather indices (FFMC, ISI and FWI)
where higher values indicate drier fuels and greater potential for wildfire behaviour (Van
Wagner, 1987).
Therefore, this study considers a ‘warming, drying trend’ where significant findings
of the variables assessed were in the direction of a drier fire environment, or more hazardous
conditions, and a ‘cooling, wetting’ trend where significant findings were in the direction of a
less hazardous fire environment, as indicated by a decrease in temperature, wind speed, fire
weather indices, or an increase to relative humidity.
Due to the use of longer time periods (16 years) which mask short term variability,
and comparison of multiple variables, the results can be used to infer a broader trend
occurring due to climate change and inform operational decisions in wildfire management.
Spread Day Thresholds
Examining the differences in how often Fire Weather Index thresholds associated
with spread events were exceeded was challenging due to varying completeness in records
and the number of observation points in each daily period. To address this, data were pooled
into four daily periods: 1) 0000h-0500h, 2) 0600h-1100h, 3) 1200h-1700h and 4) 1800h-

13

2300h and the total percentage of records exceeding a threshold (FFMC ≥91, ISI ≥8.7, FWI ≥
19) was calculated and compared between the 1990-2005 dataset, and the 2006-2021 dataset,
by month and BEC zone. The percentage of recorded observations was selected to avoid
skewing results due to a greater number of observations in the 2006-2021 dataset. Percent
change of each daily period was then calculated to compare between the 1990-2005 and
2006-2021 time periods by month and BEC zone.

14

CHAPTER 3 - RESULTS
Mean, Median, and 95th percentile hourly comparison
Each of the six BEC zone in this study had 1728 statistical tests (24 hours x 4 months x 6
weather variables x 3 measures - mean, median, and 95th percentile) run on hourly data
comparisons from the 1990-2005 dataset and the 2006-2021 dataset for a total number of 10,368
statistical tests. Of all the significant results generated in this study, the majority were associated
with a warming and drying trend (Table 5). The SBS zone had the greatest number of significant
results with 59.6%, followed by the IDF at 54.4% and the ICH at 47.6%. Less than 1% of results
were significant with a cooling trend for CWH, ICH, IDF, and SBS. The MS had the greatest
amount of cooling with 2.1% followed by the BWBS at 1.2%.
The number of hours with a significant difference (α<0.05) in the mean, median, and 95th
percentile was totaled and summarized by variable, month, and BEC zone. Significant shifts
with a cooling and wetting trend are summarized in

15

Table 6, and significant shifts with a warming, and drying trend are summarized in Table
7. Overall, very few hours showed a cooling trend. The MS zone had the most cooling compared
to the other zones. Cooling trends occurred most often in June or September, and least often in
July and August.
Warming trends were seen most often in July for all zones. August had the second most
warming trends for BWBS, CWH, MS, and SBS whereas June had the second most warming in
both ICH and IDF. The SBS zone had the most significant warming hours overall, followed by
the IDF and ICH. Most significant warming trends were associated with changes in the mean
value, followed by the median (Temperature, RH, FFMC, ISI, and FWI). Wind speed had the
most significant shifts in the median, followed closely by the mean, and the least significant
shifts in the 95th percentile compared to the other variables. Significant changes in the 95th
percentile was most prevalent for FWI, with the SBS zone having the greatest number of
significant hours at the 95th percentile weather with a warming trend.

Table 5: Summary of significant findings of the 1728 statistical tests by BEC zone of the total
number, and percentage of significant findings for both a cooling and warming trend. Bold
values indicate the zone with the highest percentage in each category.
Cooling, Wetting
Zone
BWBS
CWH
ICH
IDF
MS
SBS

Number of
significant results
20
11
7
6
36
10

Percentage with
significant results
1.2%
0.6%
0.4%
0.3%
2.1%
0.6%

Warming, Drying
Number of
significant results
422
675
823
940
412
1030

Percentage with
significant results
24.4%
39.1%
47.6%
54.4%
23.8%
59.6%

The figures in Appendix A visually represent the diurnal data curves for mean, median,
and 95th percentile hourly values with significant results shown and the magnitude of change
visualized between the 1990-2005 and 2006-2021 datasets shown in Appendix B. Significant
findings were summarized by month (June: Table 8, July: Table 9, August: Table 10, September:
Table 11) with the maximum significant difference of mean, median, and 95th percentile between

16

the 1990-2005 and 2005-2021 datasets shown for each variable by zone. The associated hour
when the max difference was found is specified, and the overall mean of the significant
differences is shown. Negative RH values, or positive temperature, wind speed, FFMC, ISI, and
FWI values indicate a warming trend over time.
In June, the MS and IDF had the largest differences compared to the other zones (Table
8). The 95th percentile temperature in the MS increased an average of 3.5°C across all
significant hours, with a maximum of 5.1°C at 1400h. MS had the greatest change in FFMC,
with a max median difference of 40.9 occurring at 0400h. The IDF zone had the highest mean
and 95th percentile shifts in relative humidity, with the 95th percentile decreasing 12.5% at
0500h. The windspeed also showed the greatest shift in the IDF zone with the mean 0800h winds
increasing by 4.4 km/h and the median increasing by 4.8km/h.
The same trend is evident in reviewing the July data (Table 9). The IDF and MS show
the greatest changes compared to the other BEC zones. Mean FFMC, ISI, and FWI were all
greatest in the IDF zone, with the biggest peaks occurring throughout the day (0700h, 1400h, and
2300h, for FFMC, ISI, and FWI respectively).
In August, the MS zone had the greatest mean (2.9 °C increase at 1300h) and median (2.8
°C increase at 1300h) temperature change compared to the other BEC zones (Table 10). The
maximum decrease in hourly RH occurred in the SBS where the mean decreased by a maximum
of 9.8% and the median decreased by a maximum of 15%. The IDF zone saw the greatest
increased wind speed across the mean, median, and 95th percentile values with maximums
occurring at 1400h, 2100h, and 0700h respectively. The CWH zone had the greatest increases in
95th percentile FFMC, ISI, and FWI, all occurring at 1200h and the median FFMC increased by
a maximum of 16.4 at 1600h. Mean FFMC, ISI, and FWI were greatest in the IDF all occurring
at 1200h, and median ISI and FWI were also highest in the IDF at 1200h.
The data for September shows greater warming in the BWBS zone, which has the highest
mean and median temperature increases (2.7 °C maximum mean change and 3.3 °C maximum
median change, both occurring at 1000h) (Table 11). The 95th percentile temperature increases
were second highest in BWBS, surpassed by the MS zone with a 4.5 °C maximum change
occurring at 1300h. The IDF had the greatest change in WS, ISI, and the mean and median FWI.
The MS zone had the highest increase in mean FFMC at 0100h, but the overall trend in the

17

median across all significant hours showed a marginal decline. The SBS had the highest overall
change in the median across all significant hours.
The greatest September RH changes were seen across different zones. The maximum
mean change occurred in the BWBS, but the overall greatest mean change was in the IDF. The
IDF also had the greatest maximum median and overall median change. However, the 95th
percentile changes decreased the highest amount in the CWH zone (26% decrease at 0500h, and
18% overall). Of interest, the CWH zone also had the greatest change in 95th percentile relative
humidity in July and August with a reduction in July of 18.7% at 2200h and an average decline
of 13.6% across all significant hours. In August, the maximum decrease was 21% at 2200h, with
a mean of 11.6% and in September the 0500h relative humidity decreased by 26%, with an 18%
overall decrease.
The frequency of significant findings per hour are summarized in Figure 3 to Figure 5.
These figures provide a visual representation of how often each hour was associated with a
significant finding in the comparisons of the six variables (temperature, relative humidity, wind
speed, FFMC, ISI, and FWI) across the four months tested in this study (June, July, August,
September). The significant findings are shown for the comparisons of mean (Figure 3), median
(Figure 4), and 95th percentile (Figure 5). The hours associated with the maximum significant
difference in Table 8 to Table 11 were summarized in Figure 6. Generally, the maximum
significant differences occurred throughout all hours in the day, however, there was a slight
increase during daytime hours from 0700h to 1800h. The ICH and BWBS zones had the highest
count of maximum significant differences at 1400h. The IDF, MS, SBS, and CWH had the
highest count of maximum significant differences at 1200h.
The significant findings were further summarized by periods in Figure 7 to better detect
trends in the diurnal cycle. The SBS, IDF, and ICH zones have the greatest number of
significant results. The IDF and SBS had the greatest mean warming trends during the second
period (0600-1100h). The BWBS zone had fewer significant findings in the mean comparison in
the fourth period (1800h-2300h) compared to the other periods, and the CWH had greater
significant findings in the mean comparison in the fourth period (1800h-2300h) compared to the
other three time periods.

18

There were fewer significant changes in the median compared to the mean, or 95th
percentile analyses. Unlike the mean comparison, the third period (1200h-1700h) had the most
frequent significant results for the ICH and SBS zones. The MS had the fewest significant
median results, with most occurring during the third period (1200h-1700h). The IDF followed a
similar trend as the mean comparison, with the second period (0600h-1100h) having the greatest
number of significant findings.
The SBS zone had the greatest number of significant findings in the 95th percentile
comparison, with slightly more findings occurring in the third period (1200h-1700h). The CWH
and ICH had similar results with the most significant results landing in the third and fourth
periods (spanning 1200h-2300h) compared to the earlier hours. The maximum significant
differences of mean, median, and 95th percentile occurred in the third period (1200h-1700h)
across all BEC zones but were most evident in the BWBS, CWH, ICH, and IDF.

19

Table 6: Number of hours with significant change (Mean, Median, and 95th percentile) towards a cooling and wetting trend between
1990-2005 and 2006-2021 by BEC zone, month, and weather variable (Temperature, Relative Humidity, Wind Speed, Fine Fuel
Moisture Code, Initial Spread Index, and Fire Weather Index). The number of hours are colour coded on a gradient from light grey to
dark blue from 0 to 24 hours to highlight fields with a higher count of significant findings with a cooling trend.
Temp Decrease
M
Mdn
95

RH Increase
M
Mdn
95

JUNE
JULY
AUGUST
SEPTEMBER

0
0
0
0

1
0
0
0

0
0
0
0

1
0
0
2

1
0
0
1

2
0
1
2

JUNE
JULY
AUGUST
SEPTEMBER

0
0
0
0

1
0
0
0

0
0
0
0

0
0
0
0

0
0
0
0

1
0
0
0

JUNE
JULY
AUGUST
SEPTEMBER

0
0
0
0

3
0
0
0

0
0
0
0

0
0
0
1

0
0
0
0

0
0
0
1

JUNE
JULY
AUGUST
SEPTEMBER

0
0
0
0

2
0
0
0

0
0
2
0

0
0
0
0

0
0
0
0

0
0
0
1

JUNE
JULY
AUGUST
SEPTEMBER

0
0
1
0

0
0
1
1

0
0
0
0

0
0
0
2

0
0
0
2

2
0
2
1

JUNE
JULY
AUGUST
SEPTEMBER

1
0
0
1

1
0
0
1

1
0
0
0

0
0
0
0

0
0
0
0

0
0
0
1

Cooling and wetting trend
WS Decrease
FFMC Decrease
M
Mdn
95
M
Mdn
95
BWBS
0
0
0
0
0
2
0
0
1
0
0
0
0
0
2
0
0
0
0
0
2
0
0
0
CWH
0
0
1
0
0
2
0
0
0
0
0
0
0
0
1
0
0
0
0
0
1
0
1
0
ICH
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
IDF
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
MS
0
0
5
0
0
0
0
0
4
0
0
0
0
0
1
0
0
3
0
0
1
0
3
0
SBS
0
0
2
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
1
0
0
0

ISI Decrease
M
Mdn
95

FWI Decrease
M Mdn 95

0
0
0
0

0
0
0
0

0
0
0
0

0
0
0
0

0
0
0
1

0
0
0
1

0
0
0
0

1
0
0
1

0
0
0
0

0
0
0
0

1
0
0
0

0
0
0
0

0
0
0
0

0
0
0
0

0
0
0
0

0
0
0
1

0
0
0
0

0
0
0
0

0
0
0
0

0
0
0
0

0
0
0
0

0
0
0
0

0
0
0
0

0
0
0
0

0
0
0
1

0
0
0
3

0
0
1
0

0
0
0
0

0
0
0
2

0
0
0
0

0
0
0
0

0
0
0
0

0
0
0
0

0
0
0
0

0
0
0
0

0
0
0
0

20

Table 7: Number of hours with significant change (Mean, Median, and 95th percentile) towards a warming and drying trend between
1990-2005 and 2006-2021 by BEC zone, month, and weather variable (Temperature, Relative Humidity, Wind Speed, Fine Fuel
Moisture Code, Initial Spread Index, and Fire Weather Index). The number of hours are colour coded on a gradient from green to red
from 0 to 24 hours to highlight fields with a higher count of significant findings with a warming trend.
Temp Increase
M
Mdn
95

RH Decrease
M
Mdn
95

JUNE
JULY
AUGUST
SEPTEMBER

2
12
4
5

1
8
3
3

4
8
4
6

5
12
5
5

4
10
3
6

2
6
5
2

JUNE
JULY
AUGUST
SEPTEMBER

4
12
10
13

0
11
6
9

6
5
8
9

6
18
14
12

7
16
16
13

3
6
6
7

JUNE
JULY
AUGUST
SEPTEMBER

3
15
4
7

0
15
6
4

5
10
3
7

14
17
9
8

13
17
8
8

9
11
1
2

JUNE
JULY
AUGUST
SEPTEMBER

4
15
5
5

4
16
3
3

5
6
2
2

11
18
14
6

13
18
14
7

7
9
1
2

JUNE
JULY
AUGUST
SEPTEMBER

4
11
3
1

3
8
3
1

6
1
2
3

5
10
4
1

3
8
2
1

1
6
3
3

JUNE
JULY
AUGUST
SEPTEMBER

4
17
12
10

2
15
8
7

1
12
4
10

16
22
20
12

17
22
19
13

5
14
8
6

Warming and drying trend
WS Increase
FFMC Increase
M
Mdn
95
M
Mdn
95
BWBS
7
6
2
5
1
4
6
10
0
14
10
10
10
14
1
7
2
5
3
5
0
9
5
2
CWH
14
12
3
3
0
2
16
16
3
11
4
9
20
21
1
15
4
9
12
15
1
0
0
11
ICH
24
24
3
18
8
11
24
24
3
23
12
11
24
24
3
17
5
3
24
23
2
9
3
1
IDF
24
24
9
19
9
10
24
23
10
22
24
5
24
24
8
18
13
2
24
24
10
11
2
1
MS
5
8
0
13
5
6
12
10
0
17
3
5
8
12
1
14
4
3
8
10
2
5
1
1
SBS
24
23
1
22
11
7
22
24
0
20
9
21
23
24
3
21
12
16
22
21
0
15
6
9

ISI Increase
M
Mdn
95

FWI Increase
M Mdn
95

4
12
6
5

3
10
4
7

4
8
5
3

4
19
12
6

2
9
5
6

2
12
9
2

3
16
14
8

1
20
15
0

3
9
12
12

8
21
16
15

2
18
14
0

9
12
14
14

18
24
16
5

10
20
12
4

14
15
9
2

15
24
14
8

6
23
13
2

19
17
10
4

21
24
22
12

18
24
22
14

11
16
5
7

21
24
21
13

14
24
19
11

14
14
8
7

7
17
10
1

5
11
6
1

3
7
3
2

7
18
11
5

8
15
8
2

4
7
5
3

21
24
23
12

15
18
16
6

9
23
20
9

19
24
20
16

13
19
16
9

13
24
19
10

21

Table 8: Maximum difference between 1990-2005 and 2006-2021 for the Mean, Median, and
95th percentile values for each variable (Temperature, Relative Humidity, Wind Speed, Fine Fuel
Moisture Code, Initial Spread Index, and Fire Weather Index) by BEC zone in June. The
associated hour where the max occurs is shown, and the mean difference across all significant
findings. The top 10% of each difference is in bold. NA indicates no significant findings.
JUNE
MEDIAN

MEAN
Max
Diff

Hour
of Max

Mean
Diff

BWBS
CWH
ICH
IDF
MS
SBS

1.7
1.9
1.3
1.9
3.0
1.1

9
16
18
15
12
13

1.4
1.2
1.2
1.4
1.9
0.6

BWBS
CWH
ICH
IDF
MS
SBS

-7.7
-8.1
-7.7
-10.8
-10.3
-8.1

9
8
7
9
10
11

-4.2
-5.6
-5.5
-7.1
-8.4
-5.4

BWBS
CWH
ICH
IDF
MS
SBS

2.0
2.4
2.8
4.4
2.3
2.7

17
15
20
8
12
12

1.8
1.2
2.0
3.0
2.0
1.7

BWBS
CWH
ICH
IDF
MS
SBS

10.4
10.4
16.2
21.7
27.6
15.9

14
12
12
14
4
9

8.8
9.4
9.3
11.4
15.8
9.4

BWBS
CWH
ICH
IDF
MS
SBS

1.2
1.0
1.1
1.8
2.1
1.1

14
12
12
14
12
12

0.7
0.8
0.7
1.1
1.1
0.7

BWBS
CWH
ICH
IDF
MS
SBS

3.1
2.5
2.6
5.1
4.8
2.5

14
20
12
21
12
1

2.1
1.7
1.7
3.2
3.0
1.7

Max
Hour
Diff
of Max
TEMP
1.5
9
-1.6
18
-0.8
16
1.8
15
2.5
10
1.2
14
RH
-11.0
2
-10.0
6
-11.0
2
-11.0
2
-12.0
10
-10.0
6
WS
2.5
9
2.7
4
3.7
2
4.8
8
3.2
8
3.3
10
FFMC
7.5
14
NA
NA
21.7
6
21.2
14
40.9
4
17.0
1
ISI
1.9
14
0.7
3
0.8
14
2.5
20
1.6
12
1.2
1
FWI
4.7
14
2.7
12
1.7
18
7.0
14
5.6
12
2.9
1

95th PERCENTILE
Mean
Diff

Max
Diff

Hour
of Max

Mean
Diff

0.2
-1.6
-1.0
0.6
1.8
0.4

3.4
4.3
3.0
3.5
5.1
1.5

9
16
15
17
14
6

2.2
2.7
2.4
2.3
3.5
0.1

-5.0
-6.2
-6.7
-7.7
-9.3
-6.7

-7.4
-9.6
-7.6
-12.5
-6.8
-7.6

9
3
6
5
12
6

0.9
-3.3
-5.7
-6.6
5.3
-5.6

1.9
1.9
2.5
3.6
2.1
2.2

3.2
3.5
3.8
6.9
-3.0
3.5

2
21
23
11
16
22

2.8
1.3
3.0
4.3
-4.5
-0.1

7.5
NA
14.5
8.3
27.4
9.6

1.3
2.7
3.3
2.3
6.2
2.7

12
21
12
13
12
21

0.3
0.2
2.0
1.8
3.3
1.8

1.2
0.0
0.5
1.4
1.3
0.8

2.4
3.7
3.6
4.5
5.9
2.8

14
12
12
11
12
21

1.9
2.5
2.4
2.6
3.2
2.0

3.6
1.0
1.2
4.1
3.2
2.0

6.9
9.8
8.9
10.6
11.5
7.6

22
21
12
16
19
23

5.7
6.3
5.3
6.9
10.1
5.3

22

Table 9: Maximum difference between 1990-2005 and 2006-2021 for the Mean, Median, and
95th percentile values for each variable (Temperature, Relative Humidity, Wind Speed, Fine Fuel
Moisture Code, Initial Spread Index, and Fire Weather Index) by BEC zone in July. The
associated hour where the max occurs is shown, and the mean difference across all significant
findings. The top 10% of each difference is in bold. NA indicates no significant findings.
JULY
MEDIAN

MEAN
Max
Diff

Hour
of Max

Mean
Diff

BWBS
CWH
ICH
IDF
MS
SBS

3.1
2.1
2.8
2.9
3.2
2.1

16
14
14
15
13
15

1.6
1.5
1.5
1.5
2.0
1.4

BWBS
CWH
ICH
IDF
MS
SBS

-11.0
-10.6
-10.5
-14.1
-12.4
-10.1

10
17
14
7
18
12

-6.4
-6.2
-7.1
-7.8
-8.8
-6.1

BWBS
CWH
ICH
IDF
MS
SBS

1.9
2.2
2.4
4.5
2.6
2.2

8
18
2
15
9
21

1.5
1.4
1.9
2.9
1.9
1.5

BWBS
CWH
ICH
IDF
MS
SBS

15.6
11.7
16.9
24.3
19.6
16.8

12
2
14
7
13
12

9.2
9.4
10.8
11.8
12.8
9.4

BWBS
CWH
ICH
IDF
MS
SBS

1.5
1.5
2.4
2.7
2.1
1.7

16
14
14
14
18
12

0.8
0.8
1.1
1.6
1.2
0.9

BWBS
CWH
ICH
IDF
MS
SBS

4.3
4.9
6.6
10.0
8.0
5.2

12
22
14
23
21
18

2.2
3.0
3.9
6.0
4.5
3.2

Max
Hour
Diff
of Max
TEMP
3.1
16
2.6
12
3.4
14
4.4
1
3.7
13
2.6
8
RH
-12.0
10
-17.5
17
-13.0
4
-21.0
7
-12.5
8
-11.5
15
WS
2.8
13
2.6
8
3.7
2
5.1
23
3.8
10
3.3
10
FFMC
10.9
13
6.4
14
13.8
13
10.7
7
5.6
21
14.9
11
ISI
1.6
16
1.6
17
2.8
14
3.4
15
2.8
21
1.9
18
FWI
3.9
20
4.6
14
8.0
14
13.3
23
12.1
21
5.0
12

95th PERCENTILE
Mean
Diff

Max
Diff

Hour
of Max

Mean
Diff

1.8
1.9
1.8
1.9
2.6
1.4

3.2
3.6
3.0
3.1
3.4
2.8

10
13
22
1
16
18

2.4
2.7
2.1
1.9
3.4
2.0

-6.9
-8.0
-9.2
-9.6
-9.4
-7.1

-11.0
-18.7
-15.9
-7.0
-17.3
-12.9

14
22
2
7
5
8

-7.3
-13.6
-8.1
-5.1
-7.5
-8.4

1.6
1.8
2.6
3.5
2.3
1.9

-3.2
3.9
3.2
5.8
-3.0
NA

18
14
19
13
12
NA

-3.2
3.6
2.9
3.9
-4.4
NA

6.8
4.5
7.9
3.7
4.1
8.5

3.6
3.8
3.5
2.5
3.0
4.7

12
12
6
12
12
12

2.3
3.0
1.9
1.5
2.3
2.4

0.9
0.8
1.2
2.1
1.5
0.9

4.7
3.7
4.3
3.8
4.5
4.3

16
13
22
12
5
12

2.4
2.4
2.8
2.4
3.1
2.4

2.4
3.1
4.3
8.6
6.0
3.0

11.6
16.5
15.3
11.9
17.3
13.3

12
15
2
7
5
18

7.4
8.6
9.8
6.5
8.8
8.7

23

Table 10: Maximum difference between 1990-2005 and 2006-2021 for the Mean, Median, and
95th percentile values for each variable (Temperature, Relative Humidity, Wind Speed, Fine Fuel
Moisture Code, Initial Spread Index, and Fire Weather Index) by BEC zone in August. The
associated hour where the max occurs is shown, and the mean difference across all significant
findings. The top 10% of each difference is in bold.
AUGUST
MEDIAN

MEAN
Max
Diff

Hour
of Max

Mean
Diff

BWBS
CWH
ICH
IDF
MS
SBS

1.3
1.4
1.9
1.5
2.9
1.5

17
7
14
2
13
13

1.1
1.2
1.2
1.3
1.3
0.9

BWBS
CWH
ICH
IDF
MS
SBS

-6.6
-8.1
-7.1
-9.6
-8.8
-9.8

10
10
14
9
15
11

-4.9
-6.0
-5.0
-5.8
-7.8
-5.5

BWBS
CWH
ICH
IDF
MS
SBS

2.4
2.3
2.7
3.8
2.7
2.6

14
10
12
14
10
6

1.5
1.4
1.9
2.8
1.9
1.5

BWBS
CWH
ICH
IDF
MS
SBS

11.9
16.2
16.2
22.4
16.4
18.7

10
7
14
12
3
8

9.5
10.5
9.6
10.6
11.5
9.6

BWBS
CWH
ICH
IDF
MS
SBS

0.9
1.7
1.7
2.6
1.4
1.5

12
12
14
12
19
12

0.7
0.9
0.8
1.2
1.1
0.7

BWBS
CWH
ICH
IDF
MS
SBS

3.0
5.9
4.9
8.9
6.0
4.9

21
12
19
12
22
0

2.0
3.4
3.1
4.7
4.2
3.0

Max
Hour
Diff
of Max
TEMP
1.9
8
1.9
8
2.2
6
2.2
15
2.8
13
1.9
8
RH
-10.0
6
-12.0
10
-9.0
22
-14.5
7
-12.0
10
-15.0
8
WS
2.7
12
2.8
13
3.5
20
4.1
21
2.9
21
3.4
1
FFMC
7.0
10
16.4
16
12.0
11
5.4
9
3.3
19
14.0
8
ISI
1.0
14
1.8
15
2.1
14
3.1
12
1.8
19
1.6
19
FWI
3.3
14
6.6
16
6.0
14
12.4
12
9.4
22
6.7
19

95th PERCENTILE
Mean
Diff

Max
Diff

Hour
of Max

Mean
Diff

1.5
1.3
1.4
1.6
1.5
1.1

3.1
2.9
1.8
2.1
2.7
2.0

17
11
0
18
17
13

2.1
2.5
1.4
0.2
2.1
1.6

-9.3
-6.7
-7.3
-7.2
-10.5
-6.7

-6.1
-21.0
-5.1
-6.3
-6.9
-14.0

2
22
12
12
10
8

-2.6
-11.6
-5.1
-6.3
1.6
-7.8

1.9
1.9
2.6
3.4
2.2
2.0

4.9
2.6
3.3
5.8
5.0
4.3

9
11
3
7
10
6

-1.0
-0.3
2.7
3.6
0.6
1.8

6.9
8.1
6.9
2.8
2.7
6.9

3.2
6.1
2.1
2.4
2.0
3.5

12
12
12
12
11
12

1.9
3.2
1.4
0.7
0.0
1.9

0.7
0.8
1.0
1.6
1.2
0.9

2.7
5.6
3.3
2.9
3.8
3.2

12
12
14
10
11
12

2.0
2.4
2.4
2.4
1.4
1.9

2.2
4.1
3.6
6.5
5.5
3.3

10.5
17.0
11.1
7.9
9.6
11.7

21
12
19
12
6
0

6.7
8.6
6.9
5.7
9.2
7.4

24

Table 11: Maximum difference between 1990-2005 and 2006-2021 for the Mean, Median, and
95th percentile values for each variable (Temperature, Relative Humidity, Wind Speed, Fine Fuel
Moisture Code, Initial Spread Index, and Fire Weather Index) by BEC zone in September. The
associated hour where the max occurs is shown, and the mean difference across all significant
findings. The top 10% of each difference is in bold. NA indicates no significant findings.
SEPTEMBER
MEDIAN

MEAN
Max
Diff

Hour
of Max

Mean
Diff

BWBS
CWH
ICH
IDF
MS
SBS

2.7
2.0
1.6
1.7
1.7
1.5

10
4
13
3
12
10

1.9
1.2
1.0
1.4
1.7
0.9

BWBS
CWH
ICH
IDF
MS
SBS

-9.3
-6.2
-7.5
-9.1
-8.2
-7.9

10
11
13
8
12
11

-3.7
-4.6
-3.5
-6.1
2.0
-4.9

BWBS
CWH
ICH
IDF
MS
SBS

1.9
1.8
2.6
4.3
2.7
2.2

1
22
20
15
4
23

1.6
1.2
1.8
2.9
2.0
1.5

BWBS
CWH
ICH
IDF
MS
SBS

13.6
NA
13.9
11.8
15.6
11.9

14
NA
13
13
1
6

9.0
NA
7.9
8.5
11.1
8.3

BWBS
CWH
ICH
IDF
MS
SBS

0.8
0.6
0.6
1.4
0.7
0.5

18
19
13
12
1
22

0.6
0.4
0.4
0.9
-0.1
0.4

BWBS
CWH
ICH
IDF
MS
SBS

2.9
3.3
2.6
4.6
3.4
2.8

18
19
23
0
20
22

1.9
1.7
1.3
3.4
2.8
1.3

Max
Hour
Diff
of Max
TEMP
3.3
10
1.8
16
1.9
9
2.2
15
2.5
9
1.4
15
RH
-12.0
10
-7.0
23
-7.5
11
-13.5
22
-6.0
22
-10.0
6
WS
2.9
13
2.7
4
3.5
20
4.7
13
4.6
22
3.3
6
FFMC
14.1
14
-16.4
9
17.3
9
6.2
7
18.1
1
17.6
10
ISI
1.0
14
-0.4
9
0.5
13
1.3
16
1.1
1
0.6
13
FWI
2.5
2
NA
NA
1.4
0
5.9
4
3.4
20
1.8
23

95th PERCENTILE
Mean
Diff

Max
Diff

Hour
of Max

Mean
Diff

2.4
1.4
1.2
1.7
0.5
0.8

4.3
3.8
3.1
3.3
4.5
2.9

18
18
17
15
13
14

3.0
2.8
1.9
3.1
3.3
2.0

-4.6
-3.2
-4.7
-8.3
5.2
-5.0

-8.4
-26.0
-8.4
-11.0
-7.5
-14.9

10
5
8
7
13
14

2.1
-18.0
-0.8
-0.8
-1.9
-7.1

1.9
1.9
2.5
3.7
2.3
2.0

-3.3
3.6
3.1
6.1
4.7
-3.5

5
8
8
20
18
16

-3.8
-1.9
2.8
4.3
0.9
-3.5

8.4
-16.4
12.5
4.5
-0.1
13.3

2.6
6.5
1.9
2.9
1.9
3.3

18
3
12
12
5
12

2.6
4.5
0.2
2.9
1.9
2.1

0.6
-0.4
0.4
0.9
-0.6
0.4

3.1
3.3
1.7
3.5
2.5
2.1

18
12
12
12
23
17

2.5
2.0
1.5
2.0
2.3
1.5

1.4
NA
1.3
4.2
-0.5
0.9

10.5
14.0
7.7
9.2
9.2
9.8

18
19
23
12
2
22

3.7
8.6
6.0
6.9
7.8
5.0

25

Figure 3: The count of significant findings (α<0.05) with a warming trend per hour for a 24-hour
period analyzed by BEC zone between 1990-2005 and 2006-2021. Variables in the analysis were
the means of temperature, relative humidity, wind speed, FFMC, ISI, and FWI. Monthly data
(June, July, August, and September) was analyzed individually and is being presented in
aggregate.

26

Figure 4: The count of significant findings (α<0.05) with a warming trend per hour for a 24-hour
period analyzed by BEC zone between 1990-2005 and 2006-2021. Variables in the analysis were
the medians of temperature, relative humidity, wind speed, FFMC, ISI, and FWI. Monthly data
(June, July, August, and September) was analyzed individually and is being presented in
aggregate.

27

Figure 5: The count of significant findings (α<0.05) with a warming trend per hour for a 24-hour
period analyzed by BEC zone between 1990-2005 and 2006-2021. Variables in the analysis were
the 95th percentiles of temperature, relative humidity, wind speed, FFMC, ISI, and FWI.
Monthly data (June, July, August, and September) was analyzed individually and is being
presented in aggregate.

28

Figure 6: The aggregated count per hour of significant findings (α<0.05) with a warming trend
when the maximum difference occurred by BEC zone. Analyses included a comparison of mean,
median, and 95th percentile temperature for six variables (relative humidity, wind speed, FFMC,
ISI, and FWI) for each month (June, July, August, September) over two time periods (1990-2005
and 2006-2021) for a total of 72 comparisons completed for each hourly timestep.

29

Figure 7: The count of significant findings (α<0.05) with a warming trend, grouped by periods, for each BEC ZONE. Analyses
included comparisons of hourly temperature, relative humidity, wind speed, FFMC, ISI, and FWI for each month (June, July, August,
and September) over two time periods (1990-2005 and 2006-2021). Counts are summarized by mean, median, 95th percentile, and
hour of maximum difference.

30

Spread Day Thresholds
A comparison between the 1990-2005 dataset and the 2006-2021 dataset of the
percentage of observations exceeding known thresholds for fire spread events was summarized
by zone for FFMC (Figure 8 to Figure 13), ISI (Figure 14 to Figure 19) and FWI (Figure 20 to
Figure 25).
The ICH and IDF zones had the greatest percentage of values that exceeded the threshold
for FFMC (≥91). The greatest number of thresholds were exceeded during period 4 (1800h2300h), followed by period 3 (1200h-1700h) with the lowest amount of FFMC thresholds
exceeded in period 2 (0600h-1100h). There was an observed trend of more thresholds being
exceeded in 2006-2021 compared to 1990-2005. When reviewing the percentage change, the
BWBS had the greatest percent change compared to all other zones, and it was highest in period
2 (0600h-1100h) for July, August, and September. Overall, in the BWBS, there was a small
percentage of values exceeding, but it had the greatest change compared to the other zones.
For the ISI threshold of ≥8.7, most thresholds were exceeded during the third period
(1200-1700h), followed by fourth (1800-2300h), and the least in the second (0600-1100h). The
zones with the highest percentage of thresholds exceeded were the IDF, then ICH, and finally
MS. Overall, higher percentages of thresholds were observed in 2006-2021 compared to 19902005 except a decreased ISI threshold observations most common in September for the CWH,
ICH and SBS. The IDF zone had the greatest percent change in ISI during the June 0600h-1100h
period and the second greatest change occurred in the third June period (1200-1700h) in the ICH
zone.
The FWI threshold of ≥19 was exceeded most often in the IDF zone. The MS zone had
the second highest percentage followed by the ICH. Period 3 (1200h-1700h) had the highest
number of values over the threshold, followed by period 4 (1800-2300h), period 1 (0000h0500h), and the least observed thresholds occurring during period 2 (0600-1100h). The greatest
percent change occurred in the BWBS and SBS in July. The BWBS had the greatest change in
FWI threshold exceedance in the first period (0000h-0500h) )in July, and SBS in the fourth
period (1800-2300h) in July and first (0000h-0500h) in August. Cooling trends occurred in CWH
in June and September but had a higher change to warmer conditions exceeding thresholds in the
third (1200h- 1700h) period in July.

31

Figure 8: BWBS BEC Zone: Percentage of observed FFMC values exceeding a threshold of ≥91 across four daily periods: 1) 0000h0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 1800-2300h by month for data grouped from 1990-2005 and 2006-2021. The secondary
Y-axis shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the more recent dataset (2006-2021).

32

Figure 9: CWH BEC Zone: Percentage of observed FFMC values exceeding a threshold of ≥91 across four daily periods: 1) 0000h0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 1800-2300h by month for data grouped from 1990-2005 and 2006-2021. The secondary
Y-axis shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the more recent dataset (2006-2021).

33

Figure 10: ICH BEC Zone: Percentage of observed FFMC values exceeding a threshold of ≥91 across four daily periods: 1) 0000h0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 1800-2300h by month for data grouped from 1990-2005 and 2006-2021. The secondary
Y-axis shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the more recent dataset (2006-2021).

34

Figure 11: IDF BEC Zone: Percentage of observed FFMC values exceeding a threshold of ≥91 across four daily periods: 1) 0000h0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 1800-2300h by month for data grouped from 1990-2005 and 2006-2021. The secondary
Y-axis shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the more recent dataset (2006-2021).

35

Figure 12: MS BEC Zone: Percentage of observed FFMC values exceeding a threshold of ≥91 across four daily periods: 1) 0000h0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 1800-2300h by month for data grouped from 1990-2005 and 2006-2021. The secondary
Y-axis shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the more recent dataset (2006-2021).

36

Figure 13: SBS BEC Zone: Percentage of observed FFMC values exceeding a threshold of ≥91 across four daily periods: 1) 0000h0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 1800-2300h by month for data grouped from 1990-2005 and 2006-2021. The secondary
Y-axis shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the more recent dataset (2006-2021).

37

Figure 14: BWBS BEC Zone: Percentage of observed ISI values exceeding a threshold of ≥8.7 across four daily periods: 1) 0000h0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 1800-2300h by month for data grouped from 1990-2005 and 2006-2021. The secondary
Y-axis shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the more recent dataset (2006-2021).

38

Figure 15: CWH BEC Zone: Percentage of observed ISI values exceeding a threshold of ≥8.7 across four daily periods: 1) 0000h0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 1800-2300h by month for data grouped from 1990-2005 and 2006-2021. The secondary
Y-axis shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the more recent dataset (2006-2021).

39

Figure 16: ICH BEC Zone: Percentage of observed ISI values exceeding a threshold of ≥8.7 across four daily periods: 1) 0000h0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 1800-2300h by month for data grouped from 1990-2005 and 2006-2021. The secondary
Y-axis shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the more recent dataset (2006-2021).

40

Figure 17: IDF BEC Zone: Percentage of observed ISI values exceeding a threshold of ≥8.7 across four daily periods: 1) 0000h0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 1800-2300h by month for data grouped from 1990-2005 and 2006-2021. The secondary
Y-axis shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the more recent dataset (2006-2021).

41

Figure 18: MS BEC Zone: Percentage of observed ISI values exceeding a threshold of ≥8.7 across four daily periods: 1) 0000h0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 1800-2300h by month for data grouped from 1990-2005 and 2006-2021. The secondary
Y-axis shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the more recent dataset (2006-2021).

42

Figure 19: SBS BEC Zone: Percentage of observed ISI values exceeding a threshold of ≥8.7 across four daily periods: 1) 0000h0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 1800-2300h by month for data grouped from 1990-2005 and 2006-2021. The secondary
Y-axis shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the more recent dataset (2006-2021).

43

Figure 20: BWBS BEC Zone: Percentage of observed FWI values exceeding a threshold of ≥19 across four daily periods: 1) 0000h0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 1800-2300h by month for data grouped from 1990-2005 and 2006-2021. The secondary
Y-axis shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the more recent dataset (2006-2021).

44

Figure 21: CWH BEC Zone: Percentage of observed FWI values exceeding a threshold of ≥19 across four daily periods: 1) 0000h0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 1800-2300h by month for data grouped from 1990-2005 and 2006-2021. The secondary
Y-axis shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the more recent dataset (2006-2021).

45

Figure 22: ICH BEC Zone: Percentage of observed FWI values exceeding a threshold of ≥19 across four daily periods: 1) 0000h0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 1800-2300h by month for data grouped from 1990-2005 and 2006-2021. The secondary
Y-axis shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the more recent dataset (2006-2021).

46

Figure 23: IDF BEC Zone: Percentage of observed FWI values exceeding a threshold of ≥19 across four daily periods: 1) 0000h0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 1800-2300h by month for data grouped from 1990-2005 and 2006-2021. The secondary
Y-axis shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the more recent dataset (2006-2021).

47

Figure 24: MS BEC Zone: Percentage of observed FWI values exceeding a threshold of ≥19 across four daily periods: 1) 0000h0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 1800-2300h by month for data grouped from 1990-2005 and 2006-2021. The secondary
Y-axis shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the more recent dataset (2006-2021).

48

Figure 25: SBS BEC Zone: Percentage of observed FWI values exceeding a threshold of ≥19 across four daily periods: 1) 0000h0500h, 2) 0600-1100h, 3) 1200h-1700h and 4) 1800-2300h by month for data grouped from 1990-2005 and 2006-2021. The secondary
Y-axis shows the percentage change of the FFMC values from the earlier dataset (1990-2005) to the more recent dataset (2006-2021).

49

CHAPTER 4 – DISCUSSION
Overall Findings
The first Intergovernmental Panel on Climate Change (IPCC) report was released in
1990 and confirmed the link between human activities and global climate change leading to an
increase in average temperatures. This analysis confirms significant changes towards a broader
warming trend associated with climate change has occurred across many BEC zones in British
Columbia from weather observations recorded between 2006-2021 compared to observations
from 1990-2005. Following the period used in this analysis, the global average surface
temperature in 2023 was the warmest on record (NOAA, 2024). British Columbia broke records
for the warmest summer with 84% of the province experiencing abnormally dry conditions, with
exceptional drought recorded across some regions. The Upper Fraser West, Nechako, and Skagit
basin all recorded their lowest annual precipitation since 1950. As a result, the province had low
streamflow, low soil moisture, and record-breaking wildfires.
Using annual suppression costs as one indicator of wildfire season severity, it is evident
that fire seasons are worsening. Annual wildfire suppression costs in British Columbia have
shown increasing variability and higher total costs in recent periods compared to historical
records (1970-2000 compared to 2001-2013; Stocks & Martell, 2016). Wildfire costs are
expected to continue increasing under climate change, with extreme seasons predicted to occur
every 1-2 years compared to every 10 years historically (Hope et al., 2016). Kirchmeier-Young
et al. (2019) attributed the extreme wildfire season of 2017 to anthropogenic climate change
noting extremes in warm and dry temperatures lead to extremes in fire behaviour. A study by
Jain et al. (2021) also found a trend of increasing global mean 95th percentile FWI and ISI in a
study spanning 1979-2020. Their study showed eight of the most extreme years on record
occurred between 2011-2020 which aligns with the trends shown in this study for British
Columbia.
Regional Findings
This study suggests that the SBS, IDF, and ICH zones are seeing more significant
changes compared to the other zones, however, a warming trend was detected across all zones in
the study with few findings showing a cooling trend. Differences between the zones show the
importance of not aggregating data for large areas with complex topography (Meyn et al., 2010).

50

BEC zones are the recommended climate unit for assessing climate change at a provincial scale
(DeLong et al., 2010). Although fire effects will vary based on localized variations in the fire
environment and vegetative community, these results can be applied to management from a
broad lens by providing a better understanding of regional climate shifts. Fire management
personnel working in different fire zones across the province can use these findings to
understand the changes occurring in their regions. Effectively adapting to climate change
requires an understanding of conditions, and this study provides insights on climate change
impacts across different regions of the province allowing for localized adaptation.
Meyn et al. 2010 found a decreasing trend in fire frequency and area burned in the ICH,
IDF, CWH, and SBS for a period spanning 1920-2000. They attributed this decrease to wetter
summers during this period. The significant warming trends found in this study may suggest a
shift away from wetter summers detected in the second period used for comparison (2006-2021
dataset). Findings from Parisien et al. (2023) found climate change to be a primary factor in
increased wildfire activity in British Columbia following a shift occurring around 2005 towards
increased evaporative demand and warming temperatures. These climatological changes were
found to be a factor in the extreme severity of four recent wildfire seasons (2017, 2018, 2021,
and 2023). The results of this study confirm a shift towards a warmer, drier climate across
British Columbia.
By comparing the results of shifts across the mean, median, and 95th percentile, it is
possible to understand different patterns over time. The mean is more sensitive to outliers and
extreme values and may be more likely to show changes based on short-term variations. The
median values are less sensitive to extremes and better reflect a shift towards more frequently
changed values which can indicate a change in the typical conditions over some time. Across the
mean, median, and 95th percentile, the SBS, IDF, and ICH all had the greatest number of
significant findings compared to the other zones. This shows these zones are exhibiting both
greater extremes, and an overall increase towards more frequent warm and dry conditions.
This study detected significant changes across both mean and median wind speeds across
almost all hours (>21 hours) tested in the ICH, IDF, and SBS zones. Comparisons in the 95th
percentile winds were rarely significant in all zones (<3 hours), except in the IDF (<10 hours).
This finding is contrary to expectations. Shaw and Miyawaki (2023) found fast winds increase

51

2.5 times more than average winds whereby fast winds get faster. Their study projects an
increase in upper-level jet stream winds with climate change. Increases in strong dry, surface
winds can have critical implications for wildfire growth and spread, exceeding suppression
capacity (Beverly et al. 2011, Potter and McEvoy 2021). The weather stations used in this study
rely on a 10-minute average wind speed recorded in the last 10 minutes of each hour. True
extreme values of gusty or erratic winds may have been missed or had their effect diluted due to
the nature of data collection.
Jain et al. (2021) found the strongest drivers of extreme FWI and ISI included relative
humidity, temperature, and vapour pressure deficit. Wind speed and precipitation were
responsible for fewer significant results in their study, ranging from only 11-13% of the global
area. In British Columbia, their results showed an increasing trend in extreme fire weather in the
south of the province due to increases in temperature outpacing changes in moisture.
The Chiodi et al. (2021) study spanned 1980-2019 and found the greatest overnight
vapour pressure deficits in foothill regions near arid plateaus across the western United States.
This study had the greatest number of significant findings in the SBS, IDF, and ICH zone, and
the largest magnitude of changes in the MS and IDF during June and July. The MS is on the
leeward side of the coast mountains with the SBS and IDF zones comprising the interior plateau.
The impacts of climate change may be exacerbated by the rain shadow effect leading to a greater
number of significant results in these BEC zones, and greater magnitude of changes. Simpson et
al. (2024) found the discrepancy between climate model predictions and observed near-surface
humidity was also greatest in arid and semi-arid regions. The models assume that the warmer air
will carry more moisture, however, observations show vapour remaining constant despite
warming temperatures. The resulting increase in vapour pressure deficit has implications for
wildland fire management. For every 1°C increase in temperature, a 15% increase in seasonal
precipitation is required to offset fuel drying (Flannigan et al. 2016).
Historical Fire Context
British Columbia has a history of notable fire events (Government of British Columbia,
2025). During the first time period used in this study, 1990-2005, a series of significant wildfire
events occurred in the interior. In 1994, the Penticton wildfire burned 5,500 hectares and
damaged 18 structures. The Salmon Arm Wildfire of 1998 burned 6,000 hectares and damaged

52

40 buildings. During 2003, two extreme interface fires occurred: the Okanagan Mountain Park
Wildfire, and the McClure Wildfire. Both of these fires resulted in mass evacuations, significant
loss of infrastructure (homes, businesses and trestle bridges) and grew to final sizes exceeding
25,000 hectares.
During the second time period, 2006-2021, the historical fire records in British Columbia
demonstrate substantially more challenging conditions (Government of British Columbia, 2025).
Reviewing season summaries from recent years shows far greater area burned, financial impact
through suppression costs and social impact of structure loss and evacuations. Several notable
events are mentioned across the years 2009-present compared to relatively few recorded fires of
note in the historical records. In 2017, the area burned broke records, exceeding 1.2 million
hectares. Notable fires included the Elephant Hill Wildfire in the Kamloops Fire Zone, which
grew to a over 1,000 hectares within the first 24 hours of ignition, and a final size of 191,865
hectares. In the Cariboo Fire Zone, the plateau complex which resulted from 20 separate
lightning fires growing together, ended at a final combined area of 545,131 hectares across the
Chilcotin Plateau. In 2018, area burned records were exceeded once again across the province
with a total area of 1,345,284 hectares. The greatest area burned occurred in the Northwest,
Coastal and Prince George fire centers. In 2021 the fire season was exacerbated by belowaverage fall and winter precipitation and an extended and historic record-breaking heat wave
spanning from mid-June to mid-July. Multiple lightning storms resulted in new fire starts and
hot, dry, windy weather facilitated substantial fire growth into August. The Kamloops, Cariboo
and Prince George Fire centers had the greatest area burned.
Findings from this study showed an increase in the frequency of spread event thresholds
being exceeded. All zones had an increase in the percentage of observations exceeding
thresholds in July and August between the hours of 1200h-2300h. Although fire effects and fire
behaviour will vary within and between each zone, this increased frequency of values above
threshold shows conditions are changing towards more volatile conditions. Fire history recorded
during the second period shows a reality with increasing area burned and more challenging
suppression conditions requiring substantial resources.
This research is focused on historical data spanning 1990-2021. Since then, British
Columbia recorded the most destructive wildfire season in history. In 2023, 2.84 million

53

hectares burned across the province (Government of British Columbia, 2025). The area burned
was 10 times greater than the 20-year average annual area burned and was twice as high as the
previous record-breaking season which occurred in 2018 with a total of 1.35 million hectares.
Notable fire events included the Donnie Creek Wildfire which became the largest wildfire ever
recorded in British Columbia’s history (Government of British Columbia, 2023), and the
Newcastle Creek Wildfire on Vancouver Island which was observed to be burning as deep as 1
meter below ground (Government of British Columbia, 2025). The Donnie Creek fire burnt
primarily in the BWBS zone. In this study, the BWBS zone showed less warming compared to
other zones, but had the greatest percent change in the FFMC spread event thresholds between
time periods, particularly between 0600h and 1100h in July, August, and September. This zone,
which previously had very few hours above thresholds, is showing a large shift. This means
areas that previously experienced a long fire return interval with infrequent fires, may be more
susceptible to burning in future climates.
The Newcastle Creek fire was another fire of note in 2023 which occurred within the
CWH zone. Although this fire did not exhibit large growth, the fire effects were severe due to
the extent of deep burning. Deeply burning fires are ecologically damaging and require
exceptional time and expense to extinguish. This BEC zone had several significant results across
June, July, and August with increasing temperature and windspeed and a decrease in RH. These
sustained periods of warmer, drier weather are unusual in a coastal environment, but likely to
occur more frequently in a future climate.
Timing
This study is the first to review changes across space and time in British Columbia with
the intent of informing operational wildfire management decisions. The analysis is highly
detailed providing results for each BEC zone by each month through the primary summer fire
season in British Columbia, and by each hour during that month. This detailed approach can be
used to determine what areas of the province may require greater resourcing on a monthly scale,
and inform the most beneficial operational shift schedule for resources working on active fires.
Typical scheduling overlaps with peak burn period, where resources often need to be stood down
from action due to safety concerns. With the increasing fire intensity and severity associated
with climate change, fire behaviour during peak burn often cannot be controlled. At wildfire

54

intensity class 4 (4,000-10,000 kW/m) fire are extremely difficult to control and often have
prolific spot fires 500-1000m ahead of the fire front (Tedim et al. 2018). At higher intensity
classes (>10,000kW/m), control becomes impossible, and spot fires can start several kilometers
ahead of the main fire front (Tedim et al. 2018). Prioritizing sustained action suppression
resources during time periods where fires are easier to control could result in better outcomes on
large, challenging wildfires.
Monthly
This study showed differences in the strength of warming relationships across June, July,
August, and September. The most significant warming trends appeared during July, followed by
August. Other studies typically take a seasonal approach to understanding fire weather, for
example, June-August (Meyn, 2010), May-August (Xiao and Zhuang, 2007) and Spring (March,
April, May), Summer (June, July, August) and Fall (September, October, November) (Wang et al.
2023). Understanding changes seasonally is important, however, the approach used in this study
provides further detail to fire management agencies and can be used to understand when
resources may be challenged and where and when to prepare for additional personnel or
equipment needs. The challenging 2023 fire season in British Columbia was exacerbated by
heatwaves and record-breaking warm and dry temperatures across May through September
(Government of British Columbia, 2024).
Hourly
The largest significant changes were detected during the third period (1200h-1700h)
across all six BEC zones in this study. This suggests the magnitude of changes caused by
climate change is greatest during the afternoon hours for the zones and months analyzed.
However, the zones differed with the BWBS and IDF having the second highest magnitude in the
second period (0600-1100), whereas the MS and SBS had the second highest magnitude in the
fourth period (1800h-2300h). These differences show how climate change is affecting areas
across British Columbia differently. The implications may result in fires exhibiting greater
growth during the peak burn period across BWBS and IDF due to antecedent warming of fuels
earlier in the day. In the MS and SBS, the peak burn period could extend later into the evening.
The results could be applied regionally to inform and schedule suppression resources during
periods where they are likely to make the greatest impact.

55

The understanding that fire behaviour reduces overnight has been challenged in recent
years (Luo et al. 2024). Following reports from wildland personnel on observations of overnight
fire spread, Chiodi et al. (2021) conducted a study that found prominent increases in overnight
vapor pressure deficits (VPD), driven by increased temperature and decreased vapour pressure
where the observed overnight VPD exceeded climate change model predictions. The change in
nighttime fire activity has been outpacing daytime fire activity when analyzed through MODIS
observations across the conterminous United States (Freeborn et al., 2022). Luo et al. (2024)
also found an increase in fire behaviour associated with overnight burning attributed to drought
conditions allowing fires to burn over a greater period and result in larger fire sizes. This study
showed spread event thresholds were more likely to be exceeded in the later dataset (2006-2021)
compared to the past conditions (1990-2005). A study in India reviewing hourly weather data
has also shown warming trends during night and early mornings and an overall increase in
warming hours leading to increased duration of heatwaves (Rohini et al., 2022).
Limitations & Future Research
This study relied on data from a network of weather stations maintained by the British
Columbia Wildfire Service. Between the dates used in this study (1990-2021), many stations
would have undergone routine maintenance and instrumentation changes which may have
impacted the accuracy of observations. Instrument upgrades and maintenance are unlikely to
have had a significant influence on the results. A second limitation of the dataset was the
required cleaning for quality control which is important when working with a large dataset. It is
possible that some true extremes were eliminated from the dataset and erroneously classified as
outliers, however, the conservative approach to using a z-score of 4 makes it unlikely that the
results were impacted. Finally, eight BEC zones were omitted from the analysis due to
insufficient weather station data. Future research could examine the ERA5 reanalysis dataset
which includes hourly climate data across the entire globe at a resolution of 0.25° x 0.25° from
1940 to the present (Hersbach et al. 2023). This dataset relies on the physical modeling of
climate variables to fill gaps where observations are lacking to create a comprehensive dataset
across the earth's surface.
A notable exclusion from this study was data from stations in the Ponderosa Pine BEC
zone. The PP zone includes weather stations which had record breaking temperatures during the

56

2021 heat dome which heavily impacted British Columbia. Lytton, British Columbia recorded a
temperature of 49.6°C on June 29, 2021 and broke records for the highest temperature ever
recorded in Canada. Other stations in the PP BEC zone with record breaking highs include
Kamloops which recorded 47.3°C, Kelowna with 45.7°C, and Osoyoos with 46.9°C. The
maximum temperature value used in this study was 42°C. This shows the importance of
expanding this assessment across all BEC zones in British Columbia. This study showed
significant warming trends across the province, but the situation may be more severe in zones
which were unable to be assessed through station data alone.
Variables not included in this study include hourly rainfall, the Duff Moisture Code
(DMC), Drought Code (DC), and Build-up Index (BUI). Hourly rainfall is accounted for
through calculations of the FFMC, ISI, and FWI so it was indirectly assessed through focusing
on those variables. As a stand-alone variable for analysis at the hourly scale in this study it was
omitted due to the spatial variation in weather systems bringing precipitation. The DC, DMC,
and BUI are not significantly associated with spread events (Wang et al. 2023). Drought is a
critical fire weather element associated with extreme fire behaviour, especially when paired with
strong, unstable winds and low relative humidity (Werth et al., 2016); however, droughts are a
consequence of weather patterns occurring over relatively long durations which would be
inappropriate to study at the hourly scale used in this study.
Thresholds in this study were applied equally across all BEC zones. Wang et al. 2023
found spread events can occur under less severe conditions both as latitude increases and
throughout coastal areas. Their study, which used Canadian ecozones as analysis units found
differences across the spread event thresholds between coastal areas (Pacific Maritime), interior
British Columbia (Montane Cordillera), and Northern British Columbia (Boreal Ecozones). For
all three spread event indices (FFMC, ISI, and FWI), the greatest values were needed in the
Montane Cordillera zone and included an FFMC ≥ 92, ISI ≥ 11, and FWI ≥ 26. The Pacific
maritime ecozone exceeded spread event thresholds at FFMC of 90-90.9, ISI 5-5.9, and FWI
range of 9-13.9. The boreal regions had FFMC of 88-88.9, ISI range of 6-9.9, and FWI of 1623.9. There is a large variation in these spread event thresholds applied to Canadian Ecozones.
Future work could consider an approach similar to the Wang et al. (2023) Canada-wide study to
better inform spread event thresholds by BEC zone throughout British Columbia. A study

57

conducted in Alberta found an ISI ≥ 9 to be an important factor in late spring fires (Tymstra et
al., 2021). Understanding regionally specific thresholds can help fire managers make more
informed decisions. When assessing all three indices (FFMC, ISI, and FWI) the single best
indicator of spread events appears to be FFMC (Wang et al. 2023). Results from this study are
valuable to show the trends and magnitude of change in the variables analyzed.

58

CHAPTER 5 – CONCLUSION
This study examined differences in weather at an hourly scale across British Columbia and
confirmed that climate change has already led to significant warming between two historical
periods (1990-2005 and 2006-2021) with subsequent spread event thresholds being more likely
to be exceeded. Geographically, the BEC zones with the most significant findings included the
SBS, IDF and ICH. The varying results between zones highlights the nuanced effect of climate
change and the regional implications with differences in daily impacts, magnitude and rate of
change. The results can be used at a finer management scale to understand what climate change
impacts have already been identified and inform future management decisions. For example,
these results could inform personnel and suppression equipment needs throughout a fire season
by month and guide changes to the standard daily firefighting operational periods for greater
safety and effectiveness in wildfire management. A changing climate is resulting in spread event
thresholds being surpassed throughout more hours in a day, meaning the conventional
understanding of peak burn will not be the only challenging operational period, and fires will be
less likely to slow or stop spread overnight.
Compared to other climate-based studies, this research was unique in analyzing weather data
at an hourly scale to understand how a changing climate may be impacting the fire environment
through the diurnal cycle using BEC zones as analysis units. The comparison of these two
sixteen-year periods provides an understanding of how climate change impacts are materializing
across the province of British Columbia and can provide fire management agencies with
information on how anticipated fire behaviour and growth may be impacted with the nuance of
time and space taken into consideration. Future research could identify spread event thresholds
for each unique BEC zone and use the gridded ERA-5 dataset to fill gaps in the BEC zones
which were excluded from this study due to insufficient station observations. Similar studies
could be expanded to other provinces and territories across Canada, using regionally appropriate
analysis units.

59

REFERENCES
Abatzoglou, J. T., Williams, A. P., & Barbero, R. (2019). Global emergence of anthropogenic
climate change in fire weather indices. Geophysical Research Letters, 46(1), 326–336.
https://doi.org/10.1029/2018GL080959
Agee, J. K. (1996). The influence of forest structure on fire behavior. In Forest Vegetation
Management Conference (pp. 52–68).
Balch, J. K., Abatzoglou, J. T., Joseph, M. B., Koontz, M. J., Mahood, A. L., McGlinchy, J.,
Cattau, M. E., & Williams, A. P. (2022). Warming weakens the nighttime barrier to global fire.
Nature, 602(7897), 442–448. https://doi.org/10.1038/s41586-021-04325-1
Beverly, J. L., Flannigan, M. D., Stocks, B. J., & Bothwell, P. (2011). The association between
Northern Hemisphere climate patterns and interannual variability in Canadian wildfire activity.
Canadian Journal of Forest Research, 41, 2193–2201. https://doi.org/10.1139/X11-131
Centre for Forest Conservation Genetics. (n.d.). Subzone/variant climate data [Data set]. UBC
Centre for Forest Conservation Genetics. Retrieved November 23, 2024, from
https://cfcg.forestry.ubc.ca/resources/cataloguing-in-situ-genetic-resources/subzonevariantclimate-data/
Chiodi, A. M., Potter, B. E., & Larkin, N. K. (2021). Multi-decadal change in western US
nighttime vapor pressure deficit. Geophysical Research Letters, 48(15).
https://doi.org/10.1029/2021GL092830
Countryman, C. M. (1972). The fire environment concept (No. 15). U.S. Department of
Agriculture, Forest Service, Pacific Southwest Forest and Range Experiment Station.
DeLong, S. C., Griesbauer, H., Mackenzie, W., & Foord, V. (2010). Corroboration of
biogeoclimatic ecosystem classification climate zonation by spatially modelled climate data. BC
Journal of Ecosystems and Management, 10(3), 49–64.
www.forrex.org/publications/jem/ISS52/vol10_no3_art7.pdf
Easterling, D. R., Horton, B., Jones, P. D., Peterson, T. C., Karl, T. R., Parker, D. E., Salinger, M.
J., Razuvayev, V., Plummer, N., Jamason, P., & Folland, C. K. (1997). Maximum and minimum
temperature trends for the globe. Science, 277, 364–367.
https://doi.org/10.1126/science.277.5324.364
Ellis, T. M., Bowman, D. M. J. S., Jain, P., Flannigan, M. D., & Williamson, G. J. (2021). Global
increase in wildfire risk due to climate-driven declines in fuel moisture. Global Change Biology.
https://doi.org/10.1111/gcb.16006
Flannigan, M. D., Krawchuk, M. A., De Groot, W. J., Wotton, B. M., & Gowman, L. M. (2009).
Implications of changing climate for global wildland fire. International Journal of Wildland Fire,
18(5), 483–507. https://doi.org/10.1071/WF08187

60

Flannigan, M. D., Wotton, B. M., Marshall, G. A., De Groot, W. J., Johnston, J., Jurko, N., &
Cantin, A. S. (2016). Fuel moisture sensitivity to temperature and precipitation: Climate change
implications. Climatic Change, 134(1–2), 59–71. https://doi.org/10.1007/s10584-015-1521-0
Forestry Canada Fire Danger Group. (1992). Development and structure of the Canadian Forest
Fire Behavior Prediction System. Forestry Canada. Forestry Canada, Headquarters, Fire Danger
Group and Science and Sustainable Development Directorate, Ottawa. Information Report ST-X3.
Freeborn, P. H., Jolly, W. M., Cochrane, M. A., & Roberts, G. (2022). Large wildfire-driven
increases in nighttime fire activity observed across CONUS from 2003–2020. Remote Sensing of
Environment, 268. https://doi.org/10.1016/j.rse.2021.112777
FTS, Inc. (2015). FTS fixed RAWS remote automated weather station PDF.
https://ftsinc.com/wp-content/uploads/2015/08/FTS-Fixed-RAWS-brochure_web.pdf
Government of British Columbia. (2025). Wildfire season summary. BC Wildfire Service.
Retrieved March 15, 2025, from https://www2.gov.bc.ca/gov/content/safety/wildfirestatus/about-bcws/wildfire-history/wildfire-season-summary
Government of British Columbia. (2023). How 2023 compares to previous wildfire seasons. BC
Wildfire Service. Retrieved November 11, 2024, from https://blog.gov.bc.ca/bcwildfire/how2023-compares-to-previous-seasons/
Guindon, L., Gauthier, S., Manka, F., Parisien, M.-A., Whitman, E., Bernier, P., Beaudoin, A.,
Villemaire, P., & Skakun, R. (2021). Trends in wildfire burn severity across Canada, 1985 to
2015. https://doi.org/10.23687/b1f61b7e-4ba6-4244-bc79-c1174f2f92cd
Hersbach, H., Bell, B., Berrisford, P., Biavati, G., Horányi, A., Muñoz Sabater, J., Nicolas, J.,
Peubey, C., Radu, R., Rozum, I., Schepers, D., Simmons, A., Soci, C., Dee, D., & Thépaut, J.-N.
(2023). ERA5 hourly data on single levels from 1940 to present. Copernicus Climate Change
Service (C3S) Climate Data Store (CDS). https://doi.org/10.24381/cds.adbb2d47
Hope, E. S., McKenney, D. W., Pedlar, J. H., Stocks, B. J., & Gauthier, S. (2016). Wildfire
suppression costs for Canada under a changing climate. PLoS ONE, 11(8).
https://doi.org/10.1371/journal.pone.0157425
Intergovernmental Panel on Climate Change. (1990). Climate change: The IPCC scientific
assessment. Cambridge University Press.
Intergovernmental Panel on Climate Change. (2023). Summary for policymakers. In Climate
change 2021: The physical science basis. Contribution of Working Group I to the Sixth
Assessment Report of the Intergovernmental Panel on Climate Change. Retrieved from
https://www.ipcc.ch/report/ar6/syr/
Jain, P., Castellanos-Acuna, D., Coogan, S. C. P., Abatzoglou, J. T., & Flannigan, M. D. (2021).
Observed increases in extreme fire weather driven by atmospheric humidity and temperature.
Nature Climate Change. https://doi.org/10.1038/s41558-021-01224-1

61

Jolly, W. M., Cochrane, M. A., Freeborn, P. H., Holden, Z. A., Brown, T. J., Williamson, G. J., &
Bowman, D. M. J. S. (2015). Climate-induced variations in global wildfire danger from 1979 to
2013. Nature Communications, 6. https://doi.org/10.1038/ncomms8537
Kirchmeier-Young, M. C., Gillett, N. P., Zwiers, F. W., Cannon, A. J., & Anslow, F. S. (2019).
Attribution of the influence of human-induced climate change on an extreme fire season. Earth's
Future, 7(1), 2–10. https://doi.org/10.1029/2018EF001050
Lawson, B. D., & Armitage, O. B. (2008). Weather guide for the Canadian forest fire danger
rating system. Natural Resources Canada, Canadian Forest Service, Northern Forestry Centre,
Edmonton, Alberta.
Luo, K., Wang, X., de Jong, M., & Flannigan, M. (2024). Drought triggers and sustains overnight
fires in North America. Nature, 627(8003), 321–327. https://doi.org/10.1038/s41586-024-070285
Mair, P., & Wilcox, R. (2020). Robust statistical methods in R using the WRS2 package.
Behavior Research Methods, 52. https://doi.org/10.3758/s13428-019-01246-w
Meidinger, D., & Pojar, J. (1991). Ecosystems of British Columbia. BC Ministry of Forests,
Research Branch, Special Report Series 6.
Meyn, A., Taylor, S. W., Flannigan, M. D., Thonicke, K., & Cramer, W. (2010). Relationship
between fire, climate oscillations, and drought in British Columbia, Canada, 1920-2000. Global
Change Biology, 16(3), 977–989. https://doi.org/10.1111/j.1365-2486.2009.02061.x
NOAA National Centers for Environmental Information. (2024, January). Monthly global
climate report for annual 2023. National Centers for Environmental Information. Retrieved
October 22, 2024, from https://www.ncei.noaa.gov/access/monitoring/monthlyreport/global/202313
Parisien, M. A., Barber, Q. E., Bourbonnais, M. L., Daniels, L. D., Flannigan, M. D., Gray, R.
W., Hoffman, K. M., Jain, P., Stephens, S. L., Taylor, S. W., & Whitman, E. (2023). Abrupt,
climate-induced increase in wildfires in British Columbia since the mid-2000s. Communications
Earth and Environment, 4(1). https://doi.org/10.1038/s43247-023-00977-1
Podur, J., & Wotton, B. M. (2011). Defining fire spread event days for fire-growth modelling.
International Journal of Wildland Fire, 20(4), 497–507. https://doi.org/10.1071/WF09001
Potter, B. E., & McEvoy, D. (2021). Weather factors associated with extremely large fires and
fire growth days. Earth Interactions, 25, 160–176. https://doi.org/10.1175/EI-D-21-0008.1
R Core Team. (2022). R: A language and environment for statistical computing. R Foundation
for Statistical Computing, Vienna, Austria. https://www.R-project.org/
Rohini, P., Rajeevan, M., & Soni, V. K. (2022). Trends in diurnal variation of surface air
temperatures over India during hot weather (April–June) season. Natural Hazards, 114(2), 1815–
1827. https://doi.org/10.1007/s11069-022-05447-z

62

Rothermel, R. C., Hartford, R. A., & Chase, C. H. (1994). Fire growth maps for the 1988 Greater
Yellowstone Area fires (Gen. Tech. Rep. INT-304). USDA Forest Service, Intermountain
Research Station.
Shaw, T. A., & Miyawaki, O. (2023). Fast upper-level jet stream winds get faster under climate
change. Nature Climate Change, 14(1), 61–67. https://doi.org/10.1038/s41558-023-01884-1
Simpson, I. R., McKinnon, K. A., Kennedy, D., Lawrence, D. M., Lehner, F., & Seager, R. I.
(2024). Observed humidity trends in dry regions contradict climate models. Proceedings of the
National Academy of Sciences. https://doi.org/10.1073/pnas
Stocks, B. J., & Martell, D. L. (2016). Forest fire management expenditures in Canada: 19702013 (Vol. 92).
Sullivan, A. L. (2017). Inside the Inferno: Fundamental processes of wildland fire behaviour:
Part 1: Fundamental processes of wildland fire behaviour. Current Forestry Reports, 3(2), 150–
171. https://doi.org/10.1007/s40725-017-0058-z
Taylor, S. W., & Alexander, M. E. (2003). Considerations in developing a national forest fire
danger rating system. In Proceedings of the XII World Forestry Congress (Quebec City, QC).
Taylor, S. W., & Alexander, M. E. (2006). Science, technology, and human factors in fire danger
rating: The Canadian experience. International Journal of Wildland Fire, 15, 121–135.
https://doi.org/10.1071/WF05021
Tedim, F., Leone, V., Amraoui, M., Bouillon, C., Coughlan, M., Delogu, G., Fernandes, P.,
Ferreira, McCaffrey, S., McGee, T., Parente, J., Paton, D., Pereira, M., Ribeiro, L., Viegas, D., &
Xanthopoulos, G. (2018). Defining Extreme Wildfire Events: Difficulties, Challenges, and
Impacts. Fire. 1. 9. 10.3390/fire1010009.
Tymstra, C., Jain, P., & Flannigan, M. D. (2021). Characterisation of initial fire weather
conditions for large spring wildfires in Alberta, Canada. International Journal of Wildland Fire,
30(11), 823–835. https://doi.org/10.1071/WF21045
Van Wagner, C. E. (1983). Fire behaviour in northern conifer forests and shrublands. In R. A.
Wein & D. A. MacLean (Eds.), The role of fire in northern circumpolar ecosystems (pp. 65–80).
John Wiley & Sons Ltd.
Van Wagner, C. E., & Pickett, T. L. (1985). Equations and FORTRAN program for the Canadian
forest fire weather index system (For. Tech. Rep. No. 33). Environment Canada, Canadian
Forestry Service, Petawawa National Forestry Institute.
Van Wagner, C. E. (1987). Development and structure of the Canadian Forest Fire Weather Index
System (For. Tech. Rep. 35). Canadian Forest Service, Ottawa, ON.
Wang, X., Oliver, J., Swystun, T., Hanes, C. C., Erni, S., & Flannigan, M. D. (2023). Critical fire
weather conditions during active fire spread days in Canada. Science of the Total Environment,
869. https://doi.org/10.1016/j.scitotenv.2023.161831

63

Wang, X., Parisien, M.-A., Flannigan, M.D., Parks, S.A., Anderson, K.R., Little, J.M. and Taylor,
S.W. (2014), The potential and realized spread of wildfires across Canada. Glob Change Biol,
20: 2518-2530. https://doi-org.ezproxy.tru.ca/10.1111/gcb.12590
Wang, X., Thompson, D. K., Marshall, G. A., Tymstra, C., Carr, R., & Flannigan, M. D. (2015).
Increasing frequency of extreme fire weather in Canada with climate change. Climatic Change,
130(4), 573–586. https://doi.org/10.1007/s10584-015-1375-5
Werth, P. A., Potter, B. E., Alexander, M. E., Cruz, M. G., Clements, C. B., Finney, M. A.,
Forthofer, J. M., Goodrick, S. L., Hoffman, C., Jolly, W. M., McAllister, S. S., Ottmar, R. D., &
Parsons, R. A. (2016). Synthesis of knowledge of extreme fire behavior: Volume 2 for fire
behavior specialists, researchers, and meteorologists (Gen. Tech. Rep.). USDA Forest Service
Pacific Northwest Research Station. http://www.fs.fed.us/pnw
Wotton, B. M. (2009). Interpreting and using outputs from the Canadian Forest Fire Danger
Rating System in research applications. Environmental and Ecological Statistics, 16(2), 107–
131. https://doi.org/10.1007/s10651-007-0084-2
Wotton, B. M., Flannigan, M. D., & Marshall, G. A. (2017). Potential climate change impacts on
fire intensity and key wildfire suppression thresholds in Canada. Environmental Research
Letters, 12(9). https://doi.org/10.1088/1748-9326/aa7e6e
Xiao, J., & Zhuang, Q. (2007). Drought effects on large fire activity in Canadian and Alaskan
forests. Environmental Research Letters, 2(4). https://doi.org/10.1088/1748-9326/2/4/044003

64

APPENDIX A – RESULTS OF MEAN, MEDIAN AND 95th PERCENTILE ANALYSIS –
1990-2005 COMPARED TO 2006-2021

The following figures visually represent the diurnal data curves for mean, median, and 95th
percentile hourly values with significant results shown by an asterisk.

65

66

67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

APPENDIX B – MAGNITUDE OF CHANGE IN RESULTS OF MEAN, MEDIAN AND 95th
PERCENTILE ANALYSIS – 1990-2005 COMPARED TO 2006-2021

The following figures visually represent the change between the mean, median, and 95th
percentile hourly values with significant results shown by an asterisk.

84

85

86

87

88

89

90

91

92

93

94

95

96

97

98

99

100

101