Conference PaperPDF Available

Are High Severity Fires Increasing in Southern Australia?

Authors:

Abstract and Figures

Wildfires worldwide are becoming more frequent but are they also becoming more severe? Here we used remotely sensed burn-severity data from wildfires in Victoria, southeastern Australia to address that question. We selected 162 wildfires of more than 1000 ha that occurred over the past 30 years across a wide range of forest types. Spectral indices derived from Landsat pre-and post-fire imagery were used to map fire severity. Our results show a significant increase in the absolute and proportional area burnt by high-severity fire over the last three decades. This study demonstrates that wildfires in the temperate forests of southern Australia are becoming more severe. Such change in fire regimes may have critical consequences for the sustainability and resilience of the studied forests. B. N. Tran, M. A. Tanase, L. T. Bennett and C. Aponte, "Are High Severity Fires Increasing in Southern Australia?," IGARSS 2020 - 2020 IEEE International Geoscience and Remote Sensing Symposium, Waikoloa, HI, USA, 2020, pp. 4630-4633, doi: 10.1109/IGARSS39084.2020.9324121.
Content may be subject to copyright.
ARE HIGH SEVERITY FIRES INCREASING IN SOUTHERN AUSTRALIA?
Bang Nguyen Tran 1,2, Mihai A. Tanase 3,4, Lauren T. Bennett 1, Cristina Aponte 1,4
1 School of Ecosystem and Forest Sciences, University of Melbourne, Australia
2 Faculty of Environment, Vietnam National University of Agriculture, Vietnam
3 Department of Geology, Geography and Environment, University of Alcala, Spain
4 National Institute for Research and DeveORSPHQWLQ)RUHVWU\³0DULQ'UDFHD´Romania
ABSTRACT
Wildfires worldwide are becoming more frequent but are
they also becoming more severe? Here we used remotely
sensed burn-severity data from wildfires in Victoria,
southeastern Australia to address that question. We selected
162 wildfires of more than 1000 ha that occurred over the
past 30 years across a wide range of forest types. Spectral
indices derived from Landsat pre- and post- fire imagery
were used to map fire severity. Our results show a
significant increase in the absolute and proportional area
burnt by high-severity fire over the last three decades. This
study demonstrates that wildfires in the temperate forests of
southern Australia are becoming more severe. Such change
in fire regimes may have critical consequences for the
sustainability and resilience of the studied forests.
Index Terms² Burn severity, eucalypt, Landsat,
spectral index, temperate forest
1. INTRODUCTION
Wildfires are among the most important disturbances
worldwide with significant impacts on the carbon cycle,
biodiversity and human wellbeing [1, 2]. It has been shown
that forest wildfires are becoming more frequent and, in
some cases, also more extensive [3, 4], a change that has
been linked to climate change [4, 5]. Whether wildfires are
also becoming more severe is still unclear.
Fire severity is an important attribute of a fire regime. High-
severity forest fires can lead to carbon losses, elevated tree
mortalities, greater vulnerability to subsequent fires and
conversion of forest to less productive states [6-10].
Analyzing the retrospective temporal trend of high-severity
fire is therefore essential to better understand shifts in fire
regimes and how novel regimes might affect forest
landscapes.
2. MATERIALS & METHODS
2.1. Study area & fire history
This study focused on wildfires that occurred in the state of
Victoria, southeastern Australia over the past 30 years.
Using the fire history records provided by the Department of
Environment, Land, Water & Planning (DELWP) of
Victoria, we selected wildfires with a minimum area of 1000
ha that burnt between 1987 and 2017, totaling 162 wildfires
(Fig. 1). These fires occurred across seven bioregions of
varying soil and climate characteristics and affected a range
of temperate eucalypt forests differing in species
composition, structure, and regeneration strategy.
MDD
VM
SEH
VM
VVP
SEC
SCP
VM
VM
AA
p
0 50 10025
Kilometers
Legend
)LUHFHQWURLGV %LRUHJLRQ
$$
0''
6&3
6(&
6(+
90
993
ǻǼ
ǻǼ
Figure 1. Map of the State of Victoria, southeastern
Australia. (i) Victoria highlighted (grey) in the map of
Australia; (ii) Locations of study areas within the state of
Victoria in south-eastern Australia. Dots represent the center
of each of the 162 wildfires analysed in this study. Colours
relate to different bioregions.
 
IGARSS 2020 - 2020 IEEE International Geoscience and Remote Sensing Symposium | 978-1-7281-6374-1/20/$31.00 ©2020 IEEE | DOI: 10.1109/IGARSS39084.2020.9324121
Authorized licensed use limited to: University of Melbourne. Downloaded on February 22,2021 at 06:57:17 UTC from IEEE Xplore. Restrictions apply.
2.2 Fire severity mapping and data analyses
Fire severity mapping was conducted using Landsat TM,
ETM+ and Landsat 8 images acquired within two months
before and after the fires to minimize differences in forest
phenology and general atmospheric conditions at the time of
acquisition. The images were selected and obtained through
the US Geological Survey (USGS) EarthExplorer as higher-
level surface reflectance products for each fire. The images
were masked for clouds and shadows using the Fmask
algorithm.
Four spectral indices (NBR, NDVI, NDWI, and MSAVI)
and their temporal differences (i.e. delta versions, which
calculate the change between pre-fire and post-fire spectral
index values) were computed for each of the 162 wildfires.
These indices are commonly used to assess fire severity [11,
12] and were identified in our previous study [13] as optimal
spectral indices for fire-severity mapping for the studied
forest types.
Figure 2. Fire severity map as a result of random forest
model for Kinglake area in 2009.
Fire severity was mapped using a random forest model
previously developed by the authors [14] for the same area.
Briefly, the model used a training dataset of 2238 reference
plots encompassing 13 forest types across Victoria. Plot fire
severity was assessed in situ or was visually interpreted on
very high resolution orthophotos as Unburnt: less than 1%
of eucalypt and non-eucalypt crowns scorched; Low
severity: light scorch of 1 - 35% of eucalypt and non-
eucalypt crowns; Moderate severity: 30 - 65% of eucalypt
and non-eucalypt crowns scorched; or High severity: 70 -
100% of eucalypt and non-eucalypt crowns burnt [15].
Accuracy of high severity mapping on an independent
validating dataset across all 13 forest types was very high,
with 0.06 and 0.18 commission and omission error
respectively [14].
Changes in the area burnt by high-severity fire (total and
proportional) over time were examined using linear models
with individual fires as the sampling unit and fire size as a
covariate. Data were transformed when needed to meet
assumptions of normality. All statistical tests were
conducted in the statistical programming language R [16].
3. RESULTS & DISCUSSION
Fire history records of the past three decades in Victoria
show a significant increase in the number of wildfires larger
than 1000 ha but no significant increase in their size (Fig.
3). In contrast to the trend in the overall fire size, our
analysis of the high-severity fire mapping indicates that
within those fires greater than 1000 ha both the area burnt
by high-severity fire and the proportion that high-severity
area represents have increased significantly (p<0.05 and
p<0.1 respectively) from 1987 to 2017 (Fig. 4).
Figure 3. Temporal trends over the past three decades in the
number of wildfires greater than 1000 ha (upper panel) and
the fire size of such fires (lower panel).
The increase in high-severity fires in south-eastern Australia
is of great concern. The impacts of high severity fires in
temperate Eucalyptus forest include soil erosion, reduction

Authorized licensed use limited to: University of Melbourne. Downloaded on February 22,2021 at 06:57:17 UTC from IEEE Xplore. Restrictions apply.
in carbon sequestration and carbon carrying capacity,
destruction of habitat, and loss of social and aesthetical
values [17-20]. The magnitude of such impacts depends on
the forest species composition, their regeneration strategy,
the occurrence of repeated fires at short intervals, and the
environmental conditions during post-fire regeneration [21-
23]. At landscape scales, nevertheless, the impact will
ultimately depend on the area burnt by high-severity fires.
Figure 4. Temporal trends in the area burnt by high severity
fire presented as total area (upper panel) and percentage of
burnt area (lower panel). Dots represent each of the 162
wildfires examined in this study.
Changes in the area of high-severity fire like those described
here have been predicted to occur as a result of climate
change for decades [24-26]. If that trend is to continue, as
expected under current climate change projections, it is
likely to result in large-scale changes in key structural
attributes of even the most fire-tolerant forests.
Our results are of critical importance for forest managers
and policy makers to acknowledge and act on the described
changes in the fire regimes to ensure that Victorian
temperate forests continue to provide us with ecosystem
services essential for our wellbeing.
5. ACKNOWLEDGEMENTS
The authors would like to acknowledge the financial support
of the Melbourne Research Scholarship program, the
Vietnam International Education Cooperation Department
(VIED) scholarship, and the Integrated Forest Ecosystem
Research program, supported by the Victorian Department
of Environment, Land, Water and Planning. We also
acknowledge the support of many staff and students from
the School of Ecosystem and Forest Sciences at the
University of Melbourne.
6. REFERENCES
[1] L. Giglio, J.T. Randerson and G.R. van der Werf:
"Analysis of daily, monthly, and annual burned area using
WKH IRXUWKဨJHQHUDWLRQ JOREDO ILUH HPLVVLRQV GDWDEDVH
(GFED4)". Journal of Geophysical Research:
Biogeosciences 2013, 118(1):317-328.
[2] C.A. Harper, W.M. Ford, M.A. Lashley, C.E. Moorman
and M.C. Stambaugh: "Fire effects on wildlife in the Central
Hardwoods and Appalachian regions, USA". Fire Ecology
2016, 12(2):127-159.
[3] S.L. Stephens, J.K. Agee, P.Z. Fule, M. North, W.
Romme, T. Swetnam and M.G. Turner: "Managing forests
and fire in changing climates". Science 2013, 342(6154):41-
42.
[4] M.D. Flannigan, M.A. Krawchuk, W.J. de Groot, B.M.
Wotton and L.M. Gowman: "Implications of changing
climate for global wildland fire". International journal of
wildland fire 2009, 18(5):483-507.
[5] C. Aponte, W.J. de Groot and B.M. Wotton: "Forest
fires and climate change: causes, consequences and
management options". International Journal of Wildland
Fire 2016, 25(8):i-ii.
[6] L.T. Bennett, M.J. Bruce, J. Machunter, M. Kohout, S.J.
Krishnaraj and C. Aponte: "Assessing fire impacts on the
carbon stabilit\ RI ILUHဨWROHUDQW IRUHVWV Ecological
Applications 2017, 27(8):2497-2513.
[7] J.M. Earles, M.P. North and M.D. Hurteau: "Wildfire
and drought dynamics destabilize carbon stores of
ILUHဨVXSSUHVVHG IRUHVWV Ecological applications 2014,
24(4):732-740.
[8] M.D. Hurteau and M.L. Brooks: "Short-and long-term
effects of fire on carbon in US dry temperate forest
systems". BioScience 2011, 61(2):139-146.

Authorized licensed use limited to: University of Melbourne. Downloaded on February 22,2021 at 06:57:17 UTC from IEEE Xplore. Restrictions apply.
[9] M.D. Hurteau, S. Liang, K.L. Martin, M.P. North, G.W.
Koch and B.A. Hungate: "Restoring forest structure and
SURFHVV VWDELOL]HV IRUHVW FDUERQ LQ ZLOGILUHဨSURQH
southwestern ponderosa pine forests". Ecological
applications 2016, 26(2):382-391.
[10] L.T. Bennett, M.J. Bruce, J. MacHunter, M. Kohout,
M.A. Tanase and C. Aponte: "Mortality and recruitment of
fire-tolerant eucalypts as influenced by wildfire severity and
recent prescribed fire". Forest Ecology and Management
2016, 380:107-117.
[11] J.W. Van Wagtendonk, R.R. Root and C.H. Key:
"Comparison of AVIRIS and Landsat ETM+ detection
capabilities for burn severity". Remote Sensing of
Environment 2004, 92(3):397-408.
[12] S. Veraverbeke, W.W. Verstraeten, S. Lhermitte and
R. Goossens: "Evaluating Landsat Thematic Mapper
spectral indices for estimating burn severity of the 2007
Peloponnese wildfires in Greece". International journal of
wildland fire 2010, 19(5):558-569.
[13] B. Tran, M. Tanase, L. Bennett and C. Aponte:
"Evaluation of spectral indices for assessing fire severity in
Australian temperate forests". Remote Sensing 2018,
10(11):1680.
[14] N. Tran, M. Tanase, L. Bennett and C. Aponte: "Fire-
severity classification across temperate Australian forests:
random forests versus spectral index thresholding". Remote
Sensing for Agriculture, Ecosystems, and Hydrology XXI
2019, 11149:111490U.
[15] DELWP: Fire History Records of Fires Primarily on
Public Land. In. Melbourne, Victoria, Australia: Department
of Environment Land Water and Planning; 2017.
[16] R.C. Team: "R: A language and environment for
statistical computing". 2013.
[17] B.T. Bormann, P.S. Homann, R.L. Darbyshire and
B.A. Morrissette: "Intense forest wildfire sharply reduces
mineral soil C and N: the first direct evidence". Canadian
Journal of Forest Research 2008, 38(11):2771-2783.
[18] D.M. Bowman, B.P. Murphy, M.M. Boer, R.A.
Bradstock, G.J. Cary, M.A. Cochrane, R.J. Fensham, M.A.
Krawchuk, O.F. Price and R.J. Williams: "Forest fire
management, climate change, and the risk of catastrophic
carbon losses". Frontiers in Ecology and the Environment
2013, 11(2):66-67.
[19] W.H. Cai and J. Yang: "High-severity fire reduces
early successional boreal larch forest aboveground
productivity by shifting stand density in north-eastern
China". International journal of wildland fire 2016,
25(8):861-875.
[20] X. Hu, J. Zhu, C. Wang, T. Zheng, Q. Wu, H. Yao and
J. Fang: "Impacts of fire severity and post-fire reforestation
on carbon pools in boreal larch forests in Northeast China".
Journal of Plant Ecology 2015, 9(1):1-9.
[21] .- $QGHUVRQဨ7HL[HLUD $' 0LOOHU -( 0RKDQ
T.W. Hudiburg, B.D. Duval and E.H. DeLucia: "Altered
dynamics of forest recovery under a changing climate".
Global change biology 2013, 19(7):2001-2021.
[22] T.A. Fairman, C.R. Nitschke and L.T. Bennett: "Too
much, too soon? A review of the effects of increasing
wildfire frequency on tree mortality and regeneration in
temperate eucalypt forests". International journal of
wildland fire 2016, 25(8):831-848.
[23] L.D. Prior, G.J. Williamson and D.M. Bowman:
"Impact of high-severity fire in a Tasmanian dry eucalypt
forest". Australian Journal of Botany 2016, 64(3):193-205.
[24] M.D. Flannigan and C.E.V. Wagner: "Climate change
and wildfire in Canada". Canadian Journal of Forest
Research 1991, 21(1):66-72.
[25] M.S. Torn and J.S. Fried: "Predicting the impacts of
global warming on wildland fire". Climatic change 1992,
21(3):257-274.
[26] T. Beer, A. Gill and P. Moore: Australian bushfire
GDQJHU XQGHU FKDQJLQJ FOLPDWH UHJLPHV ,Q µ*UHHQKRXVH
planning for climatH FKDQJH¶(G *, 3HDUPDQ SS ±
427. In.: CSIRO: East Melbourne; 1988.

Authorized licensed use limited to: University of Melbourne. Downloaded on February 22,2021 at 06:57:17 UTC from IEEE Xplore. Restrictions apply.
... Over the last decades, an increasing trend in larger and more severe wildfires has been generally observed over many fire-prone regions, namely in the Mediterranean basin [1][2][3], California [4][5][6], and Australia [7,8]. More frequent hot and dry summer conditions, combined with climate change and high fuel accumulation over time and space, have already been shown to promote large and intense fire seasons across these regions [9][10][11]. ...
Article
Full-text available
Mediterranean European countries, including Portugal, are considered fire-prone regions, being affected by fire events every summer. Nonetheless, Portugal has been recording large burned areas over the last 20 years, which are not only strongly associated with hot and dry conditions, but also with high fuel availability in the ecosystems. Due to recent catastrophic fire seasons, Portugal has been implementing preventive policies during the pre-fire season, which, in turn, can optimize combat strategies during the fire season. In this context, our study contributes to fire prevention by identifying the regions with the highest potential to burn. The application of a Principal Component Analysis (PCA) to a range of climatological, ecological, and biophysical variables, either provided by remote sensing or reanalysis products, and known to be linked with diverse fire-vulnerability factors, allows the objective identification of the regions with the highest susceptibility to burn. The central and southernmost areas of Portugal present a stronger signal in the PCA, suggesting a likely high exposure to future fire events. The fuel accumulation over several months, in conjunction with elevation and fire weather conditions, are the terms out of the retained PCs that can explain most of the variability. The quality assessment performed for the burned areas in 2022 showed that they occurred in highly susceptible areas, highlighting the usefulness of the proposed methodology.
Conference Paper
Full-text available
Machine learning and spectral index (SI) thresholding approaches have been tested for fire-severity mapping from local to regional scales in a range of forest types worldwide. While index thresholding can be easily implemented, its operational utility over large areas is limited as the optimum index may vary with forest type and fire regimes. In contrast, machine learning algorithms allow for multivariate fire classifications. This study compared the accuracy of fire-severity classifications from SI thresholding with those from Random Forests (RF). Reference data were from 3730 plots within the boundaries of eight major wildfires across the six temperate forest ‘functional’ groups of Victoria, south-eastern Australia. The reference plots were randomly divided into training and validation datasets (60/40) for each fire-severity class (unburnt, low, moderate, high) and forest functional group. SI fire-severity classifications were conducted using thresholds derived in a previous study based on the same datasets. A RF classification algorithm was trained to derive fire-severity levels based on appropriate spectral indices and their temporal difference. The RF classification outperformed the SI thresholding approach in most cases, increasing overall accuracy by 11% on a forest-group basis, and 16% on an individual wildfire basis. Adding more predictor variables into the RF algorithm did not improve classification accuracy. Greater overall accuracies (by 12% on average) were achieved when in situ data (rather than data from other fires) were used to train the RF algorithm. Our study shows the utility of Random Forest algorithms for streamlining fire-severity mapping across heterogeneous forested landscapes.
Article
Full-text available
Spectral indices derived from optical remote sensing data have been widely used for fire 14-severity classification in forests from local to global scales. However, comparative analyses of 15 multiple indices across diverse forest types are few. This represents an information gap for fire 16 management agencies in areas like temperate southeastern Australia, which is characterized by a 17 diversity of natural forests that vary in structure and in the fire-regeneration strategies of the 18 dominant trees. 19 We evaluate 10 spectral indices across eight areas burnt by wildfires in 1998, 2006, 2007, and 2009 in 20 southeastern Australia. These wildfire areas encompass 13 forest types, which represent 86% of the 21 7.9M ha region's forest area. Forest types were aggregated into six forest groups based on their fire-22 regeneration strategies (seeders, resprouters) and structure (tree height and canopy cover). Index 23 performance was evaluated for each forest type and forest group by examining its sensitivity to four 24 fire-severity classes (unburnt, low, moderate, high) using three independent methods (anova, 25 separability, and optimality). For the best-performing indices, we calculated index-specific 26 thresholds (by forest types and groups) to separate between the four severity classes and evaluated 27 the accuracy of fire-severity classification on independent samples. 28 Our results indicated that the best-performing indices of fire severity varied with forest type and 29 group. Overall accuracy for the best-performing indices ranged from 0.50 to 0.78, and kappa values 30 ranged from 0.33 (fair agreement) to 0.77 (substantial agreement) depending on the forest group 31 and index. Fire severity in resprouter open forests and woodlands was most accurately mapped 32 using the delta Normalized Burnt ratio (dNBR). In contrast, dNDVI (delta Normalised difference 33 vegetation index) performed best for open forests with mixed fire responses (resprouters and 34 seeders), and dNDWI (delta Normalised difference water index) was the most accurate for obligate 35 seeder closed forests. Our analysis highlighted low sensitivity of all indices to fire impacts in 36 Rainforest. 37 We conclude that the optimal spectral index for quantifying fire severity varies with forest type, but 38 that there is scope to group forests by structure and fire-regeneration strategy to simplify fire-39 severity classification in heterogeneous forest landscapes. 40
Article
Full-text available
Climate warming is predicted to increase fire activity across the Eurasian boreal larch forest in the 21st century, which could have serious consequences on carbon storage. Quantifying the effects of fire disturbance on forest structure and aboveground net primary productivity (ANPP) could aid in our ability to predict future carbon storage on a regional and biome level. In this study, we examined the spatial heterogeneity of forest structure and ANPP on sites of varying fire severity and topographic position in a recently burned landscape in the Great Xing'an Mountains, China. Results indicated that after 11 years of vegetation regrowth, fire severity significantly affected forest regeneration ANPP. Spatial heterogeneities in forest regeneration ANPP were explained by both tree sapling density and understorey vegetation abundance. Although understorey vegetation productivity on average contributed 50% of total ANPP after fire, the increase in understorey productivity with fire severity could not offset the decrease in tree productivity in severely burned stands where tree sapling density was limited. Our results suggest that high-severity fire can decrease forest regeneration ANPP by altering forest structure in the early post-fire successional stage and that this shift in forest structure may influence future forest productivity trajectories over an extended period.
Article
The carbon stability of fire-tolerant forests is often assumed but less frequently assessed, limiting potential to anticipate threats to forest carbon posed by predicted increases in forest fire activity. Assessing the carbon stability of fire-tolerant forests requires multi-indicator approaches that recognise the myriad of ways that fires influence the carbon balance including combustion, deposition of pyrogenic material, and tree death, post-fire decomposition, recruitment, and growth. Five years after a large-scale wildfire in south-eastern Australia, we assessed the impacts of low- and high-severity wildfire, with and without prescribed fire (≤ 10 years before), on carbon stocks in multiple pools, and on carbon stability indicators (carbon stock percentages in live trees and in small trees, and carbon stocks in char and fuels) in fire-tolerant eucalypt forests. Relative to unburnt forest, high-severity wildfire decreased short-term (five-year) carbon stability by significantly decreasing live tree carbon stocks and percentage stocks in live standing trees (reflecting elevated tree mortality), by increasing the percentage of live tree carbon in small trees (those vulnerable to the next fire), and by potentially increasing the probability of another fire through increased elevated fine fuel loads. In contrast, low-severity wildfire enhanced carbon stability by having negligible effects on above-ground stocks and indicators, and by significantly increasing carbon stocks in char and, in particular, soils, indicating pyrogenic carbon accumulation. Overall, recent preceding prescribed fire did not markedly influence wildfire effects on short-term carbon stability at stand scales. Despite wide confidence intervals around mean stock differences – indicating uncertainty about the magnitude of fire effects in these natural forests – our assessment highlights the need for active management of carbon assets in fire-tolerant eucalypt forests under contemporary fire regimes. Decreased live tree carbon and increased reliance on younger cohorts for carbon recovery after high-severity wildfire, could increase vulnerabilities to imminent fires, leading to decisions about interventions to maintain the productivity of some stands. Our multi-indicator assessment also highlights the importance of considering all carbon pools, particularly pyrogenic reservoirs like soils, when evaluating the potential for prescribed fire regimes to mitigate the carbon costs of wildfires in fire-prone landscapes. This article is protected by copyright. All rights reserved.
Article
Mixed-species eucalypt forests of temperate Australia are assumed tolerant of most fire regimes based on the impressive capacity of the dominant eucalypts to resprout. However, empirical data to test this assumption are rare, limiting capacity to predict forest tolerance to emerging fire regimes including more frequent severe wildfires and extensive use of prescribed fire. We quantified tree mortality and regeneration in mixed-species eucalypt forests five years after an extensive wildfire that burnt under extreme fire weather. To examine combined site-level effects of wildfire and prescribed fire, our study included factorial replications of three wildfire severities, assessed as crown scorch and understorey consumption shortly after the wildfire (Unburnt, Low, High), and two times since last preceding fire (<10 years since prescribed fire, >30 years since any fire). Our data indicate that while most trees survived low-severity wildfire through epicormic resprouting, this capacity was tested by high-severity wildfire. Five years after the wildfire, percentage mortalities of eucalypts in all size intervals from 10 to >70 cm diameter were significantly greater at High severity than Unburnt or Low severity sites, and included the near loss of the 10 to 20 cm cohort (93% mortality). Prolific seedling regeneration at High severity sites, and unreliable basal resprouting, indicated the importance of seedling recruitment to the resilience of these fire-tolerant forests. Recent prescribed fire had no clear effect on forest resistance (as tree survival) to wildfire, but decreased site-level resilience (as recruitment) by increasing mortalities of small stems. Our study indicates that high-severity wildfire has the potential to cause transitions to more open, simplified stand structures through increased tree mortality, including disproportionate losses in some size cohorts. Dependence on seedling recruitment could increase vulnerabilities to subsequent fires and future climates, potentially requiring direct management interventions to bolster forest resilience.
Article
Changing climate and a legacy of fire-exclusion have increased the probability of high-severity wildfire, leading to an increased risk of forest carbon loss in ponderosa pine forests in the southwestern USA. Efforts to reduce high-severity fire risk through forest thinning and prescribed burning require both the removal and emission of carbon from these forests, and any potential carbon benefits from treatment may depend on the occurrence of wildfire. We sought to determine how forest treatments alter the effects of stochastic wildfire events on the forest carbon balance. We modeled three treatments (control, thin-only, and thin and burn) with and without the occurrence of wildfire. We evaluated how two different probabilities of wildfire occurrence, 1% and 2% per year, might alter the carbon balance of treatments. In the absence of wildfire, we found that thinning and burning treatments initially reduced total ecosystem carbon (TEC) and increased net ecosystem carbon balance (NECB). In the presence of wildfire, the thin and burn treatment TEC surpassed that of the control in year 40 at 2%/yr wildfire probability, and in year 51 at 1%/yr wildfire probability. NECB in the presence of wildfire showed a similar response to the no-wildfire scenarios: both thin-only and thin and burn treatments increased the C sink. Treatments increased TEC by reducing both mean wildfire severity and its variability. While the carbon balance of treatments may differ in more productive forest types, the carbon balance benefits from restoring forest structure and fire in southwestern ponderosa pine forests are clear.