Global and Regional Trends and Drivers of Fire Under Climate Change

Increases in burned area of approximately 50% or higher have been seen in some extratropical forest ecoregions including in the Pacific US and high-latitude forests during 2001-2019.

“Correspondingly, increases in BA of ∼50% or higher have been seen in some extratropical forest ecoregions including in the Pacific US and high-latitude forests during 2001-2019, though interannual variability remains large in these regions.”

The frequency and severity of fire weather has increased in recent decades.

“The frequency and severity of fire weather has increased in recent decades”

Burned area has reduced by 27% globally in the past two decades.

“Overall, BA has reduced by 27% globally in the past two decades, due in large part to a decline in BA in African savannahs.”

Burned area in savannahs relates more strongly to patterns of fuel production or to the fragmentation of naturally fire-prone landscapes by agriculture than to fire weather.

“For example, BA in savannahs relates more strongly to patterns of fuel production or to the fragmentation of naturally fire-prone landscapes by agriculture.”

Over 8,885 studies published since 1990 matched a combination of keywords considered most relevant to this review according to the Web of Science.

“over 8,885 studies published since 1990 matched a combination of the keywords that we consider most relevant to this review (”

Historical trends in burned area vary considerably across models.

“and historical trends in BA also vary considerably across models.”

Current fire models incompletely reproduce the observed spatial patterns of burned area based on their existing representations of the relationships between fire and its bioclimatic and human controls.

“Current fire models incompletely reproduce the observed spatial patterns of BA based on their existing representations of the relationships between fire and its bioclimatic and human controls”

Fire weather is the dominant control on variability in burned area in many mesic forest ecoregions.

“and is the dominant control on variability in burned area (BA) in many mesic forest ecoregions”

Burned area trends in tropical forests relate more strongly to deforestation rates and forest degradation than to changing fire weather.

“Similarly, BA trends in tropical forests relate more strongly to deforestation rates and forest degradation than to changing fire weather.”

At macroregional scales, increases in FWSL have been proportionally greatest in Europe, central and boreal Asia, southern hemisphere South America and temperate North America.

“At these macroregional scales, increases in FWSL have been proportionally greatest in Europe, central and boreal Asia, southern hemisphere South America and temperate North America.”

The decadal trend in FWSL seen here (~+7% decade-1) lies in the range reported by Jolly et al. (2015) for various combinations of meteorological data set and index.

“The decadal trend in FWSL seen here (∼+7% decade -1 ) lies in the range reported by Jolly et al. (2015) for various combinations of meteorological data set and index.”

• cites Jolly et al., 2015Not checked yet

  Not checked in the free run.

In our analysis, we observed increases in FWSL (+27%; 4 days year-1) and FWI95d (+17%; 3 days year-1) in east Siberian forests during 1979-2019.

“In our analysis, we observed increases in FWSL (+27%; 4 days year -1 ) and FWI 95d (+17%; 3 days year -1 ) in east Siberian forests during 1979-2019 (Figures 2- 4, Table 1).”

We also found significant positive trends in both FWSL and FWI95d in the majority of GFED macroregions.

“We also found significant positive trends in both FWSL and FWI 95d in the majority of GFED macroregions (Figure 2; Table 1).”

Significant declines in burned area in African savannahs occurred despite significant increases in fire weather, especially extreme fire weather, in the region.

“The significant declines in BA in African savannahs have occurred despite significant increases in fire weather, and especially extreme fire weather, in the region (Section 3.1)”

Despite the increases in fire weather seasons and fire weather extremes that have been observed in virtually all world regions, BA has shown a range of regional trends.

“Despite the increases in fire weather seasons and fire weather extremes that have been observed in virtually all world regions, BA has shown a range of regional trends.”

The decline in BA is most pronounced in northern African savannahs, where BA fell by 41% during 2001-2019.

“The decline in BA is most pronounced in northern African savannahs, where BA fell by 41% during 2001-2019.”

Declines in burned area in the Mediterranean occurred despite significant increases in FWSL and FWI 95d.

“The declines in BA in the Mediterranean have occurred despite significant increases in FWSL and FWI 95d”

In the paper's analysis, there were shallower and non-significant increases in burned area of 49% in Pacific US forests.

“In our analysis, we observed shallower (and non-significant) increases in BA of 49% in the Pacific US forests”

The insignificant negative correlations seen between fire-season burned area and fire-season FWI in northern African savannahs are clear outliers in the paper's analysis.

“The insignificant negative correlations seen between fire season BA and fire season FWI in northern African savannahs are clear outliers in our analysis.”

In the paper's analysis, there was a non-significant increase in burned area of 63% in Pacific Canadian forests during 2001-2019.

“In our analysis, we observed a non-significant increase in BA of 63% in Pacific Canadian forests during 2001-2019”

The correlations between total burned area and mean FWI were significant in Alaskan forests, Pacific Canadian forests, Pacific US forests, Indonesian forests and the Mediterranean.

“the correlations were significant in Alaskan forests, pacific Canadian forests, pacific US Forests, Indonesian forests and the Mediterranean”

In the paper's analysis, there was no substantial change in burned area in Alaskan forests during 2001-2019.

“and no substantial change in BA in Alaskan forests”

In the paper's analysis, there was no significant change in annual burned area in southeast Australian forests during 2001-2019.

“In our analysis, we observed no significant change in annual BA in southeast Australian forests during 2001-2019”

In the paper's analysis of the relationship between burned area and FWI, monthly burned area correlates positively and significantly with monthly FWI in the majority of world regions.

“In our analysis of the relationship between BA and FWI (see Text A in Supporting Information S1 for methods), we found that monthly BA correlates positively and significantly with monthly FWI in the majority of world regions”

Total burned area in a fire season correlates positively with mean FWI during the fire season in all regions except NHAF and northern African savannahs.

“the total BA in a fire season correlates positively with the mean FWI during the fire season in all regions except for NHAF and northern African savannahs”

Fire weather and burned area correlate positively on inter-annual timescales in many mesic ecoregions.

“In addition, fire weather and BA correlate positively on inter-annual timescales in many mesic ecoregions.”

In the paper's analysis, the correlations between burned area and FWI were weaker in southeast Australian forests than in many other regions of interest.

“In our analysis, we found that the correlations between BA and FWI were weaker in southeast Australian forests than in many other regions of interest”

Consistent with the generally negative relationship seen globally between tree cover and annual burned area, the paper observes negative correlations between tree biomass and burned area in the majority of world regions.

“Consistent with the generally negative relationship seen globally between tree cover and annual BA, we observe negative correlations between tree biomass and BA in the majority of world regions.”

Correlations between non-tree biomass and burned area are more often positive.

“In contrast, correlations between non-tree biomass and BA are more often positive.”

The positive correlation between lightning and BA in western and central parts of Amazonia is a peculiar regional finding with relatively little precedent from the existing literature.

“The positive correlation between lightning and BA in western and central parts of Amazonia is a peculiar regional finding with relatively little precedent from the existing literature (Figure 9, Table 5).”

Insignificant positive correlations between lightning activity and BA were seen in many parts of the extratropics, including in the forests of the Pacific US and Canada, Alaska, east Siberia.

“Insignificant positive correlations between lightning activity and BA were seen in many parts of the extratropics, including in the forests of the Pacific US and Canada, Alaska, east Siberia.”

In the paper's analysis, strong correlations between monthly BA and monthly lightning activity on a climatological basis were observed in the forests of Pacific Canada, Alaska and east Siberia.

“In our analysis, we observed strong correlations between monthly BA and monthly lightning activity on a climatological basis in the forests of Pacific Canada, Alaska and east Siberia, thus suggesting that lightning plays a key role as a fire ignition source throughout the boreal forest zone (Figure 9, Table 5).”

The strongest and most significant correlations were observed in high latitude forests of Alaska and Siberia.

“The strongest and most significant correlations were observed in high latitude forests of Alaska and Siberia.”

In the paper's analysis, insignificant negative correlations between BA and lightning activity were found across most of the tropics, including in African savannahs, but excluding some tropical forests.

“In contrast to the extratropics, our analysis shows insignificant negative correlations between BA and lightning activity across most of the tropics, including in African savannahs, but excluding some tropical forests.”

In the paper's analysis, correlations between mean annual BA and human population density at sub-grid scale are highly variable, holding a mixture of positive and negative values, such that they can average to weak values even within ecoregions.

“Correlations are highly variable on these spatial scales (e.g., holding a mixture of positive and negative values; Figure 10) such that they can average to weak values even within ecoregions (Table 6).”

Suppression has been less effective in regions dominated by high-intensity, stand-replacing fires associated with extreme fire weather conditions.

“Suppression has been less effective in regions dominated by high-intensity, stand-replacing fires associated with extreme fire weather conditions”

Positive correlations between human activities and burned area are seen in the Sahel and many parts of Southern African savannahs.

“positive correlations between human activities and BA are seen in the Sahel and many parts of Southern African savannahs (Figure 10).”

Modern suppression techniques have been effective in limiting the spread of fire in many fuel types globally for over a century, particularly where low- and moderate-intensity fires dominate the fire regime.

“Modern suppression techniques have been effective in limiting the spread of fire in many fuel types globally for over a century, particularly where low-and moderate-intensity fires dominate the fire regime.”

Increases in fire weather during the past four decades have exerted an upwards pressure on fire activity in virtually all regions of the world.

“Increases in fire weather during the past four decades have exerted an upwards pressure on fire activity in virtually all regions of the world (Section 3; Figure 7).”

In Alaskan and Pacific Canadian forests, extreme fire weather frequency is simulated to increase to 28%-44% (4-7 days year-1) at global mean annual temperature increments of 1.5-2.0°C, to 72%-82% (12-13 days year-1) at 3.0°C and 112%-144% (17-23 days year-1) at 4.0°C.

“Meanwhile, FWI 95d is simulated to increase to 28%-44% (4-7 days year -1 ) at global MAT increments of 1.5-2.0°C, to 72%-82% (12-13 days year -1 ) at 3.0°C and 112%-144% (17-23 days year -1 ) at 4.0°C.”

In Alaskan and Pacific Canadian forests, simulated increases in fire weather season length rise to 37%-62% (3-6 days year-1) at global mean annual temperature increments of 1.5-2.0°C, to 136% (7-13 days year-1) at 3.0°C and to 174%-177% (10-18 days year-1) at 4.0°C.

“In these regions, the simulated increases in FWSL rise to 37%-62% (3-6 days year -1 ) at global MAT increments of 1.5-2.0°C, to 136% (7-13 days year -1 ) at 3.0°C and to 174%-177% (10-18 days year -1 ) at 4.0°C.”

Fire weather season length is projected to increase particularly rapidly in Amazonia as higher temperature increments are breached, from +74% at 1.5°C, to +136% at 2.0°C, +282% at 3.0°C and +414% at 4.0°C.

“FWSL is projected to increase particularly rapidly in Amazonia as higher temperature increments are breached, from +74% at 1.5°C, to +136% at 2.0°C, +282% at 3.0°C and +414% at 4.0°C.”

The majority of models agree on the sign and significance of trends in fire weather season length and extreme fire weather frequency at temperature increments of 1.5°C and above.

“In general, the majority of models agree on the sign and significance of trends in FWSL and FWI 95d at temperature increments of 1.5°C and above (Tables 7 and 8).”

Population correlates negatively with burned area in coastal regions of California, southeast Australia and the Mediterranean.

“We found that population correlates negatively with BA in coastal regions of California, southeast Australia and the Mediterranean (Figure 10, Table 6).”

The impact of anthropogenic climate change on fire weather has already emerged beyond pre-industrial variability in the Mediterranean region according to our analysis.

“the impact of anthropogenic climate change on fire weather has already emerged beyond pre-industrial variability in the Mediterranean region according to our analysis.”

At 3.0°C above pre-industrial global mean annual temperature, fire weather season length and extreme fire weather frequency emerge above pre-industrial variability in Indonesian forests and southeast Australian forests.

“At 3.0°C above pre-industrial global MAT, FWSL, and FWI 95d emerge above pre-industrial variability in Indonesian forests and southeast Australian forests.”

Emergence becomes increasingly widespread as global mean annual temperature rises.

“Emergence becomes increasingly widespread as global MAT rises.”

Fire weather is yet to emerge beyond pre-industrial variability in the Pacific US forests or in southeast Australia where large pre-industrial variability obscures the signal of anthropogenic climate change despite visible trends in fire weather season length and extreme fire weather frequency.

“we found that fire weather is yet to emerge beyond pre-industrial variability in the Pacific US forests or in southeast Australia where large pre-industrial variability obscures the signal of anthropogenic climate change despite visible trends in FWSL and FWI 95d”

Both fire weather season length and modeled extreme fire weather frequency have emerged above pre-industrial variability in southern African savannahs, the Mediterranean and southern Amazonia.

“we find that both FWSL and modeled FWI 95d have emerged above pre-industrial variability in southern African savannahs, the Mediterranean and southern Amazonia (Tables 7 and 8).”

Another intriguing feature of the FireMIP trends is the modeled increase in BA in the Sahel region, which is not present in the MODIS BA data set.

“Another intriguing feature of the FireMIP trends is the modeled increase in BA in the Sahel region, which is not present in the MODIS BA data set.”

At the global scale, the median FireMIP model simulates global BA to within 3% of observations during 2001-2012.

“At the global scale, the median FireMIP model simulates global BA to within 3% of observations during 2001-2012, however, the median model underpredicts BA by 33% in southern African savannahs and 11% in northern African savannahs, representing large absolute errors totaling more than 0.5 million km 2 (22%) across the two regions.”

The median model simulations of BA were within 66% of observations in Alaskan forests, southern Amazonia, east Siberian forests and Indonesian forests.

“The median model simulations of BA were within 66% of observations in Alaskan forests, southern Amazonia, east Siberian forests and Indonesian forests.”

The FireMIP models did show a majority agreement on a significant reduction in BA in the Mediterranean (-39%), Indonesian forests (-28%), Pacific US forests (-20%), and northern African savannahs (-19%).

“Despite disagreeing on the global trend in BA during the past century, the FireMIP models did show a majority agreement on a significant reduction in BA in the Mediterranean (-39%), Indonesian forests (-28%), Pacific US forests (-20%), and northern African savannahs (-19%; Figure 13).”

The median model substantially overestimates BA in several macroregions and focus ecoregions, including by ∼400%-700% in southeast Australian forests, the Mediterranean and Pacific US forests.

“On the other hand, the median model substantially overestimates BA in several macroregions and focus ecoregions, including by ∼400%-700% in southeast Australian forests, the Mediterranean and Pacific US forests (Table 9, Figure 13).”

The simulated increases in fire weather tend to accelerate with each added increment of warming.

“The simulated increases in fire weather tend to accelerate with each added increment of warming”

Land use suppresses BA in some places, for example by fragmenting savannah vegetation.

“land use suppresses BA in some places (e.g., by fragmenting savannah vegetation)”

During 2001-2019, there were significant positive correlations between fire weather and BA in all ecoregions studied on a seasonal basis.

“During 2001-2019, we observed significant positive correlations between fire weather and BA in all ecoregions that we studied on a seasonal basis”

Globally, BA decreased by 1.1 million km2 year-1 or 27% during 2001-2019.

“Globally, BA decreased by 1.1 million km 2 year -1 or 27% during 2001-2019.”

FWSL increased by 27%-94% in southern Amazonia, the Mediterranean region and forests of southeast Australia, Alaska, east Siberia, Pacific Canada and the Pacific US during 1979-2019.

“27%-94% increases in FWSL have been observed in southern Amazonia, the Mediterranean region and forests of southeast Australia, Alaska, east Siberia, Pacific Canada and the Pacific US during 1979-2019”

Human activities are associated with elevated ignition frequency in other places, for example in tropical forests that are not naturally fire-prone.

“human activities are associated with elevated ignition frequency in other places (e.g., in tropical forests that are not naturally fire-prone)”

Extreme fire weather is generally projected to increase more sharply than FWSL.

“extreme fire weather is generally projected to increase more sharply than FWSL”

An ensemble of state-of-the-art FireMIP models reproduce global BA to within 3% of observational data.

“An ensemble of state-of-the-art FireMIP models reproduce global BA to within 3% of observational data”

According to CMIP5 climate model projections, the prevalence and extremity of fire weather have already emerged beyond their pre-industrial variability in the Mediterranean and southern Amazonia due to climate change.

“According to CMIP5 climate model projections, the prevalence and extremity of fire weather have already emerged beyond their pre-industrial variability in the Mediterranean and southern Amazonia due to climate change.”

There is poor agreement amongst the FireMIP models on the sign and magnitude of historical changes in BA at global and regional scales.

“there is poor agreement amongst the models on the sign and magnitude of historical changes in BA at global and regional scales.”

BA tends to increase with vegetation biomass in regions with low vegetation productivity (productivity-limited regions, e.g., savannah-grasslands).

“BA tends to increase with vegetation biomass in regions with low vegetation productivity (productivity-limited regions, e.g., savannah-grasslands)”

In some places, years with greatest BA tend to coincide with years of greatest fire weather.

“in some places the years with greatest BA also tend to coincide with the years of greatest fire weather”

Around half of this decline is explained by a significant decline in BA in African savannahs.

“Around half of this decline is explained by a significant decline in BA in African savannahs.”

The FireMIP ensemble collectively indicates that BA has declined by 14% globally over the past century.

“they collectively indicate that BA has declined by 14% globally over the past century”

At 3°C above the pre-industrial MAT, both FWSL and FWI95d emerge beyond pre-industrial variability in all of the focus ecoregions and macroregions that we studied.

“At 3°C above the pre-industrial MAT, both FWSL and FWI 95d emerge beyond pre-industrial variability in all of the focus ecoregions and macroregions that we studied.”

BA tends to reduce with biomass in regions with high vegetation productivity (e.g., forests).

“but reduce with biomass in regions with high vegetation productivity (e.g., forests)”

In mesic forests of the Pacific US and the high latitudes, notable increases in BA on the order of 49%-93% have occurred in tandem with significant increases in fire weather.

“In mesic forests of the Pacific US and the high latitudes, notable increases in BA on the order of 49%-93% have occurred in tandem with significant increases in fire weather.”

National or jurisdictional fire databases have important limitations for global-scale research questions, including partial spatial coverage.

“National or jurisdictional fire databases, such as those from forestry services, have often been used to reconstruct fire history, however, these have important limitations for global-scale research questions, not least their partial spatial coverage”

• cites Mouillot & Field, 2005Not checked yet

  Not checked in the free run.

According to climate models, the prevalence and extremity of fire weather has already emerged beyond its pre-industrial variability in the Mediterranean due to climate change.

“According to climate models, the prevalence and extremity of fire weather has already emerged beyond its pre-industrial variability in the Mediterranean due to climate change”

The positive climate-carbon cycle feedback involving fire is only partially offset by the negative feedback associated with increases in surface albedo and reductions in temperature as forest cover reduces.

“and this positive feedback is only partially offset by the negative feedback associated with increases in surface albedo and reductions in temperature as forest cover reduces”

• cites Z.Liu et al., 2019Not checked yet

  Not checked in the free run.

Several major wildfires experienced in recent years, including the Australian bushfires of 2019/2020, have occurred amidst fire weather conditions that were considerably more likely due to climate change.

“Moreover, several of the major wildfires experienced in recent years, including the Australian bushfires of 2019/2020, have occurred amidst fire weather conditions that were considerably more likely due to climate change.”

Various human and bioclimatic factors also control fire, modulating the relationship between burned area and fire weather in many regions.

“Various human and bioclimatic factors also control fire, modulating the relationship between BA and fire weather in many regions”

Increases in the frequency and extremity of fire weather have been globally pervasive due to climate change during 1979-2019.

“Increases in the frequency and extremity of fire weather have been globally pervasive due to climate change during 1979-2019, meaning that landscapes are primed to burn more frequently.”

The global reduction in burned area is due in large part to a decline in burned area in African savannahs.

“due in large part to a decline in BA in African savannahs.”

Climate change is exerting a pervasive upwards pressure on fire globally by increasing the frequency and intensity of fire weather.

“Overall, climate change is exerting a pervasive upwards pressure on fire globally by increasing the frequency and intensity of fire weather”

The frequency and severity of fire weather is projected to escalate with each added increment of warming.

“and is projected to escalate with each added increment of warming”

Other bioclimatic and human factors can override the relationship between burned area and fire weather.

“Nonetheless, other bioclimatic and human factors can override the relationship between BA and fire weather.”

Fire is a key disturbance mechanism that can prompt change in land cover from forest to non-forest in regions where climate becomes out of phase with the existing vegetation.

“Fire is also a key disturbance mechanism that can prompt change in land cover from forest to non-forest in regions where climate becomes out of phase with the existing vegetation”

• cites Burrell et al., 2020Not checked yet

  Not checked in the free run.

• cites Burrell et al., 2021Not checked yet

  Not checked in the free run.

• cites He et al., 2019Not checked yet

  Not checked in the free run.

Landscapes are primed to burn more frequently because of globally pervasive increases in fire weather frequency and extremity due to climate change during 1979-2019.

“meaning that landscapes are primed to burn more frequently.”

Emergence of fire weather beyond pre-industrial variability will become increasingly widespread at additional levels of warming.

“and emergence will become increasingly widespread at additional levels of warming.”

It can be difficult to disentangle the impacts of individual drivers on fire activity because fire is the result of the simultaneous occurrence of three factors: a stock of fuel; fire weather conditions sufficiently dry to desiccate the fuel; and a human or natural ignition source.

“It can be difficult to disentangle the impacts of individual drivers on fire activity because fire is the result of the simultaneous occurrence of three factors: a stock of fuel; fire weather conditions that are sufficiently dry to desiccate the fuel; and a human or natural ignition source”

• cites Abram et al., 2021Not checked yet

  Not checked in the free run.

• cites Bistinas et al., 2014Not checked yet

  Not checked in the free run.

• cites Forkel, Dorigo, et al., 2019Not checked yet

  Not checked in the free run.

The upwards pressure on fire from climate change will escalate with each increment of global warming.

“and this upwards pressure will escalate with each increment of global warming.”

The accuracy of burned area products is known to vary across land covers such that trends and variability in burned area can be assessed with greater confidence in some regions than others.

“Moreover, the accuracy of BA products is known to vary across land covers such that trends and variability in BA can be assessed with greater confidence in some regions than others.”

The latest burned area products are addressing the small-fire omission issue by merging active fire detections with burned area detections.

“The latest BA products are addressing this issue by merging active fire detections with BA detections to correct for small fires that are not detected through spectral changes”

• cites Chuvieco et al., 2018Not checked yet

  Not checked in the free run.

• cites Giglio et al., 2018Not checked yet

  Not checked in the free run.

• cites Roy et al., 2019Not checked yet

  Not checked in the free run.

Satellite products purely based on burned area detection have historically omitted many small fires below the approximately 500 m spatial resolution and consequently provide a conservative estimate of burned area.

“Satellite products that are purely based on BA detection have historically omitted many small fires that occur below the ∼500 m spatial resolution of the satellite data, and consequently they provide a conservative estimate of BA”

• cites Chuvieco et al., 2018Not checked yet

  Not checked in the free run.

• cites Giglio et al., 2010Not checked yet

  Not checked in the free run.

• cites Giglio et al., 2016Not checked yet

  Not checked in the free run.

Fire weather is one of the major controls on fire activity.

“Fire weather is one of the major controls on fire activity”

Since fire weather indices chiefly account for contemporaneous meteorological controls on fire risk, they give greatest insight into the likelihood of fire in regions where fire is limited by fuel dryness, as opposed to those regions where fire is limited by fuel productivity.

“Since fire weather indices chiefly account for the contemporaneous meteorological controls on fire risk, they give greatest insight into the likelihood of fire in regions where fire is limited by fuel dryness (Abatzoglou et al., 2018;Bedia et al., 2015), as opposed to those regions where fire is limited by fuel productivity”

• cites Abatzoglou et al., 2018Not checked yet

  Not checked in the free run.

• cites Bedia et al., 2015Not checked yet

  Not checked in the free run.

Fire weather or fire danger rating systems predominantly rely on inputs of meteorological variables and occasionally soil moisture to represent the contemporaneous and interacting influences of weather on vegetation flammability.

“Fire weather or fire danger rating systems predominantly rely on inputs of meteorological variables and occasionally soil moisture to represent the contemporaneous and interacting influences of weather on vegetation flammability”

• cites Abatzoglou et al., 2018Not checked yet

  Not checked in the free run.

• cites Abatzoglou & Kolden, 2013Not checked yet

  Not checked in the free run.

• cites Bedia et al., 2015Not checked yet

  Not checked in the free run.

Fire occurrence is determined by a delicate balance between vegetation productivity (fuel production) and the frequency of dry conditions (fuel moisture).

“Hence, fire occurrence is determined by a delicate balance between vegetation productivity (fuel production) and the frequency of dry conditions (fuel moisture).”

Shifts toward a warmer and more drought-prone climate can act as a key driver of increase in the likelihood of fire, at least up to the point where the system becomes productivity-limited.

“Shifts toward a warmer and more drought-prone climate can act as a key driver of increase in the likelihood of fire, at least up to the point where the system becomes productivity-limited.”

The relationship between lightning activity and climate has been investigated through its relationship with convective available potential energy (CAPE), convective fluxes of ice in clouds, or the intensity of in-cloud updrafts as determined by convective cloud top height.

“The relationship between lightning activity and climate has been investigated through its relationship with convective available potential energy (CAPE; a measure of the buoyancy force required for uplift of air masses in the atmosphere; Romps, 2019;Romps et al., 2014;Tippett et al., 2019), convective fluxes of ice in clouds (Finney et al., 2014(Finney et al., , 2018)), or the intensity of in-”

• cites Romps, 2019Not checked yet

  Not checked in the free run.

• cites Romps et al., 2014Not checked yet

  Not checked in the free run.

• cites Tippett et al., 2019Not checked yet

  Not checked in the free run.

Routine observation of lightning activity by satellites and networks of ground sensors are beginning to enable a more detailed understanding of the role of lightning activity as a driver of fire.

“The routine observation of lightning activity by satellites and networks of ground sensors are beginning to enable a more detailed understanding of the role of lightning activity as a driver of fire”

• cites Cecil et al., 2014Not checked yet

  Not checked in the free run.

• cites Christian, 2003Not checked yet

  Not checked in the free run.

• cites Styger et al., 2018Not checked yet

  Not checked in the free run.

Trends in atmospheric CO2 and land use change also influence vegetation cover, productivity and connectivity and thus modulate the response of vegetation productivity to climate.

“Moreover, trends in atmospheric CO 2 and land use change also influence vegetation cover, productivity and connectivity and thus modulate the response of vegetation productivity to climate.”

The world's ecoregions can be arranged along a climatic gradient of productivity-dryness that determines the key limitations on fire.

“A key concept that has emerged in recent decades is that the world's ecoregions can be arranged along a climatic gradient of productivity-dryness that determines the key limitations on fire”

• cites Archibald et al., 2009Not checked yet

  Not checked in the free run.

• cites Kelley et al., 2019Not checked yet

  Not checked in the free run.

• cites Parisien & Moritz, 2009Not checked yet

  Not checked in the free run.

Fire occurs most frequently in regions with intermediate moisture availability, where dry conditions are sufficiently frequent to intermittently desiccate the available vegetation but not consistently dry enough to limit fuel production.

“Consequently, fire occurs most frequently in regions with intermediate moisture availability, where dry conditions are sufficiently frequent to intermittently desiccate the available vegetation but not consistently dry enough to limit fuel production.”

The impact of climate change on fire activity is complicated because increases in fire weather can be either compounded or countered by changes in vegetation productivity.

“Hence, the impact of climate change on fire activity is complicated because increases in fire weather can be either compounded or countered by changes in vegetation productivity.”

State-of-the-art climate models which contribute to the Coupled Model Intercomparison Project provide the output variables required to assess climate change impacts on the frequency of fire-prone weather conditions.

“For historical and future centuries, state-of the-art climate models which contribute to the coupled model intercomparison project (CMIP) provide the output variables required to assess climate change impacts on the frequency of fire-prone weather conditions”

• cites Abatzoglou et al., 2019Not checked yet

  Not checked in the free run.

• cites Eyring et al., 2016Not checked yet

  Not checked in the free run.

• cites Meehl et al., 2009Not checked yet

  Not checked in the free run.

The overall effect of changes in climate and CO2 has been to increase biomass in many high latitude ecoregions.

“The overall effect of changes in climate and CO 2 has been to increase biomass in many high latitude ecoregions”

• cites Berner et al., 2020Not checked yet

  Not checked in the free run.

• cites Bistinas et al., 2014Not checked yet

  Not checked in the free run.

• cites C. Chen et al., 2019Not checked yet

  Not checked in the free run.

Despite known limitations, satellite burned area products have built a valuable global picture of fire activity at increasingly detailed resolution, which could not have been achieved through incomplete national or regional inventories.

“Overall, despite the known limitations to accuracy and inconsistencies across products, satellite BA products have built a valuable global picture of fire activity at increasingly detailed resolution, which could not have been achieved through incomplete national or regional inventories of fire activity.”

Variability in weather on interannual and sub-decadal timescales influences vegetation productivity and moisture levels on shorter timescales, thus determining the availability and dryness of fuels during any particular fire season.

“Meanwhile, variability in weather on interannual and sub-decadal timescales influences vegetation productivity and moisture levels on shorter timescales, thus determining the availability and dryness of fuels during any particular fire season”

• cites Alvarado et al., 2020Not checked yet

  Not checked in the free run.

• cites Forkel et al., 2017Not checked yet

  Not checked in the free run.

• cites Kelley et al., 2019Not checked yet

  Not checked in the free run.

The ignition efficiency of lightning strikes can vary strongly in space and time because other controls such as fire weather influence the flammability and receptiveness of vegetation.

“Moreover, the ignition efficiency of lightning strikes can vary strongly in space and time because other controls such as fire weather influence the flammability and receptiveness of vegetation (M.D. Flannigan & Wotton, 1991;Krawchuk et al., 2009;Sedano & Randerson, 2014;Wierzchowski et al., 2002).”

• cites M.D. Flannigan & Wotton, 1991Not checked yet

  Not checked in the free run.

• cites Krawchuk et al., 2009Not checked yet

  Not checked in the free run.

• cites Sedano & Randerson, 2014Not checked yet

  Not checked in the free run.

Within and before a fire season, prolonged moisture deficits or droughts during preceding months and seasons can reduce soil moisture reserves and make vegetation less resilient to peaks in fire weather.

“Within and before a fire season, prolonged moisture deficits or droughts during preceding months and seasons can reduce soil moisture reserves and make vegetation less resilient to peaks in fire weather”

• cites Abatzoglou & Kolden, 2013Not checked yet

  Not checked in the free run.

• cites Abatzoglou et al., 2018Not checked yet

  Not checked in the free run.

• cites Abram et al., 2021Not checked yet

  Not checked in the free run.

Dry lightning that strikes a fuel bed that has dried under fire weather conditions can result in widespread ignitions.

“Hence, dry lightning that strikes a fuel bed that has dried under fire weather conditions can result in widespread ignitions (Rorig & Ferguson, 2002).”

• cites Rorig & Ferguson, 2002Not checked yet

  Not checked in the free run.

Precipitation that accompanies wet lightning can suppress the ignition and spread of fire by increasing the moisture content of the vegetation.

“Precipitation that accompanies wet lightning can suppress the ignition and spread of fire by increasing the moisture content of the vegetation.”

A range of empirical models have been developed to relate fire activity to its drivers and constraints.

“A range of empirical models have been developed to relate fire activity to its drivers and constraints (Hantson et al., 2016).”

• cites Hantson et al., 2016Not checked yet

  Not checked in the free run.

Indigenous peoples used landscape and cultural fire intentionally for a wide range of purposes.

“Indigenous peoples used landscape and cultural fire intentionally for a wide range of purposes.”

• cites Kimmerer & Lake, 2001Not checked yet

  Not checked in the free run.

Interannual variability in BA tends to increase with economic development especially in the tropics.

“For example, interannual variability in BA tends to increase with economic development especially in the tropics (Chuvieco et al., 2021).”

• cites Chuvieco et al., 2021Not checked yet

  Not checked in the free run.

Thonicke et al. (2001) introduced an empirical model (GlobFIRM) which predicted BA as a function of fine fuel moisture, without explicitly accounting for the presence or absence of ignition sources.

“Thonicke et al. (2001) introduced an empirical model (GlobFIRM) which predicted BA as a function of fine fuel moisture, without explicitly accounting for the presence or absence of ignition sources.”

• cites Thonicke et al., 2001Not checked yet

  Not checked in the free run.

Many other empirical models, constructed on global or regional scales, show promise in reproducing observed patterns and variability in BA based on similar input data.

“Many other empirical models, constructed on global or regional scales, show promise in reproducing observed patterns and variability in BA based on similar input data (Archibald et al., 2009;Balshi et al., 2009;Forkel et al., 2017;Huang et al., 2015;Krawchuk et al., 2009;Moritz et al., 2012;Turco, Jerez, et al., 2018).”

• cites Archibald et al., 2009Not checked yet

  Not checked in the free run.

• cites Balshi et al., 2009Not checked yet

  Not checked in the free run.

• cites Forkel et al., 2017Not checked yet

  Not checked in the free run.

Much of the unexplained variability in fire-climate relationships and predictive models stems from the patterns and (un)predictability of human relationships with fire, and how they have diverged over time.

“Much of the unexplained variability in fire-climate relationships and predictive models stems from the patterns and (un)predictability of human relationships with fire, and how they have diverged over time (Ford et al., 2021).”

• cites Ford et al., 2021Not checked yet

  Not checked in the free run.

Concurrent to the reduction of Indigenous and traditional fire use, the forces of globalization introduced fire to facilitate deforestation and agriculture in regions where landscape fire was rare prior to colonization, such as in Amazonia and other tropical rainforests.

“Concurrent to the reduction of Indigenous and traditional fire use, the forces of globalization introduced fire to facilitate deforestation and agriculture in regions where landscape fire was rare prior to colonization, such as in Amazonia and other tropical rainforests (Bowman et al., 2011).”

• cites Bowman et al., 2011Not checked yet

  Not checked in the free run.

The growing availability of Earth observation data and understanding of the multiple controls on fire have enabled the development of empirical fire models and improvements to the representation of fire in process-based models, such as those used in dynamic global vegetation models (DGVMs) and Earth System models (ESMs).

“The growing availability of Earth observation data and understanding of the multiple controls on fire have enabled the development of empirical fire models and improvements to the representation of fire in processbased models, such as those used in dynamic global vegetation models (DGVMs) and Earth System models (ESMs).”

Alternative approaches based on machine learning models allow the use of a variety of potential climate, vegetation, and human-related predictors and hence to quantify the importance of the various controls on fire activity.

“Alternative approaches based on machine learning models allow the use of a variety of potential climate, vegetation, and human-related predictors and hence to quantify the importance of the various controls on fire activity (Archibald et al., 2009;Forkel, Andela, et al., 2019;Forkel et al., 2017).”

• cites Archibald et al., 2009Not checked yet

  Not checked in the free run.

• cites Forkel, Andela, et al., 2019Not checked yet

  Not checked in the free run.

• cites Forkel et al., 2017Not checked yet

  Not checked in the free run.

Most empirical fire models use pre-defined predictor variables and functional relationships to simulate fire activity and hence reflect the current understanding of the controls of fire activity.

“Most empirical fire models use pre-defined predictor variables and functional relationships to simulate fire activity and hence reflect the current understanding of the controls of fire activity.”

Empirical models trained on observational data have been used in a standalone fashion to predict fire activity under predicted changes in climate.

“Empirical models trained on observational data have been used in a standalone fashion to predict fire activity under predicted changes in climate (Archibald et al., 2009;Balshi et al., 2009;Krawchuk et al., 2009;Moritz et al., 2012;Pechony & Shindell, 2010;Turco, Rosa-Cánovas, et al., 2018;Turco et al., 2014).”

• cites Archibald et al., 2009Not checked yet

  Not checked in the free run.

• cites Balshi et al., 2009Not checked yet

  Not checked in the free run.

• cites Krawchuk et al., 2009Not checked yet

  Not checked in the free run.

Some recent studies have shown significant increases in the extreme values of FWI that are considered most dangerous for fire spread by fire management practitioners (e.g., FWI >20), especially in western Canadian forests.

“Nonetheless, some recent studies have shown significant increases in the extreme values of FWI that are considered most dangerous for fire spread by fire management practitioners (e.g., FWI >20), especially in western Canadian forests (Kirchmeier-Young et al., 2017, 2019;X. Wang et al., 2015).”

• cites Kirchmeier-Young et al., 2017, 2019Not checked yet

  Not checked in the free run.

• cites X. Wang et al., 2015Not checked yet

  Not checked in the free run.

Across all North American conifer forests, Jolly et al. (2015) reported a significant increase in FWSL of 6 days year-1 during 1979-2015 based on multiple meteorological reanalyzes and fire weather indices.

“Across all North American conifer forests (including high latitude forests in Alaska and Canada, temperate forests in the western US, and subtropical forests in the southeast US), Jolly et al. (2015) reported a significant increase in FWSL of 6 days year -1 during 1979-2015 based on multiple meteorological reanalyzes and fire weather indices.”

• cites Jolly et al., 2015Not checked yet

  Not checked in the free run.

Fréjaville and Curt (2015) showed that there had been a significant increase in fire weather severity in southern France, especially during spring.

“Fréjaville and Curt (2015) showed that there had been a significant increase in fire weather severity in southern France, especially during spring.”

• cites Fréjaville & Curt, 2015Not checked yet

  Not checked in the free run.

Most ESMs that contributed to CMIP5 and CMIP6, and which supported the IPCC's fifth assessment report on climate change impacts, have excluded fire processes.

“Nonetheless, most ESMs that contributed to the climate model intercomparison project's fifth (CMIP5; Kloster & Lasslop, 2017;Taylor et al., 2012) and sixth phases (CMIP6; Eyring et al., 2016;Lasslop, Hantson, Brovkin, et al., 2020), and which supported the IPCC's fifth assessment report on climate change impacts (Settele et al., 2014), have excluded fire processes.”

• cites Kloster & Lasslop, 2017Not checked yet

  Not checked in the free run.

• cites Taylor et al., 2012Not checked yet

  Not checked in the free run.

• cites Eyring et al., 2016Not checked yet

  Not checked in the free run.

Abatzoglou and Kolden (2013) showed that significant increases in extreme fire weather occurred during 1979-2012 across much of the western US.

“Abatzoglou and Kolden (2013) showed that significant increases in extreme fire weather occurred during 1979-2012 across much of the western US.”

• cites Abatzoglou & Kolden, 2013Not checked yet

  Not checked in the free run.

The prevailing explanation for the absence of significant trends in fire weather is that interannual variability is so large that even substantial trends are difficult to isolate robustly in the Canadian region.

“The prevailing explanation for the absence of significant trends in fire weather is that interannual variability is so large that even substantial trends are difficult to isolate robustly in this region (Amiro et al., 2004;Jain et al., 2017).”

• cites Amiro et al., 2004Not checked yet

  Not checked in the free run.

• cites Jain et al., 2017Not checked yet

  Not checked in the free run.

Bedia et al. (2012) identified significant trends in the seasonal severity rating, which is based on FWI values across annual fire seasons, on the Iberian Peninsula during 1989-2011.

“Bedia et al. (2012) identified significant trends in the seasonal severity rating, which is based on FWI values across annual fire seasons, on the Iberian Peninsula during 1989-2011.”

• cites Bedia et al., 2012Not checked yet

  Not checked in the free run.

ESMs that have included fire processes were found to reproduce satellite observations of BA poorly.

“ESMs that have included fire processes were found to reproduce satellite observations of BA poorly.”

• cites Kloster & Lasslop, 2017Not checked yet

  Not checked in the free run.

Both empirical and process-based models have been employed within a range of DGVMs, which are the terrestrial component of the ESMs used for climate simulation.

“Both empirical and process-based models have been employed within a range of DGVMs (Hantson et al., 2016(Hantson et al., , 2020;;Lasslop et al., 2019;Teckentrup et al., 2019), which are the terrestrial component of the ESMs used for climate simulation.”

• cites Hantson et al., 2016Not checked yet

  Not checked in the free run.

• cites Lasslop et al., 2019Not checked yet

  Not checked in the free run.

• cites Teckentrup et al., 2019Not checked yet

  Not checked in the free run.

Girardin et al. (2009) reported a significant increase in the forest area experiencing extreme drought codes during 1901-2002.

“In the Eurasian boreal region including Siberia, Girardin et al. (2009) reported a significant increase in the forest area experiencing extreme drought codes during 1901-2002.”

• cites Girardin et al., 2009Not checked yet

  Not checked in the free run.

Dowdy and Pepler (2018) showed that there have been significant increases in the frequency of spring and summer days with 95th percentile C-Haines index values and compound events when both fire weather and C-Haines are in their 95th percentiles in parts of southeast Australia during 1979-2016.

“Dowdy and Pepler (2018) showed that there have been significant increases in the frequency of spring and summer days with 95th percentile C-Haines index values and compound events when both fire weather and C-Haines are in their 95th percentiles in parts of southeast Australia during 1979-2016.”

• cites Dowdy & Pepler, 2018Not checked yet

  Not checked in the free run.

Jolly et al. (2015) reported a significant, but smaller, increase in FWSL of 10 days year-1 in the Mediterranean during 1979-2015 based on multiple meteorological reanalyzes and fire weather indices.

“Jolly et al. (2015) reported a significant, but smaller, increase in FWSL of 10 days year -1 in the Mediterranean during 1979-2015 based on multiple meteorological reanalyzes and fire weather indices.”

• cites Jolly et al., 2015Not checked yet

  Not checked in the free run.

Historically, the most pronounced seasons of drought-driven fire in Amazonia have been associated with natural variability in precipitation associated with El Niño events or sea surface temperatures in the North Atlantic.

“Historically, the most pronounced seasons of drought-driven fire in Amazonia have been associated with natural variability in precipitation associated with El Niño events or sea surface temperatures in the North Atlantic (Aragão et al., 2018;Brando et al., 2014;Jiménez-Muñoz et al., 2016;Lewis et al., 2011;Marengo & Espinoza, 2016;Marengo et al., 2018;Panisset et al., 2018).”

• cites Aragão et al., 2018Not checked yet

  Not checked in the free run.

• cites Brando et al., 2014Not checked yet

  Not checked in the free run.

• cites Jiménez-Muñoz et al., 2016Not checked yet

  Not checked in the free run.

The paper's results support the conclusion that African savannahs have an exceptional influence on global trends in burned area.

“Our results support the conclusion of Andela et al. (2017) and others (Forkel, Dorigo, et al., 2019) that African savannahs have an exceptional influence on global trends in BA.”

• cites Andela et al. (2017)Not checked yet

  Not checked in the free run.

• cites Forkel, Dorigo, et al., 2019Not checked yet

  Not checked in the free run.

Among the most important atmospheric factors contributing to an elevated probability of fire in southeast Australia is the development of unstable pyroconvective conditions associated with frontal systems.

“Among the most important atmospheric factors contributing to an elevated probability of fire in southeast Australia is the development of unstable “pyroconvective” conditions associated with frontal systems (Abram et al., 2021;Canadell et al., 2021;Dowdy & Mills, 2012;Sharples et al., 2016).”

• cites Abram et al., 2021Not checked yet

  Not checked in the free run.

• cites Canadell et al., 2021Not checked yet

  Not checked in the free run.

• cites Dowdy & Mills, 2012Not checked yet

  Not checked in the free run.

The interannual variability is specifically attributed to large-scale oceanic oscillations in the Pacific (ENSO), Southern (SAM) and Indian oceans (IOD), which strongly influence synoptic-scale variability in temperature and precipitation in southeast Australia, promoting large interannual variability in fire weather.

“The interannual variability is specifically attributed to large-scale oceanic oscillations in the Pacific (El Niño Southern Oscillation [ENSO]), Southern (Southern Annular Mode) and Indian oceans (Indian Ocean Dipole), which strongly influence synoptic-scale variability in temperature and precipitation in southeast Australia, promoting large interannual variability in fire weather (Abram et al., 2”

• cites Abram et al., 2021Not checked yet

  Not checked in the free run.

• cites H. Clarke et al., 2013Not checked yet

  Not checked in the free run.

• cites Harris & Lucas, 2019Not checked yet

  Not checked in the free run.

New analyses with multiple fire weather indices show a shift toward high occurrence of extreme fire weather days in the southeast of Australia since around the year 2000.

“New analyses with multiple fire weather indices show a shift toward high occurrence of extreme fire weather days in the southeast of Australia since around the year 2000 (Canadell et al., 2021;Richardson et al., 2021).”

• cites Canadell et al., 2021Not checked yet

  Not checked in the free run.

• cites Richardson et al., 2021Not checked yet

  Not checked in the free run.

Declines in burned area in the Mediterranean have generally been associated with increased suppression of fires by humans since the 1980s.

“they have generally been associated with increased suppression of fires by humans since the 1980s”

• cites Brotons et al., 2013Not checked yet

  Not checked in the free run.

• cites Ruffault & Mouillot, 2015Not checked yet

  Not checked in the free run.

• cites Silva et al., 2019Not checked yet

  Not checked in the free run.

Giannaros et al. (2021) revealed significant positive trends in the frequency of fire weather extremes in the Iberian Peninsula and eastern Balkans during 1987-2016.

“Recently, Giannaros et al. (2021) revealed significant positive trends in the frequency of fire weather extremes in the Iberian Peninsula and eastern Balkans during 1987-2016.”

• cites Giannaros et al., 2021Not checked yet

  Not checked in the free run.

The compounding effect of pyroconvection on likelihood of fire has also been highlighted in the Mediterranean.

“The compounding effect of pyroconvection on likelihood of fire have also been highlighted in the Mediterranean (Pinto et al., 2020;Tatli & Türkeş, 2014).”

• cites Pinto et al., 2020Not checked yet

  Not checked in the free run.

• cites Tatli & Türkeş, 2014Not checked yet

  Not checked in the free run.

The likelihood of pyroconvective conditions is raised by offshore cold fronts that draw dry and warm winds toward southeast Australia from the interior of the continent.

“The likelihood of pyroconvective conditions is raised by offshore cold fronts that draw dry and warm winds toward southeast Australia from the interior of the continent (Di Virgilio et al., 2019;Dowdy & Pepler, 2018;Dowdy et al., 2019;Reeder et al., 2015).”

• cites Di Virgilio et al., 2019Not checked yet

  Not checked in the free run.

• cites Dowdy & Pepler, 2018Not checked yet

  Not checked in the free run.

• cites Dowdy et al., 2019Not checked yet

  Not checked in the free run.

The significant trends in fire weather corroborate with reductions in dry season precipitation, increases in the duration of the dry season, and increases in the frequency of drought conditions that have been reported for the Amazonia region.

“The significant trends in fire weather corroborate with reductions in dry season precipitation, increases in the duration of the dry season, and increases in the frequency of drought conditions that have been reported for the region (Marengo & Espinoza, 2016;Marengo et al., 2011Marengo et al., , 2018;;Nogueira et al., 2017).”

• cites Marengo & Espinoza, 2016Not checked yet

  Not checked in the free run.

• cites Marengo et al., 2011Not checked yet

  Not checked in the free run.

• cites Nogueira et al., 2017Not checked yet

  Not checked in the free run.

The number of days with FWI values over 30, which are deemed to support extreme fire spread, increased by 1 day year-1 in the Iberian Peninsula and eastern Balkans.

“The number of days with FWI values over 30, which are deemed to support extreme fire spread, increased by 1 day year -1 in these regions.”

• cites Giannaros et al., 2021Not checked yet

  Not checked in the free run.

Deforestation rates declined overall during 2001-2019 in Amazonia following peak rates in the late 1990s and early 2000s.

“Deforestation rates declined overall during 2001-2019 in Amazonia following peak rates in the late 1990s and early 2000s”

• cites Aragão et al., 2018Not checked yet

  Not checked in the free run.

• cites Silva Junior et al., 2021Not checked yet

  Not checked in the free run.

• cites Libonati et al., 2021Not checked yet

  Not checked in the free run.

Modern-day burned area is comparable to the estimated annual burned area for western US forests during the early twentieth century.

“In the modern day, BA is comparable to the estimated annual BA for western US forests during the early twentieth century”

• cites Littell et al., 2009Not checked yet

  Not checked in the free run.

As in Pacific US forests, the increases in total burned area seen in the paper's analysis appear to mask more prominent increases in the frequency and extent of large fires as reported elsewhere based on agency statistics.

“As in Pacific US forests, the increases in total BA seen in our analysis appear to mask more prominent increases in the frequency and extent of large fires as reported elsewhere based on agency statistics.”

Trends in FWI and burned area often diverge in regions where other bioclimatic and human factors act as overriding controls on fire activity, such as African savannahs, southern Amazonia, and the Mediterranean.

“Nonetheless, trends in FWI and BA often diverge in regions where other bioclimatic and human factors act as overriding controls on fire activity (e.g., African savannahs, southern Amazonia, and the Mediterranean).”

Extreme fire weather conditions seen during the 2019/2020 bushfires can occur naturally when several ocean-atmosphere oscillations align at peak amplitude (ENSO, Southern Annular Mode and Indian Ocean Dipole).

“Extreme fire weather conditions seen during the 2019/2020 bushfires can occur naturally when several ocean-atmosphere oscillations align at peak amplitude (ENSO, Southern Annular Mode and Indian Ocean Dipole)”

• cites Abram et al., 2021Not checked yet

  Not checked in the free run.

• cites Harris & Lucas, 2019Not checked yet

  Not checked in the free run.

Legislation implemented to prevent deforestation later led to a 75% decline in deforestation rates during 2005-2015.

“Legislation implemented to prevent deforestation later led to a 75% decline in deforestation rates during the period 2005-2015”

• cites Aragão et al., 2018Not checked yet

  Not checked in the free run.

• cites Silva Junior et al., 2021Not checked yet

  Not checked in the free run.

• cites D. Nepstad et al., 2014Not checked yet

  Not checked in the free run.

Other regions of the world have seen substantial increases in fire activity and burned area in recent decades.

“Other regions of the world have seen substantial increases in fire activity and BA in recent decades.”

• cites Westerling (2006, 2016)Not checked yet

  Not checked in the free run.

• cites Abatzoglou & Williams, 2016Not checked yet

  Not checked in the free run.

• cites Williams et al., 2019Not checked yet

  Not checked in the free run.

These negative relationships are best explained by strong alternative controls on fire activity in northern African savannahs, including fuel availability and connectivity as governed by moisture availability and by fragmentation of natural landscapes by agriculture.

“These negative relationships are best explained by strong alternative controls on fire activity in this region, including fuel availability and connectivity as governed by moisture availability and by the fragmentation of natural landscapes by agriculture”

• cites Alvarado et al., 2020Not checked yet

  Not checked in the free run.

Background increases in average fire weather due to anthropogenic climate change led to a greater than 30% increase in the likelihood of 2019/2020 fire weather conditions.

“However, background increases in average fire weather due to anthropogenic climate change led to a >30% increase in the likelihood of 2019/2020 fire weather conditions”

• cites van Oldenborgh et al., 2021Not checked yet

  Not checked in the free run.

Deforestation fires and drought-driven fires are not mutually independent because deforestation has a legacy effect on the susceptibility of forest edges to drought.

“Deforestation fires and drought-driven fires are not mutually independent because deforestation has a legacy effect on the susceptibility of forest edges to drought”

• cites Brando et al., 2014Not checked yet

  Not checked in the free run.

• cites Cochrane, 2003Not checked yet

  Not checked in the free run.

• cites Silva Junior et al., 2018Not checked yet

  Not checked in the free run.

Humans suppressed fire in western US forests during the mid-to-late twentieth century through wildland firefighting and land conversion to agriculture.

“highlighting the extent to which humans have suppressed fire in the region during the mid-to-late twentieth century through wildland firefighting and land conversion to agriculture”

• cites Littell et al., 2009Not checked yet

  Not checked in the free run.

In addition to deforestation fires, wildfires associated with drought have superimposed substantial inter-annual variability on the burned-area trend in Amazonia.

“In addition to deforestation fires, wildfires associated with drought have superimposed substantial inter-annual variability on the BA trend in Amazonia.”

• cites Aragão et al., 2018Not checked yet

  Not checked in the free run.

• cites D. C. Nepstad et al., 2008Not checked yet

  Not checked in the free run.

• cites Lewis et al., 2011Not checked yet

  Not checked in the free run.

The shallower increase in Pacific US forests likely reflects the paper's consideration of total burned area as opposed to the area burned by large fires captured by agency statistics.

“which likely reflects our consideration of total BA as opposed to the area burned by the large fires as captured by agency statistics”

The decoupling of fire incidence from deforestation activity does not emerge robustly across all data sets.

“However, this “decoupling” of fire incidence from deforestation activity does not emerge robustly across all data sets”

• cites Libonati et al., 2021Not checked yet

  Not checked in the free run.

Droughts in Amazonia in 2005, 2010, and 2015 led to substantial peaks in fire activity.

“Droughts in the region in 2005, 2010, and 2015 have led to substantial peaks in fire activity”

• cites Aragão et al., 2018Not checked yet

  Not checked in the free run.

• cites D. C. Nepstad et al., 2008Not checked yet

  Not checked in the free run.

• cites Lewis et al., 2011Not checked yet

  Not checked in the free run.

The ecoregions of the world can be arranged along a fire limitation continuum ranging from systems where vegetation fuels are too sparse to sustain significant fire spread (fuel production limited) to systems where fuels are rarely dry enough to burn (fuel moisture limited).

“The ecoregions of the world can be arranged along a fire limitation continuum ranging from systems where vegetation fuels are too sparse to sustain significant fire spread (fuel production limited) to systems where fuels are rarely dry enough to burn (fuel moisture limited)”

• cites Alvarado et al., 2020Not checked yet

  Not checked in the free run.

• cites Archibald et al., 2009Not checked yet

  Not checked in the free run.

• cites Kelley et al., 2019Not checked yet

  Not checked in the free run.

Annual extent of forest fires correlates positively with precipitation in prior years in some parts of France, Greece, and Spain.

“The annual extent of forest fires has been shown to correlate positively with precipitation in prior years in some parts of France (Ruffault et al., 2016), Greece (Koutsias et al., 2012), and Spain (Turco et al., 2013), although Turco et al. (2017) concluded that antecedent conditions affecting vegetation growth explained a minor fraction of the variability in summer BA at the scale of the mediter”

• cites Ruffault et al., 2016Not checked yet

  Not checked in the free run.

• cites Koutsias et al., 2012Not checked yet

  Not checked in the free run.

• cites Turco et al., 2013Not checked yet

  Not checked in the free run.

Fire has imprinted on the traits of some plant species, which have evolved survival or rejuvenation strategies to deal with fire disturbance.

“Fire has itself imprinted on the traits of some plant species, which have evolved survival or rejuvenation strategies to deal with fire disturbance.”

• cites Bond & Keeley, 2005Not checked yet

  Not checked in the free run.

• cites Cawson & Duff, 2019Not checked yet

  Not checked in the free run.

• cites Crisp et al., 2011Not checked yet

  Not checked in the free run.

In many savannahs, weather conditions lead to desiccation and turnover of fuels in virtually every dry season, such that the quantity of vegetation produced during prior wet seasons is a greater control on interannual variability in BA than contemporaneous moisture deficit in the dry season.

“In many savannahs, weather conditions lead to the desiccation and turnover of fuels in virtually every dry season, such that the quantity of vegetation produced during prior wet seasons is a greater control on interannual variability in BA than contemporaneous moisture deficit in the dry season season (Alvarado et al., 2020;Archibald et al., 2009;van der Werf et al., 2008;Y. Chen et al., 2017).”

• cites Alvarado et al., 2020Not checked yet

  Not checked in the free run.

• cites Archibald et al., 2009Not checked yet

  Not checked in the free run.

• cites van der Werf et al., 2008Not checked yet

  Not checked in the free run.

2-year precipitation was a stronger covariate of BA than contemporaneous fire weather in northern Australian savannahs.

“Abatzoglou et al. (2018) also highlighted that 2-year precipitation was a stronger covariate of BA than contemporaneous fire weather in northern Australian savannahs.”

• cites Abatzoglou et al., 2018Not checked yet

  Not checked in the free run.

The paper's analysis implies the importance of species traits by contrasting correlations between tree biomass and BA in Pacific Canadian and Alaskan forests versus east Siberian forests, with the latter showing a negative correlation consistent with fire-resistant properties of dominant subspecies.

“The importance of species traits is similarly implied in our analysis by the contrasting correlations between tree biomass and BA in the Pacific Canadian and Alaskan forests versus the east Siberian forests, with the latter showing a negative correlation consistent with the fire-resistant properties of dominant subspecies in the region (Figure 8, Table 4).”

Monthly BA increases with increasing vegetation photosynthesis in the previous 6 months in many semi-arid regions.

“By using machine learning models, it has been shown that monthly BA increases with increasing vegetation photosynthesis in the previous 6 months in many semi-arid regions (Forkel, Andela, et al., 2019;Kuhn-Régnier et al., 2021).”

• cites Forkel, Andela, et al., 2019Not checked yet

  Not checked in the free run.

• cites Kuhn-Régnier et al., 2021Not checked yet

  Not checked in the free run.

These correlations are broadly consistent with the flammable nature of herbaceous fuels, which dry more rapidly during periods of fire.

“These correlations are broadly consistent with the flammable nature of herbaceous fuels, which dry more rapidly during periods of fire”

Regeneration of black spruce communities following fire is beginning to falter in some regions of North America as reductions in moisture availability lessen their competitive advantage over other species.

“Recently, it has been reported that the regeneration of black spruce communities following fire is beginning to falter in some regions of North America as reductions in moisture availability lessen their competitive advantage over other species (Baltzer et al., 2021).”

• cites Baltzer et al., 2021Not checked yet

  Not checked in the free run.

Plant traits can modulate fire characteristics.

“Notably, plant traits can also modulate fire characteristics.”

• cites Baltzer et al., 2021Not checked yet

  Not checked in the free run.

• cites de Groot et al., 2013Not checked yet

  Not checked in the free run.

• cites Rogers et al., 2015Not checked yet

  Not checked in the free run.

These spatial correlations suggest that denser forests in southeast Australia, the Pacific US, and the Iberian Peninsula support a greater level of fire activity, in contrast to forests in other world regions.

“These spatial correlations suggest that denser forests in these regions support a greater level of fire activity, in contrast to forests in other world regions.”

Including vegetation properties such as stand age and species composition can improve the predictive capacity of fire models in Canadian forests versus those based on fire weather alone, including for severe fires.

“Several studies have shown that including vegetation properties such as stand age and species composition can improve the predictive capacity of fire models in Canadian forests versus those based on fire weather alone (Cavard et al., 2015;Hanes et al., 2019), including in the case of severe fires (Walker et al., 2020).”

• cites Cavard et al., 2015Not checked yet

  Not checked in the free run.

• cites Hanes et al., 2019Not checked yet

  Not checked in the free run.

• cites Walker et al., 2020Not checked yet

  Not checked in the free run.

Many ecoregions that display greatest sensitivity to monthly or interannual variability in fire weather, including most mesic forests, are those without substantial fuel limitations and where fire weather occurs with low to moderate frequency.

“Many of the ecoregions that display greatest sensitivity to monthly or interannual variability in fire weather, including most mesic forests, are those without substantial fuel limitations and where fire weather occurs with low to moderate frequency”

• cites Abatzoglou et al., 2018Not checked yet

  Not checked in the free run.

• cites Bedia et al., 2015Not checked yet

  Not checked in the free run.

Parts of Spain, southern France and Italy have lightning fires exceeding human ignitions, typically in remote forested areas.

“Nonetheless, several studies have identified parts of Spain (Amatulli et al., 2007;González-Olabarria et al., 2015;Vazquez & Moreno, 1998), southern France (Curt et al., 2016;Ganteaume & Guerra, 2018) and Italy (Vacchiano et al., 2018) where lightning fires exceed human ignitions, which are typically remote forested areas (Perez-Invernon et al., 2021).”

• cites Amatulli et al., 2007Not checked yet

  Not checked in the free run.

• cites Curt et al., 2016Not checked yet

  Not checked in the free run.

• cites Vacchiano et al., 2018Not checked yet

  Not checked in the free run.

Human ignitions prevail over lightning ignitions in most urban, peri-urban and rural landscapes.

“In general, human ignitions prevail over lightning ignitions in most urban, peri-urban and rural landscapes, however, numerous studies have shown that lightning strikes are the major source of fire in the most remote parts of the world (Abatzoglou et al., 2016;Balch et al., 2017;Cattau et al., 2020;Ganteaume et al., 2013;Kasischke et al., 2010;Reeve & Toumi, 1999;Stocks et al., 2002).”

• cites Abatzoglou et al., 2016Not checked yet

  Not checked in the free run.

• cites Balch et al., 2017Not checked yet

  Not checked in the free run.

• cites Cattau et al., 2020Not checked yet

  Not checked in the free run.

Just 5% of all fires in the Mediterranean region are caused by lightning.

“One study reports that just 5% of all fires in the Mediterranean region are caused by lightning (Camia et al., 2010).”

• cites Camia et al., 2010Not checked yet

  Not checked in the free run.

Ecoregions of the US in which fires are predominantly ignited by lightning experience fires that are on average nine times larger than in areas where human ignitions dominate.

“Cattau et al. (2020) showed that ecoregions of the US in which fires are predominantly ignited by lightning experience fires that are on average nine times larger than in areas where human ignitions dominate.”

• cites Cattau et al., 2020Not checked yet

  Not checked in the free run.

Lightning strikes are the major source of fire in the most remote parts of the world.

“In general, human ignitions prevail over lightning ignitions in most urban, peri-urban and rural landscapes, however, numerous studies have shown that lightning strikes are the major source of fire in the most remote parts of the world (Abatzoglou et al., 2016;Balch et al., 2017;Cattau et al., 2020;Ganteaume et al., 2013;Kasischke et al., 2010;Reeve & Toumi, 1999;Stocks et al., 2002).”

• cites Abatzoglou et al., 2016Not checked yet

  Not checked in the free run.

• cites Balch et al., 2017Not checked yet

  Not checked in the free run.

• cites Cattau et al., 2020Not checked yet

  Not checked in the free run.

The correlation between lightning activity and BA was weak unless the impacts of fire weather on vegetation dryness were also accounted for.

“Combining these records with data from the US National Lightning Detection Network and North American Precision Lightning Network, the authors showed that the correlation between lightning activity and BA was weak unless the impacts of fire weather on vegetation dryness were also accounted for.”

• cites Abatzoglou et al., 2016Not checked yet

  Not checked in the free run.

In Canada, there is a gradient from predominantly human-ignited fires in the more populous south to predominantly lightning-ignited fires in the more remote northern regions.

“Previously, Stocks et al. (2002) examined national fire records of large fires in Canada and revealed a gradient from predominantly human-ignited fires in the more populous south of the country to predominantly lightning-ignited fires in the more remote northern regions.”

• cites Stocks et al., 2002Not checked yet

  Not checked in the free run.

Climatic and topographic conditions are favorable for dry lightning occurrence in the North American boreal region.

“Numerous studies have also highlighted that climatic and topographic conditions are favorable for dry lightning occurrence in the North American boreal region (Blouin et al., 2016;Dissing & Verbyla, 2003;Peterson et al., 2010).”

• cites Blouin et al., 2016Not checked yet

  Not checked in the free run.

• cites Dissing & Verbyla, 2003Not checked yet

  Not checked in the free run.

• cites Peterson et al., 2010Not checked yet

  Not checked in the free run.

At the scale of Victoria and New South Wales, lightning ignitions of large forest fires are outnumbered by human ignitions of small fires especially in the more populated coastal regions.

“Nonetheless, at the scale of Victoria and New South Wales, lightning ignitions of large forest fires are outnumbered by human ignitions of small fires especially in the more populated coastal regions (Attiwill & Adams, 2013; K. M. Collins et al., 2015;Williamson et al., 2016).”

• cites Attiwill & Adams, 2013Not checked yet

  Not checked in the free run.

• cites K. M. Collins et al., 2015Not checked yet

  Not checked in the free run.

• cites Williamson et al., 2016Not checked yet

  Not checked in the free run.

In southeast Australian forests, BA is notably more limited by fire weather than by fuel accumulation, in contrast to semi-arid grasslands and shrublands.

“Similar observations have been made in southeast Australian forests where BA is notably more limited by fire weather than by fuel accumulation, in contrast to semi-arid grasslands and shrublands (Bradstock, 2010;Russell-Smith et al., 2007).”

• cites Bradstock, 2010Not checked yet

  Not checked in the free run.

• cites Russell-Smith et al., 2007Not checked yet

  Not checked in the free run.

Forest fires in Siberia are predominantly caused by lightning.

“The distribution of fires across populated and unpopulated regions of Russia infers that forest fires in Siberia are predominantly caused by lightning (Kharuk et al., 2011;Mollicone et al., 2006;Ponomarev et al., 2016), although the explicit separation of human and lightning -ignited fires is pending in the region.”

• cites Kharuk et al., 2011Not checked yet

  Not checked in the free run.

• cites Mollicone et al., 2006Not checked yet

  Not checked in the free run.

• cites Ponomarev et al., 2016Not checked yet

  Not checked in the free run.

Around 70% of large fires and 85% of large fire BA were associated with lightning during 1959-1997.

“Around 70% of large fires and 85% of large fire BA were associated with lightning during 1959-1997 (Blouin et al., 2016;Stocks et al., 2002).”

• cites Blouin et al., 2016Not checked yet

  Not checked in the free run.

• cites Stocks et al., 2002Not checked yet

  Not checked in the free run.

Lightning plays a key role as a fire ignition source throughout the boreal forest zone.

“In our analysis, we observed strong correlations between monthly BA and monthly lightning activity on a climatological basis in the forests of Pacific Canada, Alaska and east Siberia, thus suggesting that lightning plays a key role as a fire ignition source throughout the boreal forest zone (Figure 9, Table 5).”

The negative correlation for much of the tropics likely reflects the coincidence of wet season peaks in lightning frequency with the season of greatest vegetation moisture content (and therefore lowest flammability), which determines that lightning strikes predominantly occur at a time when fire ignition and spread are most inhibited.

“The negative correlation for much of the tropics likely reflects the coincidence of wet season peaks in lightning frequency with the season of greatest vegetation moisture content (and therefore lowest flammability), which determines that lightning strikes predominantly occur at a time when fire ignition and spread are most inhibited.”

Global-scale analysis has shown that forest BA rarely exhibits a strong positive correlation with rainfall in prior years.

“Global-scale analysis has shown that forest BA rarely exhibits a strong positive correlation with rainfall in prior years (Abatzoglou et al., 2018).”

• cites Abatzoglou et al., 2018Not checked yet

  Not checked in the free run.

During the Black Saturday fires of 2009, convection from the initial fires led to the generation of additional lightning strikes that ignited fires in neighboring forests.

“For example, during the Black Saturday fires of 2009, which remain the deadliest Australian fire outbreak on record, convection resulting from the initial fires led to the generation of additional lightning strikes that ignited fires in neighboring forests (Dowdy et al., 2017).”

• cites Dowdy et al., 2017Not checked yet

  Not checked in the free run.

Lightning ignition rate is more strongly modulated by contemporaneous fire weather conditions than in the case of human-ignited fires.

“Lightning ignition rate is more strongly modulated by contemporaneous fire weather conditions than in the case of human-ignited fires (H.Clarke, Gibson, et al., 2019).”

• cites H.Clarke, Gibson, et al., 2019Not checked yet

  Not checked in the free run.

Human management of the landscape feeds back to patterns of fire extent, frequency, and seasonality.

“Further, human management of the landscape feeds back to patterns of fire extent, frequency, and seasonality.”

• cites Archibald et al., 2013Not checked yet

  Not checked in the free run.

• cites Bowman et al., 2011Not checked yet

  Not checked in the free run.

• cites Kelley et al., 2019Not checked yet

  Not checked in the free run.

There is no global human fire data set to date.

“Efforts to parse out whether humancaused fires are intentional or accidental focus on departures from natural fire regimes (seasonality and intensity), land use/land cover linkages, and syntheses of regional ethnographic literature, although these have primarily been regional in nature and there is no global human fire data set to date (Lauk & Erb, 2009;Lavorel et al., 2007).”

• cites Lauk & Erb, 2009Not checked yet

  Not checked in the free run.

• cites Lavorel et al., 2007Not checked yet

  Not checked in the free run.

Deforestation has legacy effects on the susceptibility of tropical forest edges to fire.

“Deforestation has legacy effects on the susceptibility of tropical forest edges to fire.”

• cites Chaplin-Kramer et al., 2015Not checked yet

  Not checked in the free run.

• cites Pütz et al., 2011Not checked yet

  Not checked in the free run.

Some of the great spatial diversity of human relationships with fire has been distilled into regional patterns by global-scale studies of the co-variability between satellite-observed BA and population density and land use.

“Nonetheless, some of the great spatial diversity of human relationships with fire has been distilled into regional patterns by global-scale studies of the co-variability between satellite-observed BA and population density and land use (Andela et al., 2017;Kelley et al., 2019).”

• cites Andela et al., 2017Not checked yet

  Not checked in the free run.

• cites Kelley et al., 2019Not checked yet

  Not checked in the free run.

The correlation between lightning and BA may be an artifact caused by covariance between lightning activity and human ignitions.

“Hence, the correlation between lightning and BA may be an artifact caused by covariance between lightning activity and human ignitions.”

The slash-and-burn agriculture practice has principally been concentrated in the tropical forests of southern Amazonia and Indonesia.

“The so-called”

• cites Aragão et al., 2018Not checked yet

  Not checked in the free run.

• cites Silva Junior et al., 2021Not checked yet

  Not checked in the free run.

• cites Curtis et al., 2018Not checked yet

  Not checked in the free run.

Temporally, human-caused wildfires expand wildfire seasons well beyond the peak burning period of the natural fire regime.

“Temporally, human-caused wildfires expand wildfire seasons well beyond the peak burning period of the natural fire regime (Balch et al., 2017;Le Page et al., 2010).”

• cites Balch et al., 2017Not checked yet

  Not checked in the free run.

• cites Le Page et al., 2010Not checked yet

  Not checked in the free run.

Across these regions, human-caused wildfire distribution relates strongly to spatial proximity to human infrastructure and density in general.

“Across these regions, human-caused wildfire distribution relates strongly to spatial proximity to human infrastructure and density in general.”

• cites Catry et al., 2009Not checked yet

  Not checked in the free run.

• cites Curt et al., 2016Not checked yet

  Not checked in the free run.

• cites Ganteaume et al., 2013Not checked yet

  Not checked in the free run.

Positive correlations between BA and population density and agricultural activity in southern Amazonia are likely linked to a mixture of deforestation fires and fire escape from agricultural land.

“Hence, the positive correlations between BA and population density and agricultural activity in southern Amazonia (Andela et al., 2017;Figure 10, Table 6) are likely linked to a mixture of deforestation fires and fire escape from agricultural land.”

• cites Andela et al., 2017Not checked yet

  Not checked in the free run.

Negative correlations between BA and tree biomass in Amazonia support a link between forest degradation and fire incidence in the region.

“Negative correlations between BA and tree biomass in Amazonia also support a link between forest degradation and fire incidence in the region (Figure 8; see Text A in Supporting Information S1 for methods).”

How and where humans today use fire intentionally as a tool of land stewardship, for deforestation, and for agricultural purposes creates distinct global patterns of high fire activity where climatic controls are secondary.

“How and where humans today use fire intentionally as a tool of land stewardship, for deforestation, and for agricultural purposes creates distinct global patterns of high fire activity where climatic controls are secondary (Archibald et al., 2013;Syphard et al., 2017).”

• cites Archibald et al., 2013Not checked yet

  Not checked in the free run.

• cites Syphard et al., 2017Not checked yet

  Not checked in the free run.

Amazonian BA trends have changed in sign several times since the 1990s in response to policy implementation, regulation and recent relaxation.

“Amazonian BA trends have changed in sign several times since the 1990s in response to policy implementation, regulation and recent relaxation (Aragão et al., 2018;D. Nepstad et al., 2014;Libonati et al., 2021;Tyukavina et al., 2017;Silva Junior et al., 2021; see Section 3.2).”

• cites Aragão et al., 2018Not checked yet

  Not checked in the free run.

• cites D. Nepstad et al., 2014Not checked yet

  Not checked in the free run.

• cites Libonati et al., 2021Not checked yet

  Not checked in the free run.

Similar dynamics to Amazonia are seen in Indonesian tropical forests.

“While many studies have focused on the impacts of humans on fire activity in Amazonia, similar dynamics are seen in Indonesian tropical forests.”

• cites Field et al., 2009Not checked yet

  Not checked in the free run.

• cites Nikonovas et al., 2020Not checked yet

  Not checked in the free run.

In regions with strong fire suppression and education campaigns (e.g., North America), the number of human-caused fires has declined significantly over the last half-century.

“Notably, in regions with strong fire suppression and education campaigns (e.g., North America), the number of human-caused fires has declined significantly over the last half-century (Coogan et al., 2021;Prestemon & Butry, 2005).”

• cites Coogan et al., 2021Not checked yet

  Not checked in the free run.

• cites Prestemon & Butry, 2005Not checked yet

  Not checked in the free run.

There was a >200% increase in mean daily ignitions in the US on 4 July Independence day, linked to widespread use of fireworks.

“Perhaps the starkest example of human-caused ignition anomalies is the >200% increase in mean daily ignitions in the US on its 4 July Independence day, linked to widespread use of fireworks (Balch et al., 2017).”

• cites Balch et al., 2017Not checked yet

  Not checked in the free run.

The relationships between burned area and the density of humans, cropland and livestock conform to a conceptual framework describing the socioeconomic drivers of fire in Africa.

“The relationships seen between BA and the density of humans, cropland and livestock were found to conform to a conceptual framework describing the socioeconomic drivers of fire in Africa”

• cites Andela et al. (2017)Not checked yet

  Not checked in the free run.

In tropical and temperate pastoral rangelands, pastoralists have ignited fires near annually for millennia to remove material from the prior growing season and stimulate new growth of native, fire-adapted herbaceous species.

“In tropical and temperate pastoral rangelands, such as the cerrado and pampas of South America, the tropical savannahs of Africa and Australia and the temperate prairies and steppe of mid-North America, southeast Australia, and central Asia, pastoralists have ignited fires near annually for millennia to remove material from the prior growing season and stimulate new growth of native, fire-adapted”

• cites Bernardi et al., 2019Not checked yet

  Not checked in the free run.

• cites Le Page et al., 2010Not checked yet

  Not checked in the free run.

• cites Magi et al., 2012Not checked yet

  Not checked in the free run.

Fire timing and frequency are controlled by management practices employed in permanent croplands.

“fire timing and frequency are controlled by management practices employed in permanent croplands”

• cites Archibald et al., 2009Not checked yet

  Not checked in the free run.

• cites Grégoire et al., 2012Not checked yet

  Not checked in the free run.

• cites Tiffen, 2006Not checked yet

  Not checked in the free run.

Indonesian forest fires predominantly occur in proximity to human infrastructure.

“Indonesian forest fires predominantly occur in proximity to human infrastructure (Field et al., 2009;Nikonovas et al., 2020).”

• cites Field et al., 2009Not checked yet

  Not checked in the free run.

• cites Nikonovas et al., 2020Not checked yet

  Not checked in the free run.

Large fire events were recorded in some Indonesian forests during drought years only once the human population had expanded into those regions, despite equally strong droughts in prior years.

“Field et al. (2009) showed that large fire events were recorded in some Indonesian forests during drought years only once the human population had expanded into those regions, despite equally strong droughts being recorded in the years prior.”

• cites Field et al., 2009Not checked yet

  Not checked in the free run.

Drivers of fire in African savannahs are largely responsible for the overall declining trend of global burned area.

“As more than half of global BA occurs in African savannahs, the drivers of fire in these regions are largely responsible for the overall declining trend of global BA.”

• AssumptionRests on an assumption: More than half of global burned area occurs in African savannahs.

In some semi-arid regions where BA was historically low, overgrazing has been cited as the source of land cover conversion from relatively fire-resistant shrublands to high-flammability exotic grasslands, leading to significant increases in fire frequency and BA.

“However, in some semi-arid regions where BA was historically low, overgrazing has been cited as the source of land cover conversion from relatively fire-resistant shrublands to high-flammability exotic grasslands, leading to significant increases in fire frequency and BA (Balch et al., 2013;D'Antonio & Vitousek, 1992;Olsson et al., 2012).”

• cites Balch et al., 2013Not checked yet

  Not checked in the free run.

• cites D'Antonio & Vitousek, 1992Not checked yet

  Not checked in the free run.

• cites Olsson et al., 2012Not checked yet

  Not checked in the free run.

Deforestation fires are generally ignited at the climatological driest point of the year in order to maximize tree mortality and consumption of biomass.

“Deforestation fires are generally ignited at the climatological driest point of the year in order to maximize tree mortality and consumption of biomass (Field et al., 2016;Le Page et al., 2010).”

• cites Field et al., 2016Not checked yet

  Not checked in the free run.

• cites Le Page et al., 2010Not checked yet

  Not checked in the free run.

In agricultural systems, particularly where monoculture cereal crops dominate, humans tend to use fire to remove crop residues from the prior growing season before planting or after harvest.

“In agricultural systems, particularly where monoculture cereal crops dominate, humans tend to use fire to remove the crop residues from the prior growing season before planting or after harvest (Aguiar et al., 2011;Hall et al., 2016;McCarty et al., 2007;Korontzi et al., 2006;Moreira & Pe'er, 2018).”

• cites Aguiar et al., 2011Not checked yet

  Not checked in the free run.

• cites Hall et al., 2016Not checked yet

  Not checked in the free run.

• cites McCarty et al., 2007Not checked yet

  Not checked in the free run.

Maintenance of dry croplands, such as cereal crops and legumes, reduces BA by altering fuel structure, timing of harvest, and the speed and efficacy of fire suppression efforts in valuable croplands.

“The maintenance of dry croplands, such as cereal crops and legumes, also reduces BA by altering fuel structure, timing of harvest (i.e., removing fuel prior to peak fire potential), and the speed and efficacy of fire suppression efforts in valuable croplands (Cruz et al., 2020;Moreira et al., 2009).”

• cites Cruz et al., 2020Not checked yet

  Not checked in the free run.

• cites Moreira et al., 2009Not checked yet

  Not checked in the free run.

Similar dynamics have been reported recently in African tropical forests.

“Similar dynamics have also been reported recently in African tropical forests (Z.Zhao et al., 2021).”

• cites Z.Zhao et al., 2021Not checked yet

  Not checked in the free run.

Nomadic pastoralism often promotes burning as a tool for land management.

“nomadic pastoralism which often promotes burning as a tool for land management”

• cites Archibald et al., 2009Not checked yet

  Not checked in the free run.

• cites Grégoire et al., 2012Not checked yet

  Not checked in the free run.

• cites Tiffen, 2006Not checked yet

  Not checked in the free run.