Explainer: How climate change is affecting wildfires around the world

This year has seen unprecedented wildfires cause havoc across the world. Australia recently battled its largest bushfire on record, while parts of the Arctic, the Amazon and central Asia have also experienced unusually severe blazes.

Since this article was first published, the western US has also faced intense fires, with the state of California experiencing its worst fire season since modern records began. The states of Oregon and Washington have also seen a spike in large wildfires in 2020.

It follows on from “the year rainforests burned” in 2019. Last year saw the Amazon face its third-largest fire on record, while intense blazes also raged in Indonesia, North America and Siberia, among other regions.

A rapid analysis released this year found that climate change made the conditions for Australia’s unprecedented 2019-20 bushfires at least 30% more likely. Further analysis – visualised below in an interactive map – has shown that, globally, climate change is driving an increase in the weather conditions that can stoke wildfires.

But despite a growing field of evidence suggesting that climate change is making the conditions for fire more likely, research finds that the total area burned by wildfires each year decreased by up to a quarter in the past two decades.

Understanding this paradox requires scientists to assess a vast range of influential factors, including climate change, human land-use and political and social motivations.

In this explainer, Carbon Brief examines how wildfires around the world are changing, the influence of global warming and how risks might multiply in the future.

• When and where are most of the world’s wildfires?
• How is climate change affecting wildfire risk?
• Interactive global map which locates 73 wildfire scientific studies
• Are wildfires increasing across the globe?
• How will wildfires change in the future?

When and where are most of the world’s wildfires?

At any given time, some part of the world is on fire. The map below, from the Global Forest Watch, shows where in the world has had a fire in the past 24 hours.

A “wildfire” can be defined as any type of uncontrolled fire that is spreading across wildland, including pastureland, forests, grasslands and peatlands. (Sometimes, fires are started intentionally and in a controlled manner, including during “prescribed burning”.)

Globally, wildfires have many impacts on humans, wildlife and the economy. Wildfires are a major driver of greenhouse gas emissions and are also responsible for 5-8% of the 3.3 million annual premature deaths from poor air quality, research suggests.

Some wildfires are started naturally, chiefly by lightning. The rest are started by humans, either accidentally or by arson.

Across the world, it is estimated that just 4% of fires start naturally. However, the proportion of human-started versus lightning-started fires varies widely from region to region. For example, in the US, 84% of fires are started by humans. However, in Canada, the majority (55%) of wildfires are started by lightning.

Many regions experience distinct wildfire seasons, driven by rainy and dry periods and human practices, such as agricultural burning. However, other regions have a risk of fire year-round.

The animation below, which uses data from NASA’s moderate resolution imaging spectroradiometer (MODIS), shows where in the world wildfires occurred in each month of 2019. 

The red, orange and yellow pixels on the maps show the locations where the satellite detected actively burning fires.

The colours represent a count of the number of fires observed within a 1,000-sq-km area. White pixels show as many as 30 fires in a 1,000-sq-km area per day. Orange pixels show as many as 10 fires, while red areas show as few as 1 fire in a day.

The animation shows how the average timing of fire seasons differs in different world regions.

North America, the Amazon, southern Africa and parts of Australia tend to see an uptick in fires from around August to late November. These months coincide with the height of the dry season in southern Africa and the Amazon.

By contrast, central Africa and parts of southeast Asia see most of their fires in December through to March. This coincides with the peak of the dry season in Africa’s Sahel region, which is north of the equator.

The year 2019 saw several large and – in some cases – unprecedented wildfires.

The UK had a record-breaking year in terms of total area burned by fire, with most its fires occurring in February to April. In June to July, more than 100 intense fires broke out across the Arctic Circle, mostly in Alaska and Siberia. August saw the Amazon battle its third-largest fires on record, while September saw large fires in North America and Indonesia. Towards the end of the year, Australia began to face its largest bushfires on record.

Australia’s unprecedented fire season continued into 2020, coming to a close at the end of March. The spring saw the return of fires to rural parts of the UK – likely boosted by an unusually hot winter, which left vegetation dry. The end of March also saw large fires break out in southwestern China, which killed at least 19 people and forced almost 25,000 more to evacuate.

From March to July this year, unprecedented heat in the Arctic fanned large “zombie fires” in Siberia. The fires ripped across vast stretches of permafrost, threatening the release of millions of tonnes of long-held carbon.

Several #Arctic locations recorded land surface temperatures hitting 45C on 19 June, according to initial data from the @esa satellite Sentinel-3.

Here is what you need to know… (THREAD 1/7) https://t.co/3Y7NW283Kn

— Carbon Brief (@CarbonBrief) June 23, 2020

Analysis by Dr Mark Parrington, a senior wildfire scientist at the EU’s Copernicus Atmosphere Monitoring Service, has found that the Arctic fires seen in June 2020 were larger and more intense than those seen in June 2019. However, both the 2019 and 2020 Arctic fires are far above the average level seen in the region from 2003-18.

The scale & intensity of #Siberia/#Arctic #wildfires in June 2020 has been greater than the 'unprecedented' activity of June 2019 in #Copernicus Atmosphere Monitoring Service GFAS data based on MODIS🛰️obs https://t.co/eb31VObkdM (total of 16 Mt C ≈ 59 Mt CO2) #ArcticFires pic.twitter.com/CR7VtSWSWh

— Mark Parrington (@m_parrington) July 1, 2020

Reacting to the analysis, fellow wildfire researcher Dr Thomas Smith, from the London School of Economics, posted:

“Will 2019 & 2020 be considered extremes? Or are we seeing the beginning of a new regime?”

June 2020 also saw the return of large fires in the Amazon. This was several months before the region’s usual “fire season”, which typically runs from August to to October, says Dr Ane Alencar, director of science at the Amazon Environmental Research Institute (IPAM) in Brazil. In June, she told a press briefing:

“I think it is very important and very worrisome to think about what is going to happen during this fire season.”

Fires in the Amazon remained unusually severe in July and hit a new high in August. In September, Unearthed reported that fires in Brazil’s Pantanal, the world’s largest wetland, had reached their highest level since records began in 1998.

At the same time, the western US entered into a record-breaking fire season. Since September, fires in California have been at their highest level on record – and the entire west coast is currently on track to see more land burned by fire in 2020 than in any other year since modern records began in the 1980s.

The gif below shows the cumulative area burned by fires across the west coast from 2000 to 2020. On the map, new area burned in each consecutive year is shown in red.

The US fires have killed at least 30 people and left millions suffering from unsafe air pollution from smoke across the west coast. The fires have also wiped out populations of endangered species, including half of Washington’s pygmy rabbit population, according to the New York Times.

Though fires happen all over the world, the largest and fastest-spreading fires mostly occur in sparsely populated grasslands in Australia, Africa and central Asia, according to NASA’s fire datasets.

The map below, put together by NASA, shows the average size of fires across the world from 2003-16. On the map, dark red spots indicate large fires.

Across all of the world’s grasslands, Africa sees a high number of large fires, explains Thailynn Munroe, a fire research analyst at the Global Forest Watch, an open-source forest monitoring service.

In fact, some researchers estimate that up to 70% of the world’s fires occur on the African continent. Munroe explains to Carbon Brief:

“Most of the fires that are started in Africa are for land-clearing and agriculture. A lot of it is either clearing for pastureland in sub-Saharan Africa or for agriculture in central Africa. I think that’s why we don’t see a lot of fires in Africa in the news because it’s more just part of the way of life there.”

The timing and whereabouts of fire is also influenced by the world’s major climate systems, such as El Niño, which periodically affects weather in many world regions, including southeast Asia and South America. Munroe explains:

“Fires in southeast Asia are greatly impacted by El Niño. In 2015-16, there was a really strong El Niño event which brought hot and dry weather. That was one of the reasons why Indonesia had such an intense fire season in that period.”

In 2015, Indonesia’s wildfires spiked, causing greenhouse gas release on the same scale as Brazil’s total annual emissions. The smoke from the fires led to 19 deaths and caused up to half a million people to suffer from respiratory illness, the Guardian reported. Fires in Indonesia have been worsened by the practice of draining peatlands.

How is climate change affecting wildfire risk?

There are several ways in which climate change can raise the risk of wildfires – and the importance of each of these factors varies from region to region, says Dr Cristina Santin, a wildfires researcher from Swansea University.

However, in general, one of the most important ways that climate change can increase the risk of severe fires is by causing vegetation to dry out, she says.

When temperatures are warmer than average, rates of evaporation increase, causing moisture to be drawn out from plants on the land. This drying can create “tinderbox conditions” – meaning that, if a fire is sparked, it can spread very quickly over large areas. 

Similar conditions can also be created by long periods of drought, Santin tells Carbon Brief:

“In the fire community, we call vegetation that is available to burn ‘fuel’. If you have a forest, not all vegetation is considered fuel because, under normal circumstances, it’s not going to burn. But if you have a huge drought or heatwave, a lot of that vegetation will be very dry and, therefore, it will become fuel.”

Such conditions occurred during the 2018 northern hemisphere heatwave, which saw all-time temperature records broken across Europe, North America and Asia.

In the Attica region of Greece, wildfires ripped across large swathes of dry land at lightning speed, causing people to rush to nearby beaches and into the sea. (A study covered by Carbon Brief found that the 2018 northern-hemisphere heatwave would have been “impossible” without human-caused climate change.)

Sustained hot temperatures were also a major driver of Australia’s unprecedented 2019-20 bushfires, says Dr Friederike Otto, acting director of the Environmental Change Institute at the University of Oxford.

Otto co-authored an analysis finding that temperatures during the bushfires were 1C to 2C hotter than they would have been in the early 20th century. She tells Carbon Brief:

“We found that climate change made the bushfires at least 30% more likely – and that is a conservative estimate. When we looked at temperatures alone, we found there has been a big increase in extreme temperatures in that region because of climate change.”

Warmer than average temperatures are also likely to be the primary driver of this year’s record-breaking fire season in the western US, the Atlantic reports. 

As well as making fires more severe, warming temperatures are also making fire seasons longer in some regions, explains Dr Megan Kirchmeier-Young, a researcher of climate extremes at the Government of Canada. (Fire seasons are a stretch of a time when a particular region is most likely to see large and intense blazes.) She tells Carbon Brief:

“Climate change is affecting wildfires in two main ways. The first is an increase in the risk or the likelihood of wildfire. The second is longer fire seasons – and this is mostly coming from warming temperatures.”

A study published in 2015 found that, globally, the number of days where wildfires are likely to burn has risen as a result of climate change.

The influence of climate change on fire seasonality is especially pronounced in regions that have seasonal snow cover, including parts of North America and northern Europe, she explains:

“In some regions, especially in Canada, we also consider when we lose winter snow cover. With warming temperatures, that’s going to be happening earlier in the year and it will also be later in the year before we see cooler temperatures and the return of snow cover. So a longer fire season means more time of the year where you could have fires.”

Research led by Kirchmeier-Young found that Canada’s 2017 fire season, which saw a record 1.2m hectares of land burned, was driven by “extreme warm and dry conditions” heightened by climate change. The study estimated that the total area burned across the season was made seven to 11 times larger by climate change.

Further south, in California, a research paper published in 2020 found that the number of autumn days with weather suitable for wildfires has doubled since the 1980s as a result of climate change.

The lengthening of fire seasons in California has had a knock-on effect on plant “phenology” – the timing of key events in plants’ life cycle, says Dr Frank Lake, a fire ecologist for the United States Forest Service. He tells Carbon Brief:

“We have seen due to climatic change, often hotter, drier conditions in California. We’ve seen longer fire seasons, which has also affected plant phenology, and we have extended periods of drought. That has definitely dried out the fuels [vegetation] and increased fire activity.”

In the Amazon, human-caused climate change and shifts to how people use the land have worked in tandem to greatly increase the risk of wildfires, says Alencar. In June, she told a press briefing:

“Unlike some forests in California, Florida, or in Australia, the Amazon doesn’t burn naturally. The natural fire regime in the Amazon is said to be somewhere between 500 to 1,000 years. Our results have demonstrated that in some places in the Amazon, actually, that fire regime has changed to 12 years. So, an area that should burn naturally every 500 to 1,000 years is burning every 12.”

This year, Science Brief – a UK-based web platform run by a team of scientists – released a review of 73 scientific studies finding that climate change is increasing the risk of wildfires at a global level. 

For the review, the scientists analysed all research papers published since 2013 that investigate the link between climate change and wildfires, either in a certain region or from a global perspective.

The researchers assessed how each study compared to the statement: “Climate change increases the risk of wildfires”. 

Carbon Brief’s interactive map below includes each of the 73 studies, which are displayed according to what part of the world they focus on.

Studies that fully support the statement that climate change increases wildfire risk are represented with a dark red icon, studies that mostly support the statement are represented in dark orange and studies that inform the statement but do not fully support it are represented in light orange.

Interactive map displaying 73 studies included in Science Brief’s review on wildfires and climate change. Data source: Science Brief. Map by Tom Prater and Joe Goodman.

(The software used to make the map currently only works with a Web Mercator projection – as used by virtually all major online map providers. It is worth noting that this – like all map projections – offers a somewhat distorted view of the world.)

The map shows how the majority of the 73 studies published since 2013 looking into wildfires either fully or mostly support the statement that climate change is increasing the risk of wildfires. 

It also shows that the vast majority of regional wildfire studies have been conducted in North America, Europe and Australia – with a visible lack of studies in the fire hotspots of central Africa and southeast Asia.

The review found strong evidence that climate change is making the weather conditions needed for wildfires more likely, says Dr Matthew Jones, a senior researcher of fire emissions at the University of East Anglia and lead author of the analysis. He tells Carbon Brief:

“Our main focus in the ScienceBrief review was on fire weather and how it is impacted by climate change. We found unequivocal evidence – based on observations and models spanning key global regions – that anthropogenic climate change has increased the flammability of landscape fuels and placed upwards pressure on fire activity.”

Among the research analysed are several “attribution” studies. “Attribution” refers to a fast-growing field of climate science that aims to quantify the “fingerprint” of climate change on extreme-weather events, such as wildfires, heatwaves and floods. (Both Otto and  Kirchmeier-Young’s studies are examples of attribution, for example.)

Analysis by Carbon Brief finds that, by the end of 2019, there had been 11 studies published that look into the role of climate change in single wildfire events. Out of these studies, 10 conclude that the fires analysed were made more likely or more severe by climate change. 

However, it is worth noting that all of these studies were carried out in either North America or Australia – with, again, no research looking into the role of climate change in fire hotspots, such as central Africa or southeast Asia.

In addition, there have been relatively very few studies into the role of climate change in wildfires when compared to other extreme events.

The chart below shows the number of studies that have been conducted into different types of extreme weather events, including heatwaves, floods and extreme rainfall events and droughts, among others.

On the chart, colour indicates the proportion of studies that found climate change made the event more severe or more likely (red), less severe or less likely (yellow), where climate change had no effect (blue) and where there was insufficient data to draw conclusions (grey).

Chart shows the number of studies for each type of extreme event that fall within each category of human influence: More severe/likely (red), less severe/likely (yellow), no influence (blue) and inconclusive (grey). Chart by Carbon Brief using Highcharts.

One reason why few wildfire attribution studies have been carried out is that fire risk is affected by multiple meteorological factors, says Kirchmeier-Young. These include temperature and rainfall, but also wind speed, which can influence the pace at which fires can spread, and soil moisture, which can influence the size of wildfires.

With so many different factors to consider, it can make it difficult to come up with a study framework that encompasses the whole picture of how climate change is affecting the probability and severity of a fire event, she says:

“Attribution generally requires you to define the event that you’re looking at. With a heatwave, you may look at maximum temperatures over a particular area, which you can easily get from a climate model. But with fire, there are a lot more variables that come into play – such as temperatures, precipitation levels, wind speed – and that can be a little more challenging.”

The complicated mix of factors affecting fire risk can also make it more difficult to source data, particularly in developing world regions, adds Otto:

“You can’t really do an attribution study if you don’t have decent observations. For fire, things like wind and soil moisture are also important, and in many parts of the world you have absolutely no observations for those variables.”

These challenges may have hampered efforts to carry out more wildfire attribution studies, Otto adds.

Are wildfires increasing across the globe?

With climate change raising the risk of hot and dry weather in many parts of the world, it may seem prudent to assume that the global area burned by wildfires each year is increasing.

However, several research papers looking into wildfires at a global level have come to the opposite conclusion. A paper published in the journal Science found that, globally, the area burned by wildfires decreased by 25% between 2003 and 2015.

The animation below, which uses data from the study, illustrates the pattern of how global area burned by wildfires changed over this time period.

The chart shows how the global burnt area zigzags throughout a year-long period. This is likely down to the influence of dry seasons and seasonal human practices, says Munroe. (For more information see: “When and where are most of the world’s wildfires?”)

However, despite peaks and troughs, an overall downward trend is seen until 2015. (And unpublished data from Global Forest Watch suggests that this downward trend continued up until 2020, Munroe says.)

There are several reasons why global burnt area could be decreasing at a global level despite the effect of warming, explains Santin, who co-authored a research paper looking into this paradox in 2016.

By far the most important factor is likely to be changes in how humans use the land, especially around the Sahel region of Africa, says Santin:

“It’s not all about climate change. We’ve seen a decrease in global burnt area in the last couple of decades and that is mostly because of land-use change. There are areas – mainly in Africa – that used to be savannah but, because of growing human populations, have been converted to agricultural, grazing and urban land.”

Savannah is a term for areas of grassland with sparse trees, which are typically found on the margins of the tropics. Such ecosystems cover half of Africa’s surface. For thousands of years, wildfires have been commonplace in African savannahs, with some native tree species evolving special characteristics in response to frequent fires, such as fire-resistant bark.

However, recent years have seen an increase in the conversion of savannah to agricultural land as nations seek to feed sub-Saharan Africa’s rapidly growing population. With less wild savannah left, sub-Saharan Africa has seen a decline in the number of large wildfires, says Santin.

The map below, which uses data from the Science study, illustrates how burnt area has shifted in different world regions from 2003-15. On the map, dark blue indicates a large percentage decrease in burnt area, where deep red indicates a large percentage increase.

The map features a wide blue band stretching across northern sub-Saharan countries from Senegal and the Gambia to Sudan and South Sudan, indicating percentage decreases in area burned by fire in this region. This is likely down to land conversion, says Santin.

Other decreases in burnt area, including in central South America, Madagascar and parts of southeast Asia may also be driven by the conversion of wildland to agricultural land, she adds.

The authors of the Science paper also attribute decreases in burnt area to the conversion of savannah to agricultural land. The scientists, led by Dr Niels Andela at the NASA Goddard Space Flight Center, write:

“A shift toward more capital-intensive agriculture has led to fewer and smaller fires, driven by population increases, socioeconomic development, and demand for agricultural products from regional and global markets.”

Decreases in developed parts of the world, including northern Australia, could be down to improved fire management techniques, adds Munroe.

In northern Australia, firefighters use “prescribed burning” – a technique pioneered by indigenous communities involving the burning of small patches of grassland early on in the dry season. This creates “fire breaks” in the landscape, which can prevent fires from quickly spreading over large areas of wildland.

In the US, there has been dispute over the relative degree to which forest management techniques and climate change have contributed to a recent surge in wildfires in regions such as California.

For centuries, Native Americans managed fire risk in woodlands using prescribed burning. However, the arrival of non-native settlers saw the use of prescribed burning decrease, causing the landscape to shift to being more dominated by trees over the past century. Lake, who is of indigenous American Indian descent, tells Carbon Brief:

“At the core of the understanding of indigenous knowledge is how to live with fire.”

A study published in Science in 2006 found that the accumulation of trees – also known as “fuel loading” – and human-caused climate change are both important factors in increased fire risk across the western US. Lake says:

“We’ve reached a tipping point. There’s too much fuel loading and it’s now too warm, the ability of ignition is greater with both lightning and human-set fires, and now we have these much larger and more extensive fires…Almost universally tribal elders have said this is something that is unprecedented in our cultural legacy of knowledge.”

It is also worth noting that human land-use change can greatly increase the area burned by wildfires.

For example, recent surges in deforestation in the Amazon have been linked to greatly increased fire activity. Alencar explains:

“Fires in the Amazon usually are the last stage of deforestation. It is the most inexpensive tool that people have to convert biomass into ashes so that [they can use the land] for some pasture fields.”

According to NASA scientists, the Amazon’s severe 2019 fire season was “more consistent with land clearing than with regional drought”.

“Slash-and-burn” clearing has also been linked to damaging peat fires in Indonesia.

Fires in the country have been exacerbated further by the practice of peatland draining. In order to grow palm oil and other crops, such as timber, peatlands are often drained of their natural moisture – leaving them dry and more likely to catch alight. 

The influence of human land-use activity on global burnt area makes it an imperfect metric of how climate change is truly affecting wildfires, Prof Anthony Westerling a fire and drought researcher at the University of California, Merced, told Carbon Brief in 2018.

Another way that scientists could study the influence of climate change is by tracking fire severity.

Research published in 2017 analysed all “extreme wildfire events” from 2002-13 and found that 96% occurred during periods of hot and dry weather. (“Extreme wildfires” were deemed to be fires that caused serious economic or social damage.)

However, there is little research into how fire severity has changed at a global scale in recent decades. (A string of research papers have linked increasing fire severity in North America to climate change.)

How will wildfires change in the future?

Climate change will continue to drive temperature rise and more unpredictable rainfall in many parts of the world, meaning that the number of days with “fire weather” – conditions in which fires are likely to burn – is expected to increase in coming decades, says Kirchmeier-Young:

“As we continue to see increasing temperatures, we will continue to see an increase in the likelihood of wildfires. So, more days with increased fire risk, longer fire seasons.”

A study covered by Carbon Brief found that, by the middle of the century, there could be a 35% increase in the days with a high danger of fire across the world, on average – if little action is taken to tackle climate change.

The regions likely to see the highest increase in days with extreme fire weather in this very high emissions scenario (“RCP8.5”) include the western US, southeastern Australia, the Mediterranean and southern Africa, according to the study.

This is highlighted on the map below, which shows the projected change in the number of days with “high” fire danger from 2000-14 to the middle of the century (2014-70). The regions with the largest increase are shaded orange and red, while the areas with decreasing risk are shown in green.

Other studies have investigated how the risk of fire weather is likely to change in different world regions if the world does take action to tackle climate change.

For example, a study published in 2020 found that, in California, “climate change will further amplify the number of days with extreme fire weather by the end of this century”. However if efforts are taken to limit global warming to below 2C, which is the goal of the Paris Agreement, this “would substantially curb that increase”, the authors say.

In the Mediterranean, limiting global warming to 1.5C, which is the aspirational target of the Paris Agreement, could halve the total area burned by wildfires in the summer, when compared to a scenario where warming reaches 3C, according to a second study.

In Australia, days with fire weather akin to that seen during the 2019-2020 bushfires could become at least four times more likely under 2C of global warming, according to Otto’s recent analysis.

A landmark special report in 2018 from the Intergovernmental Panel on Climate Change (IPCC) concluded (pdf) that limiting global warming to 1.5C rather 2C would “reduce” the average wildfire risk worldwide.

However, though the risk of wildfires is likely to heighten in coming decades, it is not yet clear whether the area burned by wildfires will increase correspondingly, says Santin:

“It’s a complicated issue because it’s not only climate change controlling future fire risk, it’s also the interaction of humans and climate change.”

It is possible that efforts by humans to suppress wildfires could stem increases in the area burned by fire, despite the increased risk posed by climate change, she says.

However, it may be the case that, in some areas, the risks posed by climate change will become too large to overcome with adaptation measures, she adds:

“Something interesting that we’ve seen with the recent Australian fires is that, even in areas where efforts to suppress wildfires were undertaken, the fire conditions were so crazy – the temperatures so hot, the winds so strong – that the fires burned through every protective measure.”

Indeed, a research paper published in 2016 found that, globally, human efforts could suppress wildfire increases in the near future, but that these efforts may become ineffective later on in the century. The authors write:

“For future scenarios, global burnt area will continue to decline under a moderate emissions scenario…but start to increase again from around mid-century under high greenhouse gas emissions.”