Thanks for the in depth analysis as always. I’ve been wondering that since wildfires have increased in size and intensity over the last 20 years, whether there is now less forest to burn, and maybe that is why we are seeing a downturn in some areas, such as Greece?
I made a wild hairy ass guess that less forest to burn was not an explanation, then asked ASS (Artificial Stupidity, Stupid) if I got lucky. I think I did.
"During Greece's worst wildfire season in modern history in 2007, fires destroyed an estimated 4% of the country's total forest area."
"From 2001 to 2025, Greece lost 320 kha of tree cover, equivalent to a 9.0% of the 2000 tree cover area. This does not account for gains in tree cover over the same period."
So Greece has had more or less the same amount of forest to burn every year since 2001¹.
Thanks for putting this in the proper contexts. The simple one-chart-fits-all may work for snarky social media takes but you’ve demonstrated here that there are many nuances to the truth.
While I sometimes or often disagree with you on how serious the situation is or what action we should take, I constantly use your data and presentation thereof as it is not contended and accepted by everyone in the debate.
This is such a silly debate. If wildfires are making your cities unlivable, then they are a problem, regardless of whether there are more or less than last year.
It's far from silly, because what matters is the trend. If there are more wildfires than last year, and that was more than the year before that, the solution is different than if there are fewer year by year. As it happens, everything else equal, wildfires will keep growing worse until the global economy is decarbonized. IOW, decarbonization is necessary, though not sufficient.
Great post. To answer to the questions "Is Europe having a bad year for wildfires?" you really have to ask "compared to what?". Residents of Bordeaux probably take little comfort in knowing that residents of the Balkans are having a quiet year for fires. Across time? Cold comfort to someone whose house burned today that it was only in flames one out of the last 7000 days.
I find quantifying the trend of fires - especially the leading edge where fires emerge in areas not historically seen as at risk - is tricky as they are, by definition, very isolated events in a long tail through which we can only glimpse at the broad underlying pattern. The rising risk, driven by climate or otherwise, is theoretical until it happens.
(that said - EFFIS also maintains a Fire Danger Index, which maps what areas in Europe are at risk of wildfire impact, based on weather, fuel, presence of human settlements, etc. We should be more worried about changes here than through the fragmented lens of recorded fires).
A confounding factor is that the EU has signalled a change in its fire management approach from reactive to preventative, which will likely involve prescribed burns and allowing wildfires to burn where safe to do so. This could increase the number of small fires seen, but (hopefully) reduce large / destructive / dangerous fires. How do we compare the situation before and after this policy shift?
I think more and more Europeans will have to grow used to the presence of wildfires as normal - and the question of "worse" shifts from one of incidence to one of impact.
just a note. A forest burned last year cannot burn this year since it cannot regrow so fast. So comparing year to year is not so relevant. If all the forest of France will burn in 2026, in 2026 wildfires would not occur!
This stuff is fiendishly complicated to get your head round. I was looking last week and I think Balkans was quite bad last year, which might explain it being quieter this year?
I’ve mainly been using Effis because it has a Rapid Damage Assessment for burnt area function that is helpful for news coverage - the number it gave for Dunwich was put to local fire services and resulted in them swiftly updating their tally on the ground to higher than Effis.
There you go again, Hannah, with your fancy "facts." Get with the vibes - they world is on FIRE! Because crying wolf at every moment on every topic is the way to get clicks, and clicks are all that matter.
Not so. With the -h is the primary British spelling, without, the primary American one. Hannah being British, it stands to reason that she would use the most common way of writing of her variety of English.
That's the description. Now we need an analysis of why fire risk shifts in space and time across European countries (and elsewhere) -- temperature, rainfall, agriculture, forestry....
"The speed and intensity of wildfire outbreaks is arguably more important for the impact on communities than just the total area burned."
This is true, but how much of this is because a community has done certain things that mean wildfires are more likely, more dangerous?
The history of the South West of France is that they planted maritime pine in the 18th century to creates the Landes forest. It used to be marshy. The pine gave land, made it easier to farm. It also meant productive use. Wood, paper, pine resin.
There are no fires in the inland parts of the Landes forest. Why is that? Same global warming as Cap Ferret, if anything, slightly warmer. The difference is that the inland part is commercial. They still chop down trees to produce pine. The woodland around Cap Ferret is a nature reserve. Its purpose is not to produce wood but to create a habitat for wildlife, to give tourists somewhere nice to go on a bike.
There are also no fires in the Limousin. Again, mostly commercial forest.
The forests were tree farms for over 100 years. It's mass tourism along the coast that ended that. And when someone owns a forest, and is selling the wood, they have big incentives to stop fires. They chop down trees and sell them, and are mindful of creating firebreaks. Gaps between sections of the forest to prevent fire spreading.
One somewhat related question about the data: In mid May, several outlets (I remember "Der Spiegel" in Germany) reported that more than 150 million ha had burned globally from January to April 2026 (including 44 million ha in Asia).
These were described as record figures and attributed to GWIS data, but when I look at the current data on OWID, the numbers have changed quite a bit. For example, Asia is only at around 19 million ha even by week 23 and, as you say in the post 2026 now looks like a very low year globally so far.
Am I missing some difference between the datasets or definitions being used or some other obvious reason? If not, was there any statement or explanation for the revisions or are corrections of this size relatively common?
Thanks for the in depth analysis as always. I’ve been wondering that since wildfires have increased in size and intensity over the last 20 years, whether there is now less forest to burn, and maybe that is why we are seeing a downturn in some areas, such as Greece?
I made a wild hairy ass guess that less forest to burn was not an explanation, then asked ASS (Artificial Stupidity, Stupid) if I got lucky. I think I did.
"During Greece's worst wildfire season in modern history in 2007, fires destroyed an estimated 4% of the country's total forest area."
Then I checked a reliable source - https://globalnaturewatch.org/dashboards/country/GRC/
"From 2001 to 2025, Greece lost 320 kha of tree cover, equivalent to a 9.0% of the 2000 tree cover area. This does not account for gains in tree cover over the same period."
So Greece has had more or less the same amount of forest to burn every year since 2001¹.
¹ Subject to correction by Hannah.
Thanks for putting this in the proper contexts. The simple one-chart-fits-all may work for snarky social media takes but you’ve demonstrated here that there are many nuances to the truth.
Thank you.
While I sometimes or often disagree with you on how serious the situation is or what action we should take, I constantly use your data and presentation thereof as it is not contended and accepted by everyone in the debate.
Keep up the good work.
Thanks for this important work. Great graphs and explanations 👏
Thank you so much. This puts everything in context well as always.
Please keep doing what you're doing.
I recently introduced a statistician friend of mine to your posts and they're consuming your back catalogue with great pleasure!
This is such a silly debate. If wildfires are making your cities unlivable, then they are a problem, regardless of whether there are more or less than last year.
It's far from silly, because what matters is the trend. If there are more wildfires than last year, and that was more than the year before that, the solution is different than if there are fewer year by year. As it happens, everything else equal, wildfires will keep growing worse until the global economy is decarbonized. IOW, decarbonization is necessary, though not sufficient.
Great post. To answer to the questions "Is Europe having a bad year for wildfires?" you really have to ask "compared to what?". Residents of Bordeaux probably take little comfort in knowing that residents of the Balkans are having a quiet year for fires. Across time? Cold comfort to someone whose house burned today that it was only in flames one out of the last 7000 days.
I find quantifying the trend of fires - especially the leading edge where fires emerge in areas not historically seen as at risk - is tricky as they are, by definition, very isolated events in a long tail through which we can only glimpse at the broad underlying pattern. The rising risk, driven by climate or otherwise, is theoretical until it happens.
(that said - EFFIS also maintains a Fire Danger Index, which maps what areas in Europe are at risk of wildfire impact, based on weather, fuel, presence of human settlements, etc. We should be more worried about changes here than through the fragmented lens of recorded fires).
A confounding factor is that the EU has signalled a change in its fire management approach from reactive to preventative, which will likely involve prescribed burns and allowing wildfires to burn where safe to do so. This could increase the number of small fires seen, but (hopefully) reduce large / destructive / dangerous fires. How do we compare the situation before and after this policy shift?
I think more and more Europeans will have to grow used to the presence of wildfires as normal - and the question of "worse" shifts from one of incidence to one of impact.
just a note. A forest burned last year cannot burn this year since it cannot regrow so fast. So comparing year to year is not so relevant. If all the forest of France will burn in 2026, in 2026 wildfires would not occur!
it meant "If all the forest of France will burn in 2026, in 2027 wildfires would not occur!"
This stuff is fiendishly complicated to get your head round. I was looking last week and I think Balkans was quite bad last year, which might explain it being quieter this year?
I’ve mainly been using Effis because it has a Rapid Damage Assessment for burnt area function that is helpful for news coverage - the number it gave for Dunwich was put to local fire services and resulted in them swiftly updating their tally on the ground to higher than Effis.
Fires in Europe are breaking records: here’s what the future holds
By James Dinneen
https://www.nature.com/articles/d41586-026-02361-3?utm_source=Live+Audience&utm_campaign=10e2bbd987-nature-briefing-weekly-20260731&utm_medium=email&utm_term=0_-33f35e09ea-49657168
There you go again, Hannah, with your fancy "facts." Get with the vibes - they world is on FIRE! Because crying wolf at every moment on every topic is the way to get clicks, and clicks are all that matter.
</s>
Interesting post, as usual. But... please, please, it is savanna, not Savannah. The latter is a city in Georgia USA.
Not so. With the -h is the primary British spelling, without, the primary American one. Hannah being British, it stands to reason that she would use the most common way of writing of her variety of English.
https://www.oxfordlearnersdictionaries.com/definition/english/savannah_1
https://www.merriam-webster.com/dictionary/savanna
I've taught Geography for over 40 years in the UK. The correct spelling for the ecosystem is savanna.
That's the description. Now we need an analysis of why fire risk shifts in space and time across European countries (and elsewhere) -- temperature, rainfall, agriculture, forestry....
"The speed and intensity of wildfire outbreaks is arguably more important for the impact on communities than just the total area burned."
This is true, but how much of this is because a community has done certain things that mean wildfires are more likely, more dangerous?
The history of the South West of France is that they planted maritime pine in the 18th century to creates the Landes forest. It used to be marshy. The pine gave land, made it easier to farm. It also meant productive use. Wood, paper, pine resin.
There are no fires in the inland parts of the Landes forest. Why is that? Same global warming as Cap Ferret, if anything, slightly warmer. The difference is that the inland part is commercial. They still chop down trees to produce pine. The woodland around Cap Ferret is a nature reserve. Its purpose is not to produce wood but to create a habitat for wildlife, to give tourists somewhere nice to go on a bike.
There are also no fires in the Limousin. Again, mostly commercial forest.
It was necessary to turn the forests into tree farms to save them. /s
The forests were tree farms for over 100 years. It's mass tourism along the coast that ended that. And when someone owns a forest, and is selling the wood, they have big incentives to stop fires. They chop down trees and sell them, and are mindful of creating firebreaks. Gaps between sections of the forest to prevent fire spreading.
Thank you, I really enjoyed that post!
One somewhat related question about the data: In mid May, several outlets (I remember "Der Spiegel" in Germany) reported that more than 150 million ha had burned globally from January to April 2026 (including 44 million ha in Asia).
These were described as record figures and attributed to GWIS data, but when I look at the current data on OWID, the numbers have changed quite a bit. For example, Asia is only at around 19 million ha even by week 23 and, as you say in the post 2026 now looks like a very low year globally so far.
Am I missing some difference between the datasets or definitions being used or some other obvious reason? If not, was there any statement or explanation for the revisions or are corrections of this size relatively common?
Arrived via a skeet from Chris Baraniuk. Helpful and informative read. Thank you