It’s not just you: Summers are getting smokier.
For the third year in a row, cities like Detroit, Minneapolis, Boston, and New York are experiencing dangerously polluted air for days at a time as smoke drifts into the U.S. from wildfires in Canada.
Smoke has traveled to these places in the past, Stanford University researcher Marshall Burke told me. But the data is clear that the haze is becoming more severe.
“The worst days are worse,” said Burke, “and you can see that in the averages, the last couple of years are much, much higher across the Midwest and the East Coast than we’ve observed in the past many decades.”
Burke is one of the leading scholars studying wildfire smoke, investigating everything from its effect on air quality, public health, and behavior, to preventative and adaptive public policy responses. In one of his most recent papers, which has not yet been peer reviewed, he and his co-authors analyzed the influence of smoke on air quality over the past two decades, using satellite imagery of smoke plumes to disentangle how much of the fine particulate matter, or PM2.5, measured by air monitoring stations came from fires versus more typical sources like cars and furnaces.
The study shows a sharp increase in the amount of smoke in the air around the U.S. in just the past few years. From 2020 to 2023, the average American breathed in concentrations of smoke-related PM2.5 that were between 2.6 and 6.7 times higher than the 2006 to 2019 average.
The paper also contains a stunning set of charts that show that wildfires are eroding decades of air quality gains — and the efficacy of air quality regulation in general — and that without these smoke events, PM2.5 levels would have been significantly lower.
Courtesy of Marshall Burke
I caught up with Burke to better understand what we know about this seemingly sudden escalation of smoke events, and what we can do to better protect ourselves from them moving forward. Our conversation has been lightly edited for clarity.
Given the smoke events we’ve seen in the last three years, can we say anything about the next three years?
I don’t think you want to make bets on any specific years. The long run trend, unfortunately, suggests that the last few years are going to be more representative than the sorts of years we got 10 to 15 to 20 years ago. And that is due to the underlying physical climate that’s warming and drying out fuels and making fire spread faster and fires much larger. Larger fires generate more smoke.
Has it all been driven by Canadian wildfires?
No. The East Coast and the Midwest will get exposure from fires as far as California, often in the Northern Rockies. But the recent very bad exposure — 2023 was by far the worst year in the Midwest and East Coast — that was nearly all from Canadian fires. This year, again, it’s nearly all from Canadian fires.
Why is that?
The reason we’ve seen a lot more Canadian fires is the same reason we’ve seen a lot more fires in the U.S. West — increasing fuel aridity. As temperatures warm, forests dry out. And so when you get lightning strikes, which tend to start most of the large fires in Canada, you get faster fire spread and much larger fires.
Interestingly, we’ve seen in Canada fewer total fires over time. Often I see people posting this on Twitter — Climate change is not a problem, we’re getting fewer fires in Canada — and that’s true. I think they’ve reduced other sources of ignitions. But you still get lightning ignitions.
Burned area has gone the other way — you’ve seen an increase in burned area. So, fewer fires, but much larger fires, and these larger fires are the ones that put out a lot more smoke, and the smoke gets pushed into population centers in Canada and into the U.S.
There were really large wildfires in California before 2023. Why weren’t places on the East Coast having smoky days as a result of those?
It’s the way the wind blows and how far it has to go. In the large 2020 and 2021 fire seasons we had in the U.S. West, some of that smoke certainly was making it to the East Coast, but given the prevailing wind patterns and the distance the smoke had to travel, the influence of those fires on air quality was not as big as the recent Canadian fires.
Are there other events that cause comparable air quality degradation to wildfires?
You can get really specific things — if a train crashes and lights on fire and a given town is exposed to really high levels of whatever pollutant for a few days. Sometimes you can get dust events that have broad scale exposure. But basically never do you reach the AQI levels that we see in wildfires. Wildfires are pretty unique in their ability to expose very large numbers of people to a very high level of pollutants for days, or unfortunately now, weeks, at a time. Nothing else compares in the U.S.
If you go to other parts of the world where you have large anthropogenic sources — Indian cities, Chinese cities — it can be quite different. There’s some exceptions. Salt Lake City and places where you get inversions and you get pollution trapped for many days, you can get pretty high levels of exposure, but typically nowhere close to what you get during these acute wildfire events.
When the AQI goes back down to levels that are more common in a city after a smoke event and people feel safer going outside, are you able to measure how much of the PM2.5 remaining in the air is from a wildfire? Does it matter?
We try to measure that directly — on any given day, how much of the PM that you’re experiencing is from wildfires versus from other sources. What you see is these events can turn on really quickly, and they can also turn off really quickly, either because the wind direction changes or because it rains — if it rains, you rain out a lot of these pollutants, and then you’re breathing mostly clean air right away.
We also try to measure, how does human health respond? One thing that science doesn’t give us a crisp answer to yet is, is one day of 100 micrograms better or worse than 10 days of 10 micrograms of exposure? We don’t actually really know. What we do see is people respond very differently to those two scenarios in ways that likely affect health outcomes. On really bad days, people tend to stay inside. In California, total emergency department visits go down instead of up, and that’s because people are not getting in their cars, they’re not getting in car accidents, they’re not spraining their ankle playing football or whatever because they’re staying at home.
On lower smoke days, we see emergency department visits go up. That’s probably because people are not changing their behavior. But, maybe surprisingly, we still don’t have a crisp answer if you’re thinking about asthma or mortality or other cardiovascular outcomes.
What are some of the other questions researchers are trying to answer as this becomes more of a national issue?
All sorts of things. The immediate health impacts that you think about — respiratory outcomes have been the one that’s been measured best in a lot of different settings. Cardiovascular outcomes, I would say the evidence is surprisingly more mixed on that. There’s a long-standing literature that shows cardiovascular mortality impacts of exposure to PM, but for wildfire PM, specifically, that evidence is less clear. Sorting that out and trying to understand whether there are differences is important.
Cognitive outcomes — does it increase your risk of dementia? Does student learning go down? Does it reduce cognitive performance at work? I think there’s emerging evidence that smoke is pretty important. Exposure to air pollution, more broadly, is important, but wildfire smoke, specifically, can impact these outcomes.
Birth outcomes is another one we and others have looked at. You see a pretty clear signature of wildfire smoke in birth outcomes — increases to the risk of pre-term birth, for instance. We used to just think about sensitive populations as elderly populations or people with pre-existing conditions. And basically what the research is showing is, no, actually, everyone is sensitive in some way. The list of people who are likely affected probably includes most, if not all of us.
What are the potential policy responses to this in places that haven’t had to deal with it in the past?
I think there’s three policy buckets. This is more true in the U.S. than Canada, but our fire problem is a combination of a warming climate and a century of fire suppression that has left abundant fuel in our landscapes, so number one is dealing with climate change as best we can, and two is doing something about the accumulated fuel loads. There’s a lot we can do there — prescribed burning is one approach that we and others are studying a lot; mechanical thinning, where you go out and actually remove the fuel. Understanding when and where to do that and what the benefits are is an ongoing scientific challenge, but I think most of the evidence would suggest we’re going to need a lot more of that than we’ve done, historically.
But even if we do a lot of that, we’re going to get more of these smoke events, unfortunately. And so we need to protect ourselves when these events happen. Indoor air filtration works really well, so we need to make sure people have access to filters of various types. The evidence would suggest that we see health impacts even at pretty low levels of exposure, and so if you have a portable filter — I drive my family crazy, I’m turning ours on all the time. You should basically just be running them all the time.
What about in terms of messaging? I’m thinking about city officials or state officials, when a smoke event is coming — and maybe this is still an active area of research — but what’s the current thinking on what message to send to people?
Yeah, I think it is an ongoing area, in terms of exactly how to do this and who to target with the information. The way we typically do this is to set these thresholds, right? So, above some threshold, you get a notice, and below, you don’t. That is understandable.
But what we see in the data is that there’s not some level below which you’re fine and above which you’re screwed. What we see is the more smoke you’re exposed to, the worse off you are, and so our goal should just be to reduce our exposure as best we can. How to message that effectively is not something we have a crisp social scientific answer to yet.
A lot of the advice has historically been that you should stay at home with your windows and doors closed. In California homes that is not very protective because California homes tend to be not very tight. In my view, just telling people to close their windows and doors is not sufficient for protecting health. They need some sort of active filtration — portable air filter, central air — to do that.
The other thing that’s happened in California, and I’ve seen this with my own kids — should we cancel school on really bad days? The assumption is that kids are better protected at home than they would be in the school environment, and that’s just not obviously true. It could be the case that for many kids, schools are better. We don’t know, because we do not have comprehensive measurement of indoor air quality, and this is a huge failing that we need to fix. Just as we measure it pretty comprehensively outside, we’ve got to do the same thing inside, and we just haven’t done this.
Courtesy of Marshall Burke
