Rex Havencroft is not a real person, though his job very much is. He's a composite drawn from conversations with National Weather Service forecasters, EPA air quality specialists, and the growing cadre of meteorologists who've found themselves in the unexpected business of tracking atmospheric poison. We met him, hypothetically speaking, on an October afternoon when Manitoba wildfire smoke had turned the Minneapolis sky the color of old pennies.

You have one of those jobs that didn't really exist fifteen years ago.

Not in this form, no. I mean, air quality forecasting existed, but it was mostly about urban pollution. Car exhaust, industrial emissions, that kind of thing. Predictable stuff. Summer smog season.

Now? Now I'm essentially a smoke detective. I spend my days staring at satellite imagery of fire plumes, calculating whether aerosol layers are at 2,000 feet or 200 feet, because that difference determines whether I'm issuing a "maybe stay inside" advisory or a "do not go outside under any circumstances" warning.¹

The weird part is how much of the job is about vertical geometry. People think smoke just spreads. But it's three-dimensional. A massive plume can be sitting at 15,000 feet doing absolutely nothing to you, or it can be hugging the ground turning your neighborhood into a gas chamber. My whole day is "where is the smoke layer and is it descending."

What does that actually look like? The monitoring.

So NOAA satellites are constantly measuring aerosol optical depth, which is basically how much particulate matter is between the satellite and the ground. That feeds into EPA's air quality models and our forecast guidance. But models are just models. They're wrong all the time, especially with smoke because fire behavior is chaotic and atmospheric mixing is complicated.

What I'm really doing is combining the satellite data with ground-based sensors, radar, surface observations, and honestly just... looking at webcams. I have probably forty webcam bookmarks. Highway cameras, park cameras, random business rooftop cameras. Because the models might say the smoke is aloft, but if I can see brown haze at ground level in Duluth, I know we've got a problem.

The satellite product we rely on most is aerosol layer height. It tells us whether smoke is near the surface or not.² One of my colleagues calls it "the uncertainty killer" because without it, you're basically guessing. With it, you can say with confidence: this smoke is going to hurt people today, or this smoke is going to look scary but stay overhead.

And then you issue warnings that people ignore.

laughs Yeah. Yeah, that's the job.

I mean, not everyone ignores them. Vulnerable populations, people with asthma or COPD, they take it seriously. Schools usually follow our guidance. But go to any park when we've got a 200 AQI and you'll see people jogging. Jogging! In air that looks like someone's grinding coffee beans made of smoke.

I've started keeping a mental tally. Last summer, during the big Canadian fire event, we had 300+ AQI for three days straight. I issued warnings every six hours. And I watched on traffic cameras people having outdoor birthday parties. Grilling. There was a youth soccer game in St. Paul that they played through until halftime when some kid's inhaler stopped working.

That must be...

Maddening? It's complicated. Part of me understands it. Smoke events can last for weeks now. You can't just stop your life. People have jobs, kids have activities, life continues.

But part of me is screaming internally because I know what PM2.5 does to your cardiovascular system.³ I see the ER admission data afterward. There's always a spike in cardiac events, respiratory distress, acute renal failure. Always.

The other thing is, and this is going to sound terrible, but there's a class element. The people jogging through 250 AQI are usually affluent folks who have air purifiers at home, who can afford to take sick days if they need to, who have good health insurance. The people really getting hammered are outdoor workers. Construction crews. Agricultural workers. Delivery drivers. People who don't have a choice.

California has workplace protection rules now for wildfire smoke.

They do! If AQI for PM2.5 hits 151 to 500, employers have to provide N95 masks and training. Above 500, masks are mandatory.⁴ Which is great in theory.

In practice, I've watched construction continue through 400 AQI with maybe half the crew wearing masks, and wearing them incorrectly. A loose N95 doesn't do much.

And those rules only apply in California. Most states have nothing. So I'm issuing warnings in Minnesota knowing that the legal framework to actually protect people doesn't exist. I'm just hoping people will care enough about their lungs to make good choices.

Do you think about the fires themselves? The people evacuating?

All the time. This year, 2025, Manitoba and Saskatchewan have been absolutely catastrophic. Way above average for fires and acreage burned.⁵ I'm tracking smoke from fires that destroyed communities, that forced mass evacuations. There are people who lost everything, and my job is to track the atmospheric aftermath of their disaster as it drifts south.

How do you communicate that? The "all options are bad" reality?

Badly, mostly. Our official warnings are very technical. "Air Quality Alert issued for the following counties. PM2.5 concentrations are expected to reach Unhealthy for Sensitive Groups. Limit outdoor exertion." It's accurate but it doesn't convey the actual stakes.

What I want to say is: "The air will hurt you today. If you have any choice at all, stay inside. If you must go out, understand that you're accepting cardiovascular risk. Your body will remember this exposure."

But we can't say that. Too alarming. Might cause panic. So we say "limit outdoor exertion" and people interpret that as "maybe skip the marathon but jogging is fine."

The other communication nightmare is duration. Smoke events last for days, weeks sometimes. People can't process that. They're used to weather warnings that resolve quickly. "Tornado warning, take shelter, okay it passed." But smoke? "The air is bad today. It will be bad tomorrow. It will probably be bad next week. We don't know when it will end." That's psychologically brutal.

The worst is when we get smoke from prescribed burns. Controlled fires that are supposed to reduce wildfire risk. Forest managers plan them for optimal weather conditions, good dispersal, minimal impact.⁶ But "minimal" still means smoke, and people get furious. I'll get emails: "Why are they burning on purpose? My kid has asthma!" And I'm like, I understand your anger, but the alternative is catastrophic wildfire that's ten times worse. We're choosing between bad options.

Do you have air purifiers at your house?

laughs Three of them. HEPA filters, the whole setup. I check the AQI before I walk my dog. I'm the guy at parties explaining why you should close your windows even though it's 85 degrees out because the indoor air quality will be better than outdoor even if it's hot. I'm super fun at parties.

My wife jokes that I've become a smoke monk. She's not wrong. I think about smoke the way medieval monks thought about sin: it's everywhere, it's invisible, it's corrupting, and most people are in denial about it.

What changes in the next five years?

More of this. Definitely more of this. Fire seasons are lengthening, fires are getting bigger, and the atmospheric conditions that trap smoke near the surface are becoming more common. We're going to see more multi-week smoke events. More 300+ AQI days in places that rarely saw 150 before.

The forecasting will get better. Our models are improving, satellite data is getting more sophisticated. But I don't think behavior will change much. People adapt to new normals frighteningly fast. In five years, 200 AQI will feel routine. People will check the AQI the way they check the weather now. "Oh, it's only 180 today, not bad."

The healthcare system is what worries me. Emergency rooms are already strained during major smoke events. As this becomes more frequent, more prolonged? I don't know how the system absorbs that. We're going to see triage decisions based on air quality. "Sorry, we can't treat your non-emergency condition today because we're overwhelmed with smoke-related cardiac events."

Do you ever want to quit? Do something that doesn't involve watching disasters unfold in real-time?

Sometimes. Definitely sometimes. There are days when I'm tracking a massive plume headed for a major metro area and I'm thinking, "I'm going to issue a warning, people are going to ignore it, and then I'm going to see the hospital data showing I was right." It's exhausting being Cassandra.

But then I think about the people who do listen. The mom who keeps her asthmatic kid inside because of my warning. The construction company that gives their crew the day off. The elderly couple who decide not to do yard work. Those people exist. I can't see them, can't count them, but they're out there making better decisions because I provided information they needed.

I know how to read the satellite imagery, I understand atmospheric dynamics, I can translate complex data into actionable information.

It's just... this wasn't what I thought I'd be doing when I studied meteorology. I thought I'd be forecasting rain and snow. Instead I'm forecasting poison and watching people breathe it anyway. Welcome to the future, I guess.

Footnotes

1. https://www.nesdis.noaa.gov/news/wildfire-smoke-and-air-quality ↩

2. https://www.nesdis.noaa.gov/news/wildfire-smoke-and-air-quality ↩

3. https://www.epa.gov/emergencies-iaq/wildfires-and-indoor-air-quality-iaq ↩

4. https://www.dir.ca.gov/dosh/doshreg/Protection-from-Wildfire-Smoke/Wildfire-smoke-emergency-standard.html ↩

5. https://www.nesdis.noaa.gov/news/wildfire-smoke-and-air-quality ↩

6. https://www.nifc.gov/programs/smoke-management ↩