It's thirty-one degrees in Connecticut right now and my brother is visibly melting into the chair. Daniel's in the same heat, apparently, and he wrote in with a question that's been nagging at him. He's looking at his weather app, seeing the actual temperature, and then seeing that other number — the "feels like" — and wondering what's actually behind it. Is it calculated from temperature, humidity, and wind? Is there one common formula, or does it vary depending on which network's app you're using? And then he goes deeper — he wants to talk about thermal comfort as a concept, how it's calculated, how individual it really is. Because here's the thing he's noticed: people who live in humid climates eventually seem to get used to it. So how long does it take for the body to meaningfully adapt? That's the full spread. So let's dig into the numbers behind that little number on your phone.
The first thing to know is that "feels like" is not one thing. It's a consumer-friendly label that weather apps slap on top of what meteorologists call apparent temperature. And apparent temperature itself is a family of indices — different formulas for different conditions, developed at different times by different agencies.
So the apps are doing a bit of a branding exercise.
They are, and it's actually useful branding because most people don't want to read "heat index value of ninety-seven" or "wind chill of minus twelve." They want to know whether they need a jacket or whether they're going to keel over. But underneath the label, you're usually looking at one of two main things. In hot weather, it's the heat index, which combines temperature and relative humidity. In cold weather, it's wind chill, which combines temperature and wind speed. Some apps and national weather services use a more comprehensive apparent temperature formula that folds in solar radiation and wind across all conditions — Environment Canada does this, and some private networks borrow from it.
Daniel specifically asked whether the formula varies by network. So the answer is yes.
Yes, and it's worth understanding why. The heat index was developed by a researcher named Robert Steadman in nineteen seventy-nine. He published a long paper in the Journal of Applied Meteorology that modeled how the human body loses heat under different combinations of temperature and humidity. The core idea is straightforward: your body cools itself by sweating, and sweat works by evaporating. When the air is already saturated with moisture, evaporation slows down or stops. You keep sweating but it just drips off — it doesn't cool you. So the same air temperature feels dramatically hotter when the humidity is high.
The physics of your own sweat failing you.
Steadman built a mathematical model that takes air temperature and relative humidity and outputs what that combination "feels like" to a human body. NOAA adopted it and produced the simplified lookup table that most US weather services use. But here's a detail most people miss: the standard heat index assumes you're standing in the shade. If you're in direct sunlight, the apparent temperature can be eight to fifteen degrees Fahrenheit higher — about four to eight degrees Celsius — and that's not in the standard number.
So the "feels like" on your phone might be underestimating what you're actually experiencing if you're on a sunny sidewalk.
By a meaningful margin. Some weather providers account for this and some don't. AccuWeather, for example, uses its own proprietary formula called RealFeel that incorporates solar radiation, wind, humidity, and even the angle of the sun based on time of day and latitude. The Weather Channel uses a different blend. The BBC uses yet another. So when Daniel asks whether it varies by network — absolutely. Two different apps on the same street can give you "feels like" numbers that differ by three or four degrees.
Which is enough to change whether someone thinks "I'll go for a run" or "I'll stay inside."
And that's the public health piece. During the twenty twenty-one Pacific Northwest heat dome, heat index values were what drove emergency warnings. When the actual air temperature in Portland hit a hundred and sixteen Fahrenheit — forty-six point seven Celsius — the heat index was even worse because there was enough moisture in the air to make evaporative cooling ineffective. People who'd never experienced anything like it didn't know how to interpret what their body was telling them.
Let me pull on the cold side for a moment, because Daniel mentioned wind and humidity together. Wind chill is the cold-weather counterpart. How does that one work?
Wind chill was originally developed by two Antarctic researchers — Paul Siple and Charles Passel — in the nineteen forties. They hung water bottles outside in the wind and measured how fast they froze, then correlated that with wind speed and temperature. The idea was to quantify how wind strips away the thin layer of warm air that your body maintains next to your skin. The faster the wind, the faster that boundary layer gets replaced with cold air, and the faster you lose heat.
Water bottles. That's wonderfully low-tech.
It worked. But the original formula had problems — it was based on the freezing rate of water in a plastic cylinder, not on human skin. By the nineteen nineties, NOAA and Environment Canada jointly developed a revised wind chill index in two thousand one, using a more realistic model of human facial skin. They tested it on human subjects walking on treadmills in wind tunnels. The new formula is less extreme than the old one. A temperature of zero degrees Fahrenheit with a fifteen-mile-per-hour wind used to be reported as wind chill of minus thirty-nine. Under the two thousand one formula, it's minus nineteen. Still dangerous, but not "instant frostbite" territory. And here's the thing about humidity and cold — Daniel asked about it, and it does matter in cold weather, just less than wind. Very humid cold air can feel colder than dry cold air at the same temperature because water vapor conducts heat away from your body more efficiently than dry air. But wind chill dominates so heavily in cold conditions that humidity is usually a secondary correction.
So we've got heat index for hot and humid, wind chill for cold and windy, and then some apps use a unified apparent temperature formula that tries to cover everything. What's in that one?
Steadman again. He published a more comprehensive apparent temperature model that accounts for air temperature, humidity, wind speed, and solar radiation all at once. Environment Canada adopted a version of it, and the Australian Bureau of Meteorology uses something similar. The formula produces a single number that's supposed to represent what the air "feels like" to an average adult walking outdoors in appropriate clothing. It's the closest thing we have to a gold standard, but it's computationally expensive — it's a full heat-balance equation that models the human body's thermal equilibrium.
Computationally expensive in nineteen ninety. Today your phone does it in less time than it takes to render an emoji.
True. But weather services have legacy systems and they're conservative about changing formulas. Once you've trained the public on what "heat index one hundred five" means, you don't swap it for a new number without a long transition. So the simpler heat index persists in most US forecasts.
Let's talk about the limits of these formulas. You mentioned the shade assumption for heat index. What else do they miss?
A lot. The heat index is only defined for temperatures above eighty degrees Fahrenheit and relative humidity above forty percent. Below those thresholds, the formula doesn't apply and the "feels like" number you see is typically just the actual temperature or a different calculation. The wind chill index is only defined for temperatures at or below fifty degrees Fahrenheit and wind speeds above three miles per hour. In the gap between those two — a sixty-degree day with moderate humidity and a light breeze — there's no standard apparent temperature formula at all. Some apps just show the actual temperature. Some use an interpolation. There's no consensus.
So there's a whole range of perfectly pleasant weather where the "feels like" number is basically made up.
Or borrowed from a formula that wasn't designed for those conditions. And even within the defined ranges, the models assume a standard person. The heat index assumes a healthy adult of average height and weight, wearing light clothing, walking at about three miles per hour. If you're running, or you're elderly, or you're a child, or you're on medication that affects thermoregulation — the number on your phone is less accurate for you.
Which brings us to thermal comfort. Daniel asked about this directly, and it's the layer underneath all of this. The formulas are trying to estimate what a population experiences on average. Thermal comfort is about what you experience.
And it's a whole field of study. The central figure is a Danish researcher named P. Ole Fanger, who developed the Predicted Mean Vote model — the PMV — in the nineteen seventies. Fanger wanted to quantify thermal comfort in indoor environments, and his model is still the basis for the ISO standard that governs building climate control worldwide. The PMV takes six variables: air temperature, mean radiant temperature, air humidity, air speed, clothing insulation, and metabolic rate.
Clothing insulation and metabolic rate. Those are the ones that make it personal.
Right. Clothing insulation is measured in "clo" units. One clo is the insulation provided by a typical business suit — it keeps a resting person comfortable at about seventy degrees Fahrenheit. A pair of shorts and a t-shirt is about zero point three clo. A heavy winter coat is about two clo. Metabolic rate is measured in "met" units. One met is a person sitting quietly. Walking slowly is about two met. Running is eight to ten met. The PMV model takes all of this and outputs a number on a scale from negative three — cold — to positive three — hot — that predicts what the average person would say they feel.
The average person.
It is, and Fanger knew it. He also developed something called the Predicted Percentage Dissatisfied — the PPD — which estimates what fraction of people will be uncomfortable even at the "optimal" PMV. The lowest possible PPD is five percent. Even in a perfectly controlled environment designed to the ISO standard, one in twenty people will be too hot or too cold.
That's a beautifully honest metric. The best you can do is disappoint five percent of the room.
And in practice, it's much higher. Real buildings have drafts, sun-facing windows, hot equipment, variations in insulation. Office thermostat wars exist because the PMV model was calibrated largely on young, healthy, male subjects in laboratory conditions. Women typically have lower resting metabolic rates and tend to prefer ambient temperatures about two to three degrees Celsius warmer than men do.
So when an office sets the thermostat to what the ISO standard says is optimal, it's optimizing for a nineteen-seventies Danish lab subject.
Essentially a young man in a suit. And everyone else gets to be uncomfortable in slightly different ways. There's been a push in the last decade to revise thermal comfort standards to account for age, sex, body composition, and even cultural expectations. People who grow up in hot climates genuinely have different comfort ranges than people who grow up in cold ones, and it's not just psychological — there are physiological adaptations.
This is the part Daniel was really curious about. Acclimatization. How long does it take, and what's actually changing in the body?
The timeline is surprisingly well-studied. Heat acclimatization begins within the first few days of exposure and is substantially complete within seven to fourteen days. The changes are measurable. Your sweat rate increases. You start sweating at a lower core temperature — your body gets more proactive about cooling. Your sweat becomes more dilute because your sweat glands reabsorb more sodium, so you lose less salt. Your heart rate at a given workload drops because your plasma volume expands — you literally make more blood to improve heat transfer to the skin.
The body is basically upgrading its cooling system.
It's a remarkably elegant adaptation. A person who's been heat-acclimatized for two weeks can perform physical work in hot conditions with a much lower risk of heat exhaustion or heat stroke than someone who just arrived from a cool climate. The military and professional sports teams use deliberate heat acclimatization protocols — athletes or soldiers exercise in hot conditions for an hour or two a day, building up over ten to fourteen days.
But Daniel's question was specifically about humid climates. How does acclimatization work when the air is too wet for sweat to evaporate?
This is where it gets harder. In dry heat, sweat evaporates almost instantly, and the cooling is efficient. In humid heat, sweat pools on your skin. It doesn't evaporate. It doesn't cool you. Your body's response is to sweat even more — and that's partly adaptive and partly futile. You're losing water and electrolytes at an enormous rate, but you're not getting proportional cooling in return.
So you're just leaking.
You're leaking, and your body is working on the electrolyte conservation side of things — your sweat glands get better at reabsorbing sodium, your kidneys adjust — but the fundamental physics problem doesn't go away. Evaporation requires a vapor pressure gradient between your skin and the air. When the air is nearly saturated, that gradient is tiny. No amount of physiological adaptation can change the laws of thermodynamics.
There's a hard ceiling.
There is. The wet-bulb temperature is the measure that captures this. It's the temperature a thermometer reads when its bulb is wrapped in a wet cloth — the lowest temperature that evaporative cooling can achieve in a given air mass. When the wet-bulb temperature exceeds thirty-five degrees Celsius — about ninety-five Fahrenheit — the human body cannot cool itself at all, even with unlimited water and shade. A healthy person resting in those conditions will eventually die of hyperthermia. Acclimatization doesn't change that limit. It might buy you a little more time before organ failure, but not much.
And we've seen wet-bulb temperatures approaching that threshold.
We have. The twenty twenty-four Hajj pilgrimage in Mecca saw air temperatures above fifty degrees Celsius with high humidity. Hundreds of pilgrims died. These were people from all over the world, including many from hot climates, and the conditions simply exceeded what human physiology can handle. The wet-bulb temperature during that event was estimated to be in the high twenties to low thirties Celsius — approaching but not quite at the thirty-five-degree limit — and it was still lethal for vulnerable people exerting themselves outdoors.
So acclimatization is real, it's meaningful, but it has hard limits. What about the other direction? If you move from a humid place to a dry one, do you lose the adaptation?
You do, and fairly quickly. Heat acclimatization decays after about a week without heat exposure. After two to three weeks in a cool environment, most of the physiological adaptations are gone — plasma volume drops back down, sweat rate decreases, the earlier onset of sweating disappears. This is why people who grew up in, say, Singapore and then move to a temperate climate find that when they go back to visit, the first few days are brutal. They've lost the acclimatization.
The body is ruthlessly efficient about not maintaining adaptations it doesn't need.
Maintaining expanded plasma volume and elevated sweat gland activity costs energy. If you're not using it, your body drops it. It's the same reason athletes detrain when they stop exercising.
So let's pull this together. We've got weather apps giving us a "feels like" number that's some combination of heat index, wind chill, or apparent temperature, depending on the network and the conditions. The formulas are based on solid physics but they assume a standardized human — healthy adult, average build, appropriate clothing, shade. The actual experience varies from person to person based on metabolism, body composition, age, sex, and acclimatization state. And even the best-acclimatized body has an absolute ceiling set by wet-bulb temperature, beyond which no amount of sweating helps.
That's the picture. And I want to add one more layer. The "feels like" number you see on your phone is almost certainly not personalized. It doesn't know your age, your health status, whether you're on beta blockers that impair thermoregulation, whether you're pregnant, whether you've been in that climate for two days or two years. It's a population-level estimate being served to an individual. That's not a flaw in the science — it's just the nature of a mass-market forecast.
But as wearables get more sophisticated, you could imagine a personalized thermal stress index. Your watch knows your heart rate, your skin temperature, your activity level, your sweat rate if it's fancy enough. It knows you've been in this climate for three days and your acclimatization is partial. It could give you a "feels like" that's actually about you.
That's where this is headed. There are already research prototypes that combine personal physiological data with environmental data to predict individual heat strain. The military is interested for obvious reasons — soldiers carrying heavy loads in hot climates. Occupational health researchers are interested for outdoor workers. But the consumer version doesn't exist yet. For now, we're all looking at a number that was designed for an average person who doesn't exist.
I'm thinking about something you said earlier about the twenty twenty-one Pacific Northwest heat dome. People there weren't acclimatized. Their bodies hadn't made any of the adjustments. And the "feels like" number was telling them it was dangerously hot, but they didn't have the experiential reference point for what that number meant.
That's a really important point. The number is only useful if you know how to interpret it. In a place like Phoenix or Dubai, people have a lifetime of experience mapping "feels like one hundred ten" onto specific behaviors — stay inside between eleven and four, drink water constantly, watch for dizziness. In Portland in twenty twenty-one, a hundred and sixteen degrees with a high heat index was so far outside the range of local experience that people didn't have those mental models. The number was accurate but the cultural infrastructure around it was missing.
And that cultural infrastructure is going to need to spread as heat waves hit places that historically didn't get them.
It's already happening. Europe's heat wave in twenty twenty-three pushed "feels like" temperatures above fifty Celsius in parts of Spain. These are temperatures that would be extreme even in the Middle East, hitting populations that mostly don't have air conditioning and don't have generations of cultural knowledge about surviving extreme heat. The number on the app is the same number, but the preparedness gap is enormous.
Before we wrap up the main discussion, I want to go back to something you mentioned almost in passing. The heat index assumes you're in the shade. Direct sun can add up to fifteen degrees Fahrenheit. That's not a small correction — that's the difference between "unpleasant" and "dangerous." And yet I've never seen a weather app that surfaces that distinction clearly.
Some do. AccuWeather's RealFeel explicitly includes sun angle. But most apps just show the standard NOAA heat index, which is a shade value, and they don't tell you that's what it is. It's a communication failure. If the number is supposed to help people make decisions about outdoor activity, the shade assumption should be front and center.
Especially because if you're checking the weather to decide whether to go outside, you're probably planning to be in the sun.
Right. The use case and the assumption are misaligned. The heat index was originally designed as a public health tool — a threshold for issuing heat advisories. It was never meant to be a personalized outdoor activity guide. But it got adopted into weather apps as the "feels like" number, and now it's serving a purpose it wasn't designed for, with assumptions most users don't know about.
Which is a pretty good summary of a lot of consumer technology.
I walked into that one.
You did. So let's talk about the formula variations more concretely. Daniel asked whether it's a common formula or whether it varies by network. You said it varies. How much does it actually vary? If I check three different apps right now in the same location, what kind of spread am I likely to see?
In hot weather, typically two to five degrees Fahrenheit — one to three Celsius. It's usually not huge, but it can be larger at the extremes. The NOAA heat index table is a simplified polynomial approximation of Steadman's full model, and it has known inaccuracies at very high temperatures and humidities. Some weather providers use the full Steadman equation instead. Some add a solar radiation term. Some adjust for local conditions — elevation, proximity to water, urban heat island effects. The BBC's "feels like" includes a wind chill component even at moderate temperatures, while the standard US heat index doesn't.
So there's no regulatory body saying "this is the formula and you must use it."
There isn't. The World Meteorological Organization provides guidelines, and national weather services set standards for their own forecasts, but private weather companies are free to use whatever formula they want. The market pressure is mostly about accuracy and user trust — if your "feels like" number consistently seems wrong to users, they'll switch apps. But "seems wrong" is subjective, which is the whole problem we've been talking about.
It's a metric that's validated by the very perception it's trying to predict.
Yes. It's circular in a way that's hard to escape. You can validate the heat index against how people actually report feeling, but those reports are themselves influenced by acclimatization, clothing, activity level, and a dozen other factors. The science is real, but the precision is illusory. When your app says "feels like ninety-seven," it's giving you a point estimate for a distribution that's actually quite wide.
Hilbert, you've been quiet back there. You worked with some of this stuff, didn't you?
Hilbert: Clark Air Base, Philippines. Nineteen eighty-three to eighty-five. I was a weather observer.
Of course you were.
Hilbert: We had to calculate the heat index by hand every hour. Slide rule and a psychrometric chart — the big laminated one that curled up at the edges. You'd take the dry-bulb and wet-bulb readings from the instrument shelter, walk them over to the chart, trace the lines with your finger, and hope you didn't drift into the wrong column.
A slide rule. For heat index.
Hilbert: The slide rule had a little humidity correction dial on the side. It stuck. You had to tap it with your fingernail to get it to move, and half the time it overshot. I filed a maintenance request three times. Never got fixed.
So the heat index numbers you were reporting were... approximate.
Hilbert: One time I misread the chart. Reported a heat index of a hundred and twenty when it was really about a hundred and ten. That triggered a black flag advisory — all non-essential personnel restricted to quarters, outdoor training canceled for the entire wing. My CO was not happy when he figured it out.
How did he figure it out?
Hilbert: He walked outside and said it didn't feel like a hundred and twenty.
Validated by the very perception it was trying to predict.
Hilbert: I've been skeptical of these numbers ever since. Not the science — the science is sound. It's the human factor. The sticky dial. The guy who's been on shift for eleven hours squinting at a chart under fluorescent lights. The number on your phone is cleaner than anything we ever produced, but somewhere upstream there's still a human decision about which formula to use, which assumptions to bake in.
Do you still have the slide rule?
Hilbert: It's in a box in the garage. Next to a non-functioning theodolite and about forty feet of old anemometer cable.
Why do you have forty feet of anemometer cable?
Hilbert: I don't throw things away.
Was the dial actually broken, or was it just old?
Hilbert: It was sticky. I'm not saying it was broken. I'm saying it required a technique.
A technique.
Hilbert: You had to tap it just right. The new guys always overshot.
So the black flag advisory was a training issue.
Hilbert: I'm not here to relitigate the black flag advisory. I'm saying the number on your phone is probably fine. Just don't treat it like a thermometer reading. It's an estimate of an estimate, and somewhere in the chain there's a dial that sticks.
I'm going to be thinking about that sticky dial every time I check the weather now.
Hilbert: Good.
So next time you check the weather and see "feels like thirty-eight," remember — there's a century of physics and physiology behind that number, a slide rule in Hilbert's garage, and a whole lot of assumptions about what kind of body you have and what you're doing with it. It's not wrong. It's just more interesting than it looks.
And the gap between what the number says and what you actually feel — that's not the formula failing. That's you being a specific human instead of a statistical average. The question is whether future weather apps will start closing that gap by knowing more about you.
They will. The question is whether we want them to.
Thanks to our producer Hilbert Flumingtop, whose garage contains more retired meteorological equipment than most weather stations.
This has been My Weird Prompts. If you enjoyed the episode, leave us a review wherever you listen — it helps. You can find every episode at my weird prompts dot com, or email the show at show at my weird prompts dot com.
We'll be back soon. Stay hydrated.
