On a recent cold morning in Washington DC, I looked up the weather. The temperature was 38°F, but with winds occasionally gusting to 8 miles per hour, the wind chill was officially 32°F. Freezing.
Except it wasn't actually freezing. There weren't any puddles on the streets turning into ice. The precipitation that was falling was clearly coming down as rain. And Weather Underground reported that it "felt like" 36°F. The wind chill indicator gave a misleading picture of what things were really like outside.
There's a good reason for that: Wind chill simply doesn't mean what most people think it means.
The wind chill index is designed for a very precise, very narrow purpose. "It was developed solely to assess the risk of frostbite on unclothed parts of the body," says Krzysztof Blazejczyk, a Polish researcher who studies the thermodynamics of the human body. In other words: If the temperature is 38°F and the wind chill is 32°F, that means you'd develop frostbite on exposed skin just as quickly as you would if the temperature was 32°F and there was no wind. That's it. This formula also assumes you'll be walking directly into a steady wind continuously, with your face totally bare.
Those are very particular conditions, and they don't really describe our full range of experiences outside. So, more often than not, wind chill dramatically exaggerates the cold we actually feel.
This is not a secret. Many people have pointed this out over the years. In 2007, Slate's Daniel Engber suggested that "rather than trying to patch up wind chill's inconsistencies, we should just dump it altogether."
The real mystery, then, is why weather forecasters continue to use wind chill — even though most experts know that it's wildly flawed. In recent years, scientists have developed a number of superior alternatives that try to measure what it actually feels like to be outside, taking into account temperature, wind, sunlight, humidity. Examples include the Universal Thermal Climate Index (UTCI) or proprietary metrics like Weather Underground's "feels like."
So why haven't these alternative metrics caught on? And why can't we ever seem to quit wind chill?
How wind chill became so popular
The core idea behind wind chill was first developed in the 1940s by Paul Siple and Charles Passel, a pair of American scientists working in Antarctica. It was long known that wind caused objects to lose heat more quickly, by blowing away the layer of warmer air that surrounds them. Siple and Passel tried to measure this effect by studying the freezing rates of water bottles placed on top of their hut in Antarctica.
These calculations helped them develop what they called the "wind chill factor." For years, this metric was used mainly by scientists, because Siple and Passel expressed it in units of kilocalories per hour per square meter — a technical measurement of heat loss that was lost on most people.
Then in the 1960s, as part of an effort to better prepare soldiers for missions in cold climes, US military researchers hit upon the idea that would make wind chill famous. Why not phrase this obscure metric in terms of temperature? "Persons unaccustomed to the [wind chill] index and its history have had some trouble in using it," wrote Charles J. Eagan, a researcher at an Alaska Air Force lab, in 1964. "A way out of this difficulty for the practical user has been sought by expressing any wind chill value as an 'equivalent temperature.'"
This turned out to be a hit. Within a decade, TV and radio meteorologists were using wind chill to convey how it felt to be outside when the wind was blowing — a practice that persists today.
There was just one problem: This was never how wind chill was intended to be used. "The Antarctic experiments it was based on were very primitive," says Maurice Bluestein, an Indiana University engineer. "So when [weather forecasters] took that data and applied it to humans, it assumed totally unrealistic temperatures."
Over time, experts tried to make wind chill slightly more suitable for mass consumption — although they could never get rid of all of its flaws. One day in the 1990s, while digging out a driveway in Indianapolis, Bluestein noticed that it was surprisingly warm outside, despite the fact that meteorologists had described as a -60°F wind chill. "After a few minutes of shoveling, I was taking off my gloves and hat. It just didn't seem to be that cold," he says.
So Bluestein decided to work to improve wind chill, at the behest of the National Weather Service. In 2001, collaborating with Canadian scientist Randall Osczevski, he published the revised wind chill formula, the one we still use today. This work was based off more sophisticated models of heat loss and the human body, and it involved some experiments done with real people. As a result, it produced wind chill temperatures that weren't nearly as absurdly cold as the old formula.
Still, even Bluestein admits that the newer, updated formula doesn't fit for everyone. And it remains an imperfect gauge of what it actually feels like to be outside.
Why even modern wind chill formulas are flawed
Bluestein's updated wind chill formula is relatively simple: You plug in the temperature and wind speed, and it spits out your risk of frostbite.
But this formula is based on the assumption that each one of us inhabits the same body — about 5-foot-6, and heavyset, with the exact same size face — and therefore each of us loses heat at the same rate. It stems in part on a small experiment conducted in Canada in which 12 people walked on treadmills in a cold wind tunnel, with thermometers stuck to their faces to measure how quickly they lost heat.
This experiment actually showed that different people lose heat at dramatically different rates. "A person with more body fat, for instance, is actually at greater risk of frostbite, because heat is trapped within the body more effectively, so less reaches the skin," Bluestein explains. Still, the National Weather Service wanted a simple index for warning people about frostbite, so he and colleagues based it on people in the 5th percentile for heat loss, the worst-case scenario.
This is one of many oversimplifications that are now baked into the wind chill formula we use daily. It assumes the sun isn't shining at all, and that whenever you're outside, you're constantly walking at a speed of about 3 miles per hour straight into a steady wind. "If you are standing still or running," writes Canadian meteorologist Brad Vrolijk, "the wind chill number produced by that equation is not valid for you."
Moreover, the wind speeds generally used to calculate wind chill come from airport weather stations, but as Bluestein notes, "if you're walking in an urban environment, buildings and trees are going to cut down wind speed. At the airport, there's nothing blocking it."
Bluestein's formula was undoubtedly an improvement on the 1950s-era formula, and fulfilled the National Weather Service's request for a simple metric that doesn't require a huge set of variables (it also errs on the side of caution in warning people about frostbite). But even the new formula dramatically exaggerates what it "feels like" to be outside.
So why are we still using wind chill?
In recent years, plenty of alternate "feels like" metrics have been developed: some proprietary, like AccuWeather's RealFeel, and others developed by scientists and public weather officials, such as the Universal Thermal Climate Index (UTCI).
"UTCI takes into account ambient temperature, ambient humidity, solar radiation levels, wind speed, and combines it all with the level of clothing you'd expect someone to be wearing," says George Havenith, a British physiological researcher who helped develop the model. (Interestingly, the index assumes — based on studies of real-world behavior — that people tend to underdress for the cold.)
The UTCI formula may not be perfect, but recent studies have shown that it aligns pretty closely with the temperature people report that it "feels like" under various conditions. At the moment, though, the only national weather service that routinely uses it is Poland's.
So what accounts for its limited use — and wind chill's continued dominance? One factor is that the National Weather Service continues to promote it. And with good reason: The wind chill index is optimized for worst-case scenarios, so it gets people to take the threat of frostbite seriously.
And because the NWS always publishes it, it's the easiest metric for broadcasters to use. Technically, there's nothing stopping them from developing their own "feels like" weather metrics, the way Weather Underground does. But doing so takes more work, partly because those metrics are based on numbers — like sunlight and humidity — that vary widely over the course of the day and a person's precise location. "The more variables you put in, the harder it is for people to utilize," notes Bluestein.
Of course, you could argue that wind chill is easier to use specifically because it oversimplifies, covering up a huge amount of variation with a single number. You might even argue that UTCI and other advanced metrics do so, too, because of the inherent variability from person to person — and the impossible nature of assigning a number to the subjective question of what it "feels like" for you to be outside.
That, anyway, is why some meteorologists and weather buffs want to do away with these sorts of "feels like" metrics entirely, relying solely on a single number that tells you exactly what it promises: the temperature. "[Wind chill] is a parameter which has huge variation from person to person and depends significantly on how we dress, what our environment is like and on many factors where there will always be some level of uncertainty," writes Vrolijk, the Canadian meteorologist. "While we have gotten used to explicit declarations of the wind chill value, reality is far more fuzzy."