Wind chill calculator: the NWS formula, and what it leaves out

What is wind chill?

Wind chill is an attempt to answer a question a thermometer cannot: not how cold the air is, but how cold it feels on your skin. Moving air strips away the thin layer of warmth your body maintains around itself, so it draws heat out of you faster than still air at the same temperature. The index turns that into a single number — the temperature that would feel the same if the air were calm.

The version used across the United States and Canada was adopted jointly by the two countries' weather services for the winter of 2001/2002, replacing an index built on research the Antarctic explorers Paul Siple and Charles Passel did in the 1940s. The current one was calibrated on twelve volunteers in a wind tunnel with sensors on their faces, which is a rather better basis than the plastic cylinder of water the old one used.

This calculator uses that formula, in whichever units you prefer. It also does two things most wind chill calculators do not: it tells you when the conditions you entered are outside the range the index is defined for, and it shows you the size of the thing the index does not account for at all.

How to use it

1. Pick your units. Fahrenheit with miles per hour, or Celsius with kilometres per hour. Both answers come from the same calculation, so they cannot disagree with each other.
2. Enter the air temperature and the wind speed. Use the actual thermometer reading, not a previous wind chill figure. The sample buttons load the National Weather Service's own worked example plus four edge cases worth seeing.
3. Read the panels below the number. One says whether your reading is inside the definition. One says whether exposed skin could actually freeze — which depends on the air temperature, not the wind chill. And one gives the range the index does not model.

The number carries a range nobody prints

Here is the thing that ought to be on every wind chill calculator and is on almost none of them. The National Weather Service, on the same page where it publishes the formula, states that bright sunshine may increase the wind chill temperature by 10 to 18 degrees Fahrenheit.

That is not a rounding note. It is a variable the index does not model at all, and its size is enormous relative to the precision everyone quotes the answer to. A figure given as minus nineteen point four is really somewhere in a band nearly twenty degrees wide, depending on whether you are standing in sun or shade — and the tool reporting it to one decimal place has told you nothing about that.

It matters that this comes from the body that publishes the formula rather than from someone arguing against it. The index is useful; it is a good deal better than the one it replaced, and it is calibrated on real people rather than a can of water. But it answers a narrower question than the single confident number suggests, and a calculator that hides the caveat is selling a precision the source itself disclaims.

Two edges most calculators walk straight past

The index has bounds, and the National Weather Service states them in one sentence: it is defined only for temperatures at or below 50 degrees Fahrenheit and wind speeds above 3 miles per hour. Read that carefully. The temperature bound includes its endpoint and the wind bound does not, so exactly 3 miles per hour is outside the definition while exactly 50 degrees is inside. Enter a calm summer afternoon into most calculators and you will get a confident answer regardless; this one computes the same number and then tells you it means nothing, because the formula does not refuse — it just keeps going.

The second edge is a genuine misconception, and the National Weather Service addresses it directly. Frostbite depends on the air temperature, never on the wind chill. Wind cannot bring the temperature below freezing when the thermometer says it is above; a wind chill of minus ten with the air at forty will not freeze anything, though it can still give you hypothermia. This calculator judges that question on the air temperature alone, which is why a reading can show a wind chill well below freezing and still say frostbite is not possible.

The same logic explains why wind chill does not apply to objects. It describes the rate at which a warm body loses heat, so it means nothing for a car radiator or a water pipe. Wind cools them faster, but never below the actual air temperature — the National Weather Service gives the example itself: at minus five degrees with a wind chill of minus thirty-one, the radiator still stops at minus five.

Honest limitations

This is the North American index. Other countries publish different cold-stress measures entirely, and a number from one is not comparable with a number from another. If you are reading a forecast from outside the United States or Canada, check which index it uses before comparing it with anything here.

You will sometimes see comparisons between this formula and the pre-2001 one, usually with a dramatic figure attached. This page does not make that comparison, and the reason is worth stating: the retired formula does not appear in any of the National Weather Service documents that describe the change. The history is well sourced — the 1945 index, the joint replacement, the wind tunnel — but the old equation itself is not, so any number comparing the two would rest on something unverified.

Finally, the Celsius and Fahrenheit answers here are one calculation rather than two. Environment Canada publishes the formula with different constants for metric units, and it is genuinely the same equation with the numbers rounded for the other scale — checked across more than ten thousand combinations, the two versions never disagree by more than about a twentieth of a degree. Presenting them as two separate formulas would suggest a choice where there is not one.

Why is it free?

It is one equation with four constants, and your browser evaluates it as you type. No server is involved, so there is nothing to meter and no account to create.

Nothing is uploaded. The numbers you enter stay in this tab.