Dew Point Calculator

e_s(t) = 611.2 · exp(17.62 · t / (243.12 + t)); e = RH/100 · e_s(t); t_d = 243.12 · ln(e / 611.2) / (17.62 − ln(e / 611.2))

t

The air temperature, in °C by default, with °F in the same box. It is the temperature of the air whose dew point you want, not the temperature of the surface you are worried about — the comparison between the two is the whole point, and the page only knows the first one. Kelvin is deliberately absent: nobody reports an air temperature or a dew point in kelvin, and adding the unit would invite a reading in the wrong scale

RH

Relative humidity as a percentage, 0 to 100, and an input rather than an output on this page. It is defined as the actual vapour pressure divided by the saturation vapour pressure at the same temperature, so 60% means the air is holding six tenths of the water vapour it could hold at that temperature. Two temperatures with the same relative humidity have different dew points, and the same air has a falling relative humidity as it warms — which is exactly why the dew point is the more portable number of the two

e_s(t)

The saturation vapour pressure at the air temperature: how much water vapour the air can hold before it starts to condense, in pascals. It grows steeply with temperature, roughly doubling every 10 °C in the range people live in — around 6.1 hPa at 0 °C, 23.3 hPa at 20 °C, 42.3 hPa at 30 °C. That steepness is the reason a warm room feels muggy at a humidity that would be unremarkable outdoors in winter, and it is the first table on this page

e

The actual vapour pressure in the air, in pascals: the saturation value scaled by the relative humidity. It is the third output, printed in hectopascals because that is how meteorology and building services quote it. It is also the quantity that gets compared, because the dew point is defined as the temperature at which e_s equals this e — cooling the air does not change how much vapour is in it, it lowers the ceiling above it until the two meet

t_d

The dew point in °C: the temperature at which the air's current vapour content would be exactly saturating. Below it, water condenses out; at it, the relative humidity is 100% by definition. The main result, printed in both °C and °F. When the dew point sits close to the air temperature the air is nearly saturated and evaporation is slow — which is why a 26 °C dew point on a 30 °C day feels oppressive while the same relative humidity on a cool morning does not

17.62 / 243.12

The Magnus constants, which turn the exponential relationship between temperature and saturation vapour pressure into something a calculator can do in one line. They are fitted empirical values rather than defined ones, quoted to about ±0.6% over the −45 to +60 °C range, and that range is where this page's input limits come from. They describe saturation over a water surface. Below 0 °C, saturation over ice follows a slightly different curve, with constants 22.46 and 272.62 — see the note on frost points below

Use it when you need to know how close the air is to condensing: whether a cold water pipe, a window or a chilled beam will sweat, whether the air in a room will feel muggy, whether a paint or a coating will cure properly, or whether the air being fed into a building needs dehumidifying. It is also the number to quote when the relative humidity is not comparable between two places — a 70% reading in a warm room and a 70% reading in a cold one describe very different amounts of water, while the dew points say it directly. In a building services context the dew point sets the floor under the chilled water temperature: run water colder than the dew point through a pipe in humid air and the pipe will drip. The second table is the quick-reference version, a grid of air temperatures against humidities, which is usually how the question is asked — what is the dew point when it is 25 °C and 70% out?

1. The default: 20 °C at 60% relative humidity

   1. Air at 20 °C, relative humidity 60%
   2. Saturation vapour pressure at 20 °C: e_s = 611.2 × exp(17.62 × 20 / 263.12) = 2333 Pa
   3. Scale it by the humidity: e = 0.60 × 2333 = 1400 Pa, printed as 14.00 hPa
   4. Invert for the temperature at which 1400 Pa would be saturating: t_d = 12.0 °C
   5. Convert for the second row: 12.0 × 9/5 + 32 = 53.6 °F

   This is the reading worth memorising, because it is the one that comes up: a room at 20 °C and six tenths humid has a dew point near 12 °C, so any surface colder than about 12 °C will collect water. That is why single-glazed windows run with water on the inside in winter and why a cold drink sweats in summer — the surface is below the dew point of the air touching it, and the air in that thin layer has been cooled past what it can hold. It is also why the answer is not 'the room is 60% humid, that is fine': 60% is a statement about the air, while 12 °C is a statement about the surfaces in it.

2. A muggy summer day: 30 °C at 80%

   1. Air at 30 °C, relative humidity 80%
   2. Saturation vapour pressure at 30 °C: 4234 Pa (the first table's 42.34 hPa row)
   3. e = 0.80 × 4234 = 3387 Pa → 33.87 hPa
   4. t_d = 26.2 °C, so the air is only 3.8 °C from saturation
   5. In Fahrenheit: 79.1 °F

   A dew point of 26 °C is what a muggy day actually is, and the number explains the feeling better than the humidity does: the air is within four degrees of being unable to hold its water, so sweat evaporates slowly and every cold surface in the room is below the dew point. Weather reporting treats 20 °C as the line where it starts to feel sticky, 24 °C as oppressive, and above 26 °C as the point where outdoor activity gets genuinely risky — those thresholds are stated in °C here and in the high 60s and 70s in °F, which is why both rows are printed. Note what did not change: the vapour pressure, 33.87 hPa, is more than twice the 14.00 hPa of the previous example, and that doubling is the entire difference between a comfortable room and a stifling one.

3. Crossing 0 °C: air at the freezing point, 50%

   1. Air at 0 °C, relative humidity 50%
   2. Saturation vapour pressure at 0 °C: 611 Pa → 6.11 hPa
   3. e = 0.50 × 611 = 306 Pa → 3.06 hPa
   4. t_d = −9.2 °C — the dew point is below freezing while the air is not
   5. In Fahrenheit: 15.4 °F

   The case that catches people out: the air is at the freezing point and the dew point is nearly ten degrees below it, so this is a perfectly ordinary winter reading rather than a contradiction. It is also the case where the page's choice of constants starts to matter. −9.2 °C is a dew point computed over a water surface, which is what weather reports quote and what building services design against. The temperature at which frost would actually form on a cold surface is the frost point, and it is slightly different — for the same vapour pressure, saturation over ice happens a little closer to 0 °C than saturation over water does, so the frost point is about 1 °C above the dew point here. If you need that number, the same formula runs with the ice constants 22.46 and 272.62 in place of 17.62 and 243.12.

4. Saturated air: 20 °C at 100%

   1. Air at 20 °C, relative humidity 100%
   2. At 100% the actual vapour pressure equals the saturation value: 2333 Pa → 23.33 hPa
   3. The temperature at which 2333 Pa is saturating is 20 °C — the air temperature itself
   4. So the dew point equals the air temperature exactly, and the Fahrenheit row reads 68 °F

   At saturation the dew point and the air temperature are the same number, and that is not a coincidence or a rounding artefact — it is the definition. Cool saturated air by any amount and water comes out; warm it and the relative humidity falls below 100% with no water added or removed. This makes the 100% case the one to sanity-check a dew point calculator against, because any algebraic slip in the inversion — the two constants swapped, the ratio inside the logarithm upside down — shows up here as a number that is not equal to the input, while every other case just gives a number that is quietly wrong.

Below 0 °C the number this page returns is a dew point over a water surface, and it is not the same as the frost point. The two curves differ because ice and supercooled water have different saturation vapour pressures, and the difference is small — about 1 °C at −10 °C — but it is real, and it decides which one you should be quoting. This page uses the water-surface constants throughout (17.62 and 243.12), because weather reports and building services quote the water-surface value and because switching between the two would mean the page had to solve for the dew point before it knew which set of constants to use. If you want the frost point, run the same formula with 22.46 and 272.62, which are valid from about −65 °C to just below freezing and carry a wider tolerance of about ±1%. Both sets are Magnus-type fits: they are empirical, quoted to roughly ±0.6% over −45 to +60 °C for the water pair, which is why the temperature inputs are limited to that range. Saturation over water is also a different question from saturation in a porous material: the dew point of the air in a wall cavity, in a fuel tank or in a sealed package depends on the surfaces present and on whether there is a reservoir of liquid water. And the calculation assumes the air is at a uniform temperature — a cold surface inside warm air has its own thin layer of cooled air next to it, and it is that layer, not the room, that condenses.

What is the dew point formula?

The dew point is the condensation temperature, and it is found in two steps. The saturation vapour pressure at the air temperature comes from the Magnus formula, e_s = 611.2 · exp(17.62·t / (243.12 + t)) with t in °C and the result in pascals. The actual vapour pressure is that value scaled by the relative humidity, e = RH/100 · e_s. The dew point is then the temperature at which e would be saturating, which inverts the first expression into t_d = 243.12 · ln(e/611.2) / (17.62 − ln(e/611.2)). The page runs all three lines and shows the vapour pressure so you can see the middle step rather than take it on trust.

How do I use this dew point calculator?

Enter the air temperature — in °C or °F, whichever your thermometer reads — and the relative humidity as a percentage between 0 and 100. You get the dew point as the main result in °C, the same figure in °F underneath, and the vapour pressure in hectopascals. There is nothing to leave blank and nothing to solve for: both inputs are known quantities and the dew point is the answer. If your humidity reading comes as a wet-bulb and dry-bulb pair from a sling psychrometer, you will need to convert that pair to a relative humidity first — this page takes the percentage, not the two thermometers.

What does the dew point tell me that relative humidity does not?

It tells you how much water is actually in the air, independently of how warm the air is. Relative humidity is a ratio, so it changes when the temperature changes even though no water has been added or removed: the same air can be at 70% in the morning and 40% in the afternoon purely because it warmed up. The dew point does not move that way, so it is the number to compare between two rooms or two days. In practice the dew point also answers the question people are really asking — will this surface get wet — because the surface sweats when it is colder than the dew point, and the relative humidity alone cannot tell you that without also knowing the surface temperature.

What is a comfortable dew point?

Below about 13 °C feels dry, 13 to 16 °C is comfortable, 16 to 20 °C feels a little humid, and above 20 °C it starts to feel sticky. In Fahrenheit those lines sit around 55, 60 and 68 °F, which is why weather reports in the United States quote dew points rather than humidity in summer. Above roughly 24 °C (75 °F) the air is oppressive and outdoor exertion starts to carry real risk, which is the region the heat index page picks up: a dew point that high on a hot day is what makes the heat index climb so far above the air temperature. Building services uses a related line — chilled water and duct surfaces are kept above the design dew point so that they do not drip.

Why does the page give a dew point below 0 °C rather than a frost point?

Because it uses the saturation curve over liquid water all the way down, and that is the convention weather reports and building services follow. Saturation over ice is a slightly different curve — water molecules leave an ice surface less readily than a supercooled water surface — so for the same amount of vapour the frost point sits a little above the dew point, by about 1 °C at −10 °C. Both are legitimate; they answer different questions, and the page tells you which one it gave. If you need the frost point, swap the constants: use 22.46 in place of 17.62 and 272.62 in place of 243.12 in the same expression, which is valid from about −65 °C up to just below freezing.

Why is my measured dew point different from this?

Usually because the humidity input is the weak link rather than the arithmetic. Cheap hygrometers are commonly quoted at ±3% to ±5% relative humidity, and because the saturation curve is steep, a few percent at 20 °C moves the dew point by about a degree. A reading taken next to a radiator, a kettle or a person's breath is measuring that source rather than the room, and a sensor that has been in a damp bathroom will read high until it dries out. The Magnus formula itself is good to about ±0.6% over the range this page accepts, so the formula is rarely the explanation. If two instruments disagree, compare their dew points rather than their relative humidities — the dew point is the shared quantity, and a disagreement there is a real one.

• Dew point and relative humidity — the NPL's explanation of the two quantities, the Magnus formula in the form used here, and the accuracy of the water-surface and ice-surface constant pairs — National Physical Laboratory
• College Physics 2e, §13.6 — Humidity, Evaporation and Boiling: saturation vapour pressure, relative humidity and dew point, with the phase diagram behind them — OpenStax
• Dew point — the quantity itself, the relationship to relative humidity and to the frost point, and the meteorological thresholds quoted on this page — Wikipedia