We are in the wild and need to know how cold/hot it is. How can we measure the temperature if we don't have any thermometers? Can we build something similar, etc.?

It doesn't have to be necessarily built from "natural materials". Maybe we've a small camp with common camping equipment (except a thermometer of course).

  • 2
    there is a QI episode that discuss this. Apparently one of the way is the chirping frequency of some critter. (and quite accurate too).
    – njzk2
    Dec 2 '15 at 16:13
  • 2
    @njzk2 it's the cricket's chirp though there are multiple equations to calculate it Dec 2 '15 at 16:20

Assuming that you are not going to use if for checking fever. There are other ways to check fever.

There is no mechanism, or a piece of equipment (other than dedicated thermal transducers) I have heard of that can measure the temperature precisely. If you are just curious about how hot/cold the ambiance is, you can pull out your cellphone and get that data anytime you want.

If you just want to know if it is getting colder with the time, look for a cat, the more it curls, the colder it is getting.

  • With the one exception of recognizing 0 °C ± a few °C by frost/water freezing. Dec 5 '19 at 16:47

Yes, you can build a basic thermometer using just water, or any liquid. You first need to mostly seal it in a container (like a glass bottle) and use a straw or thin clear tube immersed in it to allow the liquid to expand up the straw. Do not let the straw touch the bottom. A 50-50 water & alcohol solution will also do the trick if you happen to have alcohol, but in a survival role alcohol should be kept for injuries, etc.

The trick in order to re-use the thermo-bottle it is to make sure that each day the exact same amount of water/alcohol is there at the start since both water and alcohol slowly do evaporate over time. (Note your regular mercury thermometer is sealed so it never escapes.) Or seal it all when not using it.

The next stage is marking the clear tube with various temperatures. Best case scenario is if you are in winter and can place it in ice cold water. That's your 0°C/32°F mark on the straw, usually the bottom of the straw. Placing the thermo-bottle in boiling water is an upper mark on the straw, just be careful not to burn yourself, or crack the bottle, or even melt the bottle or straw. Placing the thermo-bottle under your jacket for 10 minutes should give you a 37.0°C/98.6°F reading or body temperature (or you can immerse the bottle in your fresh urine which should be about 37°C/98°F).

Once you have these various marks, divide the straw into equal sections to give you an approximation of the temperature between your body temp and zero. It is still an approximation because altitude, humidity, pressure and water density can affect the reading. With alcohol as a liquid, readings below zero will also be possible.

For efficiency, practice using and marking a practice thermo-bottle and straw or tube BEFORE going into the outdoors when you do have a thermometer to help. This will help you get the kinks out when you finally need to construct one.

Other ways to measure temperature without a proper thermometer are:

  • A Galileo thermometer (more complicated tube with water and floating weighted orbs in it)

  • Evaporation rates of a water drop in direct sunshine (done ahead of time and memorized.)

  • Facial sensation (the face is better than hand at discerning temperature)

  • A basic personal chart (ice freezes at 0°C/32°F, my breath is visible at 7°C/45°F, I go to t-shirt mode at 16°C/61°F, I start to sweat when sitting at 28°C/82°F, my metal camping spoon in 43°C/110°F sun for 1 minute scalds me)

Here's a related site about the thermo-bottle (not my site):


Good luck.

PS do not throw boiling water in freezing temperatures, it's not worth the risk.

  • 1
    Many of these ideas also depend on the humidity, the size of the bottle, or even whether the wind is blowing.
    – Chris H
    Jul 23 '17 at 8:14
  • 3
    Disagree with the final point. Throwing boiling water at -40°C is great fun (off a cliff or building). So is blowing bubbles. There's no outdoor life without risk-taking.
    – gerrit
    Dec 3 '19 at 14:05
  • Good suggestion. It works better if the straw is sufficiently narrow to allow the liquid to rise part way up through capillary action. Jun 3 '20 at 17:08

You can build something that tells you when it's colder or warmer than something else, but without a gauge to help you create your graduations, you won't know what the actual temp is. I am not sure this is a necessary tool however. The body tells you by its comfort level whether you need more or less heat, or more or less water when it's hot.

You can also simply take signs from the environment, does your breath mist up, does throwing a cupful of water in the air freeze before it hits the ground (it happens in subzero weather) right around -25°F (-32°C) water thrown into the air will start to freeze before it hits the ground for example.

  • Your "subzero" statement lacks a unit, do you mean Fahrenheit?
    – gerrit
    Jul 23 '17 at 10:49
  • 1
    Also, from experience, throwing cold water down at -40°C does not make it freeze before it hits the ground. Thowing hot water down might, depending on quantity and height. If you want to waste 1 litre of boiling water throwing it down from tree-top height just to tell the difference between -20°C and -40°C, go ahead :)
    – gerrit
    Jul 23 '17 at 10:53
  • 1
    @gerrit The idea that "hot water freezes faster than cold water" and the reverse "cold water boils faster than hot water" is based off of a misguided understanding that hot water cools faster and cold water warms faster. Hot water will cool faster (higher rate of change of temperature), but it will not freeze faster than, or even overtake the temperature of, the cold water. At best, the hot water temp will approach the cold water temp and will asymptotically get closer and closer, never passing it. This is assuming all other factors are equal.
    – Loduwijk
    Sep 26 '17 at 21:50
  • 1
    @Aaron True, but the assumption all other factors are equal does not hold. Try to throw a bucket of hot and cold water off a rooftop at -40°C and observe what happens. My observation is entirely empirical.
    – gerrit
    Sep 26 '17 at 22:24

°C thermometers are calibrated relative to water temperature.
Hold some ice in hand, as soon as it starts melting, the temperature of the liquid water in the palm is just above 0°C
Let it stay in palm for a bit, until there's no temperature difference felt with other fingers; this would be around 37°C

Boil some (rain) water, stop heating when you see some bubbles & vapor. The temperature of the water is just below 100°C (at sea level)

The above facts will provide approximate calibration to your DIY thermometer.

One possible construction of a thermometer would rely on refraction of light in different media:

  1. When light enters a more dense substance (higher refractive index), it 'bends' more towards the normal line.
  2. Hot water is less dense than cold water. Calibrate the thermometer based on the angles of refraction of a straight object with different (known) temperatures of water eg. boiling point, freezing point, human body temperature.

Other constructions would rely on expansion of some fluid:

  1. For closed housings, the expansion would indicate temperature eg. mercury in a glass tube, air in a balloon, juice in a drinking straw, etc.
  2. In case of a gas in an open tube, the escaping gas with higher temperature would deflect an obstruction (eg. a leaf or paper) more.
  3. If the obstruction at the opening is attached to a fixed joint allowing centrifugal movement, then the rate of spin would indicate temperature, like car engines, or paddle wheel boats, speedometers.

Others would rely on density of a fluid eg. the height of some solid object floating in a container of the measuring fluid eg. water

Some objects bend more easily at higher temperatures, and this deflection would provide a measure of the temperature eg. bimetallic thermostat, or braided composites like fiber or resin.

  • The temperature of your hands is actually almost never 37 C, they vary widely based on environmental and circulatory condition, but in a cool environment (13 C/55F) had temp was found to be 17-18 C
    – bob1
    May 25 '20 at 21:09
  • The calibrations methods provide approximate temperatures (at sea level). 37°C is a test to check calibrations between 0-100°C, and not to be relied on as a good indicator to calibrate 37°C, since some people can maintain life at 46°C and some as low as 14°C. The crux of the answer was ways to build your own thermometer without a lab apparatus.
    – Zimba
    May 29 '20 at 13:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.