How Can People Be So Stupid? Rotating Header Image

If You Think It is Simple To Measure Outside Temperature With A Thermometer, Try This Experiment

What is the temperature outside? It certainly sounds like a simple enough question. But, have you ever actually tried to measure outside temperature? What a silly question. Of course you have. Virtually every home has an outside thermometer. Often, it is attached to the outside of the kitchen window, so you can easily see whether you need to bundle up, or put on shorts. And, that is fine. But, how many times have you looked at that thermometer, when there is snow on the ground and the local weather station is stating that it is 25 degrees Fahrenheit, and seen a reading of 35 degrees? Or, maybe even higher. Is your thermometer that bad? No. It is probably just “in the sun.” Very few people realize just how much energy the sun places on any given portion of the earth. On average, the sun deposits about 1000 watts per square meter at the surface of the earth on a clear day. That is a lot of power! (And, that is why solar panels offer so much hope for a sustainable energy future if their cost can be reduced, and/or their efficiency increased.) But, if that power is applied to a thermometer that is exposed to the sun, it will raise the temperature of the thermometer, and thus the temperature that it indicates, above the actual temperature of the air. Plus, the fact that the thermometer is mounted to your house will affect it’s reading (up or down, depending on the temperature of your house as compared to the ambient air.) In addition, since it is mounted on a window, you will have some air leaking out of the house, and that air will change the thermometer’s reading. Again, if it is winter, and the inside of your house is warmer than the outside air (hopefully your case), then that leakage air will tend to increase the temperature indicated by the thermometer.

So, we see that a thermometer mounted on your kitchen window is not a good way to measure outdoor air temperature, especially if you intend to measure temperature to an accuracy of less than 0.3 degrees F, or so. (Remember, in all of this global warming brouhaha, we are talking temperature changes of much less than 1 degree Celsius, which is about 1.6 degrees Fahrenheit. So, if we are going to have data that is meaningful in any way, we must be able to measure temperature to an accuracy of at least 0.2 degrees Celsius, or about 0.3 degrees Fahrenheit.)

So, the thermometer attached to the kitchen window, in the sun, does not work. OK, lets place the thermometer in your patio, under a roof, so that it is not in the sun. That should be better. But, is your patio on a concrete slab? Did you mount the thermometer on a wooden post that supports the roof that is exposed to the sun? Is there sufficient air circulation under your patio roof to ensure that the heat absorbed by the roof is not raising the temperature under the roof, even though the area is not enclosed, slightly? Truly, there are a lot of variables.

So, what I would suggest is a simple experiment. Go out and buy 6 to 10 identical thermometers. Then, put them in a common location in your house where they will not be exposed to sunlight. Then, after they have had a few hours to settle, record the temperature of each. If your room is at 70 degrees F, you might get readings like 70.2, 68, 69, 71, 70.8, 69.2. Thus, they are all reading about 70 degrees, but there are some accuracy problems. (If you are using a simple liquid based thermometer, you may not be able to read them to anything like this degree of accuracy. This is a problem in itself, but you can make the most accurate readings that your eyes can determine.) This data, by itself, is not very meaningful. It only means that compared to thermometer A, thermometer B reads 2.2 degrees lower, etc. You need to record this information. If you really want to be scientific, you must measure your “standards” at a number of temperatures, but this will be difficult unless you want to vary the temperature inside you home by a wide range. (You could open the windows on a winter day, or night, to let the house cool down and make additional “calibration” measurements; just don’t let the pipes burst.)

Anyway, so now you have your thermometers, and you know how each “reads” relative to the other. Now, place them in different places on your property. Put at least one in a ludicrous location, such as one where it will be in full sunlight almost all day. Place the others in locations where you think you will get a “real” surface air temperature. Then, for a period of a few weeks, record the readings of your thermometers as accurately as you can. And, make sure you make those measurements at the same time for all thermometers. (Yes, it may take a few minutes to record all measurements, but don’t record thermometer A at 10 am and thermometer B at 10:15 am. Hopefully, you can record them all within a 5 or 6 minute period.) Also, make sure you make a few measurements after the sun has gone down.

I have no way to know what your results will be, but I suspect this exercise will give you some appreciation for the difficulty of accurately measuring outside, surface air temperatures. (Officially, outside air temperatures are to be measured at 2 meters height above the ground, but the exact height is not critical, as long as it is at least 1.5 meters above the ground, and less than 3 meters above the ground.) I suspect that you will find that measurements are much closer after the sun has set. This is important, since it demonstrates the inaccuracies that can be injected due to solar heating. Also, I suspect that you will find that one location seems to be the coolest during the day. This is probably the most accurate location for taking outside surface temperature data. You will probably notice that it is located in a position that is far from any possible heat sources, such as asphault or concrete or buildings or air conditioners. It is also in a shaded location that has good air flow.

While your results may vary, it is important to recognize that almost all of your measurement errors are going to be on the high side. That is, it is virtually impossible to get an outside air temperature reading that is consistently too low, but very easy to get one that is too high. This is one reason to question the data from the GISS and the Hadley center, since they are based on very error prone ground based thermometer measurements. In addition, as has been well documented, a huge percentage of the ground based temperature data is based on sites that are improperly located in relation to artificial heating, such as asphalt parking lots, roof tops, and air conditioner outlets.


  1. energy meter says:

    I am trying to learn about measuring accuracy of energy meter overtime.
    Which brand of energy meter is best for engineering student experiment?

  2. jjMichael says:

    No. But now i will. Thanks for that.

  3. John says:

    Don’t put your thermometer against a tree in the shade – the coolness of the tree trunk will lower the temperature ~3-5 degrees below the actual air temperature on a warm day.

  4. Randy Yetto says:

    Why does the temperature drop so fast just before the sun comes up?

  5. Jeff says:

    I completely agree. I bought a simple indoor/outdoor digital thermometer, the kind with a remote outdoor radio-transmitting unit. I had to locate it well away from my house and away from any structure to get what I thought was a reasonably correct reading. Even then, it was always a bit higher than the local TV station reported… I can’t begin to consider GISS data accurate without seeing their corrective factors and how they are derived. I suspect the anthropomorphic global warming concept is based on rather flimsy data. It may be true to an extent. I suspect it is a factor, but overall, the global warming bandwagon is politically polluted. Scientifically derived truth is severely clouded by political bullshit. Therefore, I have to reserve my judgement until clearer data can be found.

Leave a Reply

Maximum 2 links per comment. Do not use BBCode.