Measuring humidity is as messy as how it can make you feel

Conference planning update.

The theme, and possible name, is going to be "Data, Behind the Scenes". The name needs a bit of work. I'll announce a call for speakers soon, likely next week, with more concrete details. Stay tuned.

I'm tentatively planning for Wednesday, September 24, 2025 as being the date for a nerdy online data conference. If anyone has reasons why that is somehow horrible (collides w/ major competing event, holiday, or something else) let me know.

Right now there's a massive heat dome event over a large part of the US. As someone who really really dislikes heat, the event has clobbered itself into every waking moment while I work in front of a fan in an 80F room hoping the humidity doesn't drench me in sweat.

So guess, what, I couldn't help but wonder how is it that we measure humidity. More specifically, if you've ever tried to buy a humidity sensor for random use around the home, they're surprisingly terrible. Cheap ones tend to say they are +/- 5% accurate, and more expensive versions maybe +/- 2 or 3%. There have been huge advances in a lot of parts of metrology in the past few decades, to the point where it's cost effective for nerds like myself to just randomly buy accurate tools as a silly hobby. Notably, thermometers can be accurate to fractions of a degree, while RH measurements depend at least partially on temperature. So I was curious as to what made the measurement of relative humidity so difficult.

To cut to the chase, the primary reason that relative humidity is hard to measure is because the concept itself is complex. Relative humidity measures how much water is in the air, compared against the maximum amount of water that can be held by that air at that temperature. In fact, the "air" part is somewhat superfluous because it has more to do with the equilibrium point of the vapor pressure of water in a space at a given temperature than the other molecules in the air. All that layman's talk about how RH measures how much water air "can hold" is apparently just widespread misconception...

As a physics/chemistry class reminder, vapor pressure is roughly the tendency for a liquid or solid to evaporate, and is related to temperature in a non-linear (mostly logarithmic) fashion. Liquid water boils when its heated to the temperature where its vapor pressure equals the atmospheric pressure, but otherwise it will slowly evaporate at a speed related to the temperature dictated by vapor pressure.

Either way, measuring relative humidity can be quite difficult because it depends on two values – air temperature and the actual water content of the air. Air is a pretty poor conductor of heat (relative to something like liquid water), and it tends to move and have variations in it that can mess with precise measurements. Similarly, figuring out how much water is in a given volume of air is tricky to do on site.

The hair hygrometer

One of the older methods for measuring the humidity used human or animal hairs pulling on a lever. Since the length of the hair would change slightly due to humidity changes, the device could provide a sense of what the humidity was like. The scales of course were pretty approximate, and obviously there was variation between different hairs. It's thankfully rare to see these anymore outside of decorative uses.

A related design uses a metal strip attached to a salt-impregnated paper strip, something like a bi-metallic strip used for temperature measurements. It operates on a similar principle where the paper strip absorbs water and bends the metal strip.

The psychrometer

Moving to more accurate means, psychrometers are devices that uses two calibrated thermometers and use the method of a wet thermometer bulb and dry thermometer bulb to come to a relative humidity value. The principal behind how they work is the fact that evaporation causes cooling (via the water taking up the heat of vaporization) and the evaporation rate (and thus cooling effect) is related to the amount of water vapor that is already in the air. More humid air makes evaporation more difficult, so the wet bulb's temperature will be closer to the dry bulb's temperature all other things being equal.

Many of us learned about psychrometers in science class through the sling psychrometer that involved putting a wet and dry bulb into a device that is then manually swung through the air quickly to create the readings. The temperature readings from the thermometers were then referenced against a chart that would give the relative humidity value for the measurements. More advanced psychrometers use fans and other aids to try to control the inevitable variation in measurements that would result from a handheld device.

I always wondered how the table that converted the readings into relative humidity were determined and apparently they were developed by Willis Carrier, inventor of modern air conditioning and the air conditioner manufacturing company bearing is name, literally wrote the book on psychrometry in "Rational Psychrometric Formulae" that spelled out detailed experimental measurements of the relationship between temperature, relative humidity, dew points, and other details. The "Carrier equation" also lets you calculate the relative humidity base on the two bulb temperatures. I didn't know that the tables were essentially idealized curve fits to a series of experiments.

When I started writing this dive into humidity, I was wondering if I was going to go into the formulae and details of psychrometrics... but apparently the reference book for the topic, "Understanding Psychrometrics" by Gatley is almost 380 pages long and way over my head. So we are all spared the gory details.

Chilled mirror hygrometers

Apparently one of the more accurate systems is a mirror chilled using electronic feedback mechanisms to very precisely the temperature where condensation happens. The now-foggy mirror would mess with a light sensor system of some kind, allowing users to precisely measure the dew point, which can be translated to the relative humidity value. From the dew point value, there are complex formulae, like the Magnus formula or refinements thereof, that can be applied that convert the dew point into a relative humidity measurement for a given ambient temperature.

This all assumes that the mirror stays clean and everything is well maintained.

What's cool is that apparently, chilled mirror hygrometers can be inserted into industrial process spaces (so long as the ambient temperature is below ~95C) to take direct measurements of the dew point.

Capacitive hygrometers

If you're looking for cheaper hygrometers, these are the most common. There can still be high quality, expensive ones that can trace calibrations back to NIST standards, or you can buy microchip versions for a few dollars. They're what's found in the cheapest humidity detectors you can buy off the internet.

These devices take a dielectric material that is hygroscopic (tends to absorb water out of the air) and sandwiches it between two conductive plates, making a capacitor. As water gets pulled into the dielectric material, the capacitance changes, which can be measured. A mapping function is then used to convert those capacitance readings into a RH value.

Since the hygrometer is measuring a secondary value (capacitance) instead a direct value that's convertible to RH like with a chilled mirror via formulae, there's always a chance that the sensor stops behaving according to how the conversion function expects the device to work. This is especially true since the material within the chips can get dirty, contaminated, or otherwise just break down. The cheap ones obviously are not individually calibrated, so there's not even a guarantee that any particular chip actually behaves like what the conversion function was designed for. That said they're often still specified to be within +/- 2% or 3% of RH value.

For many things, like keeping a display of your rare Magic card collection in good condition, they'll probably get you close enough to where we need to be.

Standing offer: If you created something and would like me to review or share it w/ the data community — just email me by replying to the newsletter emails.

Guest posts: If you’re interested in writing something, a data-related post to either show off work, share an experience, or want help coming up with a topic, please contact me. You don’t need any special credentials or credibility to do so.

"Data People Writing Stuff" webring: Welcomes anyone with a personal site/blog/newsletter/book/etc that is relevant to the data community.

About this newsletter

I’m Randy Au, Quantitative UX researcher, former data analyst, and general-purpose data and tech nerd. Counting Stuff is a weekly newsletter about the less-than-sexy aspects of data science, UX research and tech. With some excursions into other fun topics.

All photos/drawings used are taken/created by Randy unless otherwise credited.

• randyau.com — homepage, contact info, etc.

Supporting the newsletter

All Tuesday posts to Counting Stuff are always free. The newsletter is self hosted. Support from subscribers is what makes everything possible. If you love the content, consider doing any of the following ways to support the newsletter:

• Consider a paid subscription – the self-hosted server/email infra is 100% funded via subscriptions, get access to the subscriber's area in the top nav of the site too
• Send a one time tip (feel free to change the amount)
• Share posts you like with other people!
• Join the Approaching Significance Discord — where data folk hang out and can talk a bit about data, and a bit about everything else. Randy moderates the discord. We keep a chill vibe.
• Get merch! If shirts and stickers are more your style — There’s a survivorship bias shirt!