Aaah. To live in the days when temperature was still simple. I.e. before numbers were invented.

“It’s twice as hot as yesterday!” Or is it…

Language, and the way we think about temperature, is a lot of fun. For example: If it goes from 10°C to 20°C, it would be reasonable to say that it’s twice as hot… Makes sense, right? Well, that depends on how you look at it.

From 10°C to 20°C? Looking at the graph… Yep, twice as hot…

Let’s confuse things with Fahrenheit…

But… Not the whole world operates on Celsius. If you’re one of those Fahrenheit-loving types, the same temperature change would take you from 50°F to 68°F… Which, if you’re looking at the numbers, is just a 36% increase. Argh!

If you measure it, this temperature change is exactly the same — one in Celsius, and one in Fahrenheit. The Celsius scale is showing a 100% increase, but the Fahrenheit scale tips in at just a 36% increase. What the what?!

Something about atoms moving

Perhaps the only fair scale to use is to look at what heat actually is — a measurement of thermal energy, or atoms moving. The only time atoms stand still is at absolute zero, which helpfully is where the Kelvin scale starts.

But… Adding Kelvin into the mix is even more complicated:

Oh christ, here we go. Celcius: 100% increase. Fahrenheit: 36% increase. Kelvin: 20% increase

So, it should follow that the only time the temperature actually doubles, is in Kelvin, the scale that actually starts at zero…

So, if we are having a cold day of day of 10°C / 50°F / 283°K, and we were to double it in terms of Kelvin, we end up on 566°K… Mathematically correct, but that puts us at well beyond the boiling point of water (145°C / 293°F), and well above human comfort. Hmm…

Okay, but now we get it right? Not quite…

Yes, my dear pedants, I can hear you sharpen your Internet Commentary pencils already… If I’m going to sit here and argue that Kelvin is a sensible scale, then what about Rankine?

Never heard of Rankine? The Kelvin scale is basically the same as the Celsius scale, except re-adjusted so zero is absolute zero. So, of course, ten years after some smart-arse Scotsman came up with the Kelvin scale, another Scot decided to make up another scale, that did basically the same thing: Fahrenheit, except zeroed at absolute zero.

Le’Sigh… More graphs:

For fuck’s sake… Not again. Celcius: 100% increase. Fahrenheit: 36% increase. Kelvin: 20% increase, Rankine: 3.5% increase

Yep, that’s right, ladies and gentlemen… when it looked like the temperature was doubling in Celcius, the very same temperature difference only caused a 3.5% increase on the Rankine scale.

And, of course, for the sake of unwavering dedication to completeness and pedantry, I’ll do the same exercise as above… If we are having a cold day of day of 10°C / 50°F / 283°K / 510°R, and we were to double it in terms of the Rankine scale this time, we end up on 1021°R… Which is hot enough to bake pizzas: 294°C / 561°F / 834°K

In conclusion: It’s never twice as hot.

Meters, feet, ounces… Most of the scales have the common decency to at least be linear: If it doubles in one set of numbers (say, 5 meters to 10 meters), it doubles. In all sorts of measuring units:

Sniff. Can we all switch to SI units now, please?

Temperatures don’t really have that luxury, which means that it never really makes sense to compare changes as a percentage or a doubling.

In other words: Dear weatherman / random person in a bar, if you wish to avoid the ire of pedants, stay away from assertations about ‘twice as hot’ or — even worse — ‘twice as cold’ (WHAT DOES THAT EVEN MEAN?).


This has been a public service announcement from the desk of SI-standards powered pedants anonymous

Love this sort of drivel? Follow me on Twitter, there’s more where this came from.

The title image at the top of this post is Winter II (cc) Anssi Koskinen