Your pet peeve, technical or otherwise.

[vk6zgo]

That's just a round-number fetish.

Nowt wrong with that. Easy to understand and use - what's not to like?

Except that there's really nothing needing to be "understood" here. We're taught from a very early age that 32°=freezing, 212°= boiling and 68°=room temperature, and somehow we manage to retain that and don't have to refer to any cue cards to figure out whether it's hot, cold or balmy. (Same goes, of course, to those who grew up using °C.)

Like knowing that 4 quarts=1 gallon, 12"=1', 1 mile=5,280', etc., etc. Somehow we manage to muddle along here ...

Just to confuse the issue, "Imperial" gallons are larger than US gallons, so quarts & pints are proportionally larger.
an Imperial "44 gallon drum" is a US "55 gallon drum", or a"200l drum".

The immediate advantage of the Metric system is obvious in such a case, as only one conversion is needed at each end of a transaction between countries using US, Imperial  & Metric (obviously a different conversion factor in each case).

Back in the day, the difference led to US people reading mpg figures from a test drive in an Imperial country being disappointed by the result in their country, & those in an "Imperial" country being agreeably surprised that the "gas guzzling Yank tank" gave better mpg than they had been led to believe.

[themadhippy]

Back in the day, the difference led to US people reading mpg figures from a test drive in an Imperial country being disappointed by the result in their country, & those in an "Imperial" country being agreeably surprised that the "gas guzzling Yank tank" gave better mpg than they had been led to believe.

and it takes us brits longer to get drunk when supping yanky pints

[CatalinaWOW]

I certainly agree that having one worldwide standard is beneficial.  But arguing that C is better than F is just silly.  Why you would ever need to use the fraction of the difference between freezing and boiling of one substance escapes me.  In either system folks know when a temperature is miserable, and I don't think anyone thinks of it as a bit less than half way to boiling.  Both systems are arbitrary.  But for any real calculation you would use temperature difference and heat capacity in either system.  Yes the units are closer to round numbers for pure water in C, but that only covers a small fraction of real world problems.  Just add some salt or ethylene glycol and you have to go to Google either way.

[vk6zgo]

Back in the day, the difference led to US people reading mpg figures from a test drive in an Imperial country being disappointed by the result in their country, & those in an "Imperial" country being agreeably surprised that the "gas guzzling Yank tank" gave better mpg than they had been led to believe.

and it takes us brits longer to get drunk when supping yanky pints

There has been something of a "re-Imperialisation by stealth" with beer sizes in Oz, ever since the little "Trendoids" decided that they like to pretend to be in London.

To this end, the traditional Australian measures of a "glass", "middy", "pot" (& sometimes "schooner") which were painstakingly Metricated when that system came in, were replaced with the bland "Pint".

"Market forces" have dictated this be so for some reason, but what adds insult to injury is that you aren't sure if you are getting a full-blooded Pommy Pint, or an anaemic US "Pintette" for your money.

[RJSV]

   The tie-in of linking 'molecular motion' with temperature is IMO,  only a loose one,  not some literal and total LINEAR one.
You raise the temperature,  and that's an indicator tightly linked to average molecular motion,  to be sure.  But not dead-on tracking.

   If you consider the motion difference with water,  one sample at 28 ° F,  and another at
35 ° F,  there's NO WAY that's going to track as a perfect linear proportional change.
I mean,  you've got frozen molecules,  that become free.  I'd have to pound on the meaning of 'frozen' in relation to 'motion' to argue that.

   Same concept up at 212°F,  where in a gas state I would think the degrees of freedom,  to have velocity increase,  are a big, step-like change,  in average speeds,  as those newly free water vapor molecules can absorb the heat present around them.

   Perhaps that dependence,  between temperature and (water) molecular speed is more of a gross quality,  rather than being EXACT linear everywhere...?

[TimFox]

   The tie-in of linking 'molecular motion' with temperature is IMO,  only a loose one,  not some literal and total LINEAR one.
You raise the temperature,  and that's an indicator tightly linked to average molecular motion,  to be sure.  But not dead-on tracking.

   If you consider the motion difference with water,  one sample at 28 ° F,  and another at
35 ° F,  there's NO WAY that's going to track as a perfect linear proportional change.
I mean,  you've got frozen molecules,  that become free.  I'd have to pound on the meaning of 'frozen' in relation to 'motion' to argue that.

   Same concept up at 212°F,  where in a gas state I would think the degrees of freedom,  to have velocity increase,  are a big, step-like change,  in average speeds,  as those newly free water vapor molecules can absorb the heat present around them.

   Perhaps that dependence,  between temperature and (water) molecular speed is more of a gross quality,  rather than being EXACT linear everywhere...?

Here is a thorough explanation of temperature:  https://ifisc.uib-csic.es/raul/CURSOS/TERMO/Thermodynamic%20temperature.pdf
Here is a simplified version of that explanation:  https://www.nist.gov/si-redefinition/kelvin-thermodynamic-temperature
The dependence is on total energy, in a system with multiple degreees of freedom, not a gross relation with water molecular speed.
Thermodynamics is something that happened to other people.

[Analog Kid]

Makes sense. One question, though, @TimFox: is the relationship between temperature and energy in a substance independent of phase changes? Doesn't something non-linear occur when, say, water boils and changes from liquid to gas?

Hope this doesn't overcomplicate this discussion.

[TimFox]

If you are heating the sample and it goes through a phase transition, there will be a discontinuity in the temperature rise, but temperature is still defined by the total energy, as discussed in the first reference I cited.
The discussion of “degrees of freedom” is very important in statistical mechanics.

[CatalinaWOW]

Lots of bookkeeping in keeping all of this straight.  Many real life systems aren't "closed" in the sense that where all the energy is going isn't closely tracked.  A pan of water on a stove is a common example.  The temperature rises with a good approximation to linearity with energy input until boiling starts.  Then it effectively stops for the water in the pan.  The energy is going into vapor and steam which isn't readily or commonly tracked.  Not disagreeing with TimFox who is absolutely correct, just pointing out how this can get confusing.

[paulca]

Maybe I'm missing something here, but I can't understand why one can't just treat a temperature, no matter what the scale, as a percentage of some other temp. Nothing special about temp as a quantity that I can see: "X is Y% of Z" seems perfectly valid to me.

We can do it with voltage, current, frequency, etc., etc.; why not temperature?

Because 0V = 0V.  0A=0A, 0Hz = 0Hz

0*C does not equal "no heat".

What is 20% hotter than 0*C?

[paulca]

Makes sense. One question, though, @TimFox: is the relationship between temperature and energy in a substance independent of phase changes? Doesn't something non-linear occur when, say, water boils and changes from liquid to gas?

Hope this doesn't overcomplicate this discussion.

Latent heat.  "Latent" meaning "hidden".  Simply because a direct energy comparison based on temperature will not "see" this energy unless it considers the state change and measures it also.

100g of water at 100*C@STP  has a lot less energy than 100g of steam at 100*C@STP(*)

* the energy will reveal itself if you attempt the above in a lab.  You won't turn 100g of water into 100g of steam and retain standard pressure, you will either need a very large vessel or higher pressure.

[Siwastaja]

Maybe I'm missing something here, but I can't understand why one can't just treat a temperature, no matter what the scale, as a percentage of some other temp. Nothing special about temp as a quantity that I can see: "X is Y% of Z" seems perfectly valid to me.

You are missing the fact that you should not concentrate on what someone "allows" you to do; you should concentrate on asking the question: "does it make sense to do this for purpose X?" "is this useful?" "does this work?" "is this true?" Facts instead of rules.

My question to you is, in what way you find the information that +2degC is 100% hotter than +1degC, useful for any practical purpose? The only use I can imagine is trolling people with a "funny" mis-calculation. But of course the calculation and claim itself is correct and definitely not forbidden. Just useless.

[PlainName]

Because 0V = 0V.  0A=0A, 0Hz = 0Hz

0*C does not equal "no heat".

What is 20% hotter than 0*C?

What is 20% more current than 0A?

[Siwastaja]

What is 20% hotter than 0*C?

What is 20% more current than 0A?

Paulca asked wrong question. The right question is, what is 20% hotter than , not 0, but 0.1degC. Then your counterargument what is 20% more current than 0.1A makes the whole point obvious.

20% more current than 0.1A is 0.12A. It moves 20% more charge. It heats up the conductors roughly 20% more (RI^2 loss is quadratic, but heat that escapes is roughly linear, ignoring increase of convection). It charges the battery, or a capacitor from empty to full in 20% less time. So you see it's a useful number. These same laws of heating or battery charging speeds repeat, the difference is always 20% regardless if the current was originally 0.1A, 1000A or 0.00001A.

But if you increase the temperature of something (say, outdoor air; heatsink surface, semiconductor die) from 0.1degC to 0.12degC, or, by 20%, what is the practical difference? Do you put 20% more clothes on, even roughly? Do you need to use a 20% larger heatsink?

And then again, if you increase the semiconductor die temperature from 100degC to 120degC, the same 20%, then the consequence is very real.

So was the percentage of absolute Celsius temperature useful at all, for any purpose?

But don't get me wrong, there are cases where taking a percentage out of temperatures do make sense. For example, temperature differences. If, using heatsink A, Tj=120degC at Ta=20 degC (i.e., dT=100degC), and using heatsink B, Tj=110degC at Ta=30degC (i.e., dT=80degC), then you can say that heatsink B does (100-80)/80 = 25% better job than heatsink A; and it does not matter at all which unit the temperatures originally are in.

[PlainName]

Sure, I'm just being picky

Nevertheless, I think we need to keep in mind the context of things. For instance, words have specific meanings in law that the general public don't appreciate and thus use them differently. If we were lawyers we might well be whinging about that, just as we complain, as technical people, about this temperature thing. But in the context of weather forecasts for proles, how do you get across that today is going to be somewhat warmer, but not too warm, than yesterday? I think saying it's 25% warmer gets across the meaning even if pedantically it is pure cobblers.

[TimFox]

Sure, I'm just being picky

Nevertheless, I think we need to keep in mind the context of things. For instance, words have specific meanings in law that the general public don't appreciate and thus use them differently. If we were lawyers we might well be whinging about that, just as we complain, as technical people, about this temperature thing. But in the context of weather forecasts for proles, how do you get across that today is going to be somewhat warmer, but not too warm, than yesterday? I think saying it's 25% warmer gets across the meaning even if pedantically it is pure cobblers.

The television and radio weather reports here will say that tomorrow is 10 deg warmer than today (Fahrenheit), and the populace understands what to wear tomorrow.

[paulca]

Sure, I'm just being picky

Nevertheless, I think we need to keep in mind the context of things. For instance, words have specific meanings in law that the general public don't appreciate and thus use them differently. If we were lawyers we might well be whinging about that, just as we complain, as technical people, about this temperature thing. But in the context of weather forecasts for proles, how do you get across that today is going to be somewhat warmer, but not too warm, than yesterday? I think saying it's 25% warmer gets across the meaning even if pedantically it is pure cobblers.

It's, em, kinda in the name.  We use "degrees".  It will be 2 degrees warmer tomorrow.  That is where it comes from, the need for it and why we use them today.

Nobody wants to work in blackbody energy calculations and QED to work out if they are going to get burnt or pleasantly warmed.

"Degrees" and quanity-less, unit-less, arbitary scales broken down into arbitary discrete portions based on what someone made up basically.  Mr F and Mr C did it different to Mr. K who at least tried.

[CatalinaWOW]

Sure, I'm just being picky

Nevertheless, I think we need to keep in mind the context of things. For instance, words have specific meanings in law that the general public don't appreciate and thus use them differently. If we were lawyers we might well be whinging about that, just as we complain, as technical people, about this temperature thing. But in the context of weather forecasts for proles, how do you get across that today is going to be somewhat warmer, but not too warm, than yesterday? I think saying it's 25% warmer gets across the meaning even if pedantically it is pure cobblers.

It only sort of works.  It comes back to pedantic stuff, but given that human comfort is a relatively narrow band well separated from the zero point of either common temperature range percentage of temp just doesn't give very useful information.  Twenty percent warmer could be more or less comfortable.  Or about the same.   

Weather forecasters in my area deal with this by suggesting appropriate clothing (jacket, shorts, parks etc.). That allows taking, wind, humidity, cloud cover and precipitation into account, but still doesn't take individual responses into it.  I've been comfortable in sandals and shorts where other people were wearing parkas.

[Kim Christensen]

I think saying it's 25% warmer gets across the meaning even if pedantically it is pure cobblers.

Lets say it's 20c outside today and the weatherman says, "It'll be 25% warmer tomorrow" which means it'll be 25c
Then the next day, when it's 25c outside and the weatherman says, "It'll be 25% cooler tomorrow" which means it'll be 18.75c...   

[KE5FX]

You can feel free to use percentages with temperature, but you need to use the absolute Kelvin scale for that, not degrees C or F.

Of course, if the temperature goes up by 25% from one day to the next, you've got bigger problems than the outcome of Internet arguments.

[TimFox]

I think saying it's 25% warmer gets across the meaning even if pedantically it is pure cobblers.

Lets say it's 20c outside today and the weatherman says, "It'll be 25% warmer tomorrow" which means it'll be 25c
Then the next day, when it's 25c outside and the weatherman says, "It'll be 25% cooler tomorrow" which means it'll be 18.75c...   

Consider three temperatures, all of which are typical weather values:
A = 10^o C = 50^o F
B = 20^o C = 68^o F  and
C = 30^o C = 86^o F

Obviously,  (C - A) = 2 x (B - A) in both scales.
However 30/10 = 300% (in Celsius), but 86/50 = 172%  (in Fahrenheit)
another example of how not to use temperature values.

Exercise for the reader:  repeat this in absolute temperatures (Kelvin and Rankine), say, for 100, 200, and 300 K.

[TimFox]

The question above about temperature in situations involving phase change (or change of state) prompted me to think about thermometers.
Now, a thermometer measures its own temperature (excluding "thermal radiation thermometers" such as IR guns or cameras).
In classical thermodynamics (before statistical mechanics), an ungainly-named "Zeroth Law" says that if system A is in thermal equilibrium with system B, and system B is in equilibrium with system C, then system A is in equilibrium with system C.  This means we can assign "temperature" to such systems.  Before statistical mechanics, investigators defined reference values (such as 0 and 100 or 32 and 212) to physical situations (freezing and boiling) to assign numerical values on their thermometers.
To use a thermometer, one puts it in thermal equilibrium with the system being measured, and then reads it out according to its calibration.
This is obvious with a kitchen thermometer, where it starts out at room temperature in the drawer, then rises after insertion into the meat until it reaches equilibrium.
A thermometer calibrated in this manner is useful over a range with no phase transitions in the temperature sensor:  a mercury thermometer is used above -38^o C and below +350^o C to avoid freezing and boiling, respectively.
Other thermometers (such as RTDs, diodes, thermistors, and thermocouples) do not use liquids and avoid phase transitions at low temperatures.
I wondered about extremely low temperatures, below 4 K (helium boiling point).
The concept of absolute zero arose when considering the behavior of "ideal gases" (far above condensation) in Charles' Law, which became part of the Ideal Gas Law 
  pV = nkNT
where p is gas pressure, V is the volume, n is the quantity of gas in moles, k is Boltzman's constant, N is Avogadros' number, and T is absolute temperature.
Extrapolation from "normal" temperatures allowed the estimation of absolute zero.
Modern cryogenics can achieve extremely low temperatures.
Apparently, the best thermometer for extreme cold is a negative-tempco RTD (where the resistance rises at low temperatures).
This unit  https://www.lakeshore.com/products/categories/specification/temperature-products/cryogenic-temperature-sensors/ultra-low-temperature-rox  from Lakeshore Cryotronics is calibrated down to 10 mK = 0.01 K, and useful to 5 mK.  I assume the lower limit is due to self-heating:  not much is required to screw up a 5 mK reading.
The typical resistance curve is shown in  https://www.lakeshore.com/products/categories/overview/temperature-products/cryogenic-temperature-sensors/ultra-low-temperature-rox  and goes from about 2000 at 1 K up to 10,000 at 10 mK.

[PlainName]

I think saying it's 25% warmer gets across the meaning even if pedantically it is pure cobblers.

Lets say it's 20c outside today and the weatherman says, "It'll be 25% warmer tomorrow" which means it'll be 25c
Then the next day, when it's 25c outside and the weatherman says, "It'll be 25% cooler tomorrow" which means it'll be 18.75c...   

Yes. And...?

Did you read that sentance before you quoted it or was it just a handy one to paste in?

[PlainName]

Sure, I'm just being picky

Nevertheless, I think we need to keep in mind the context of things. For instance, words have specific meanings in law that the general public don't appreciate and thus use them differently. If we were lawyers we might well be whinging about that, just as we complain, as technical people, about this temperature thing. But in the context of weather forecasts for proles, how do you get across that today is going to be somewhat warmer, but not too warm, than yesterday? I think saying it's 25% warmer gets across the meaning even if pedantically it is pure cobblers.

The television and radio weather reports here will say that tomorrow is 10 deg warmer than today (Fahrenheit), and the populace understands what to wear tomorrow.

Maybe. If it's -55C out and it's going to be 10 degrees warmer, that just means your todger will last an extra second before falling off through frostbite. OTOH, it it's 43C then +10 degrees means you''re probably going to die. Really, we just want to know if it's going to be a bit warmer, lots warmer, the same, colder, etc. But giving it numbers - any ones will do - makes it much more 'official' and specific, as if the computer says.

I am not disagreeing with you on a technical level, but much of the world doesn't understand how to even spell teknikal.

[TimFox]

Sure, I'm just being picky

Nevertheless, I think we need to keep in mind the context of things. For instance, words have specific meanings in law that the general public don't appreciate and thus use them differently. If we were lawyers we might well be whinging about that, just as we complain, as technical people, about this temperature thing. But in the context of weather forecasts for proles, how do you get across that today is going to be somewhat warmer, but not too warm, than yesterday? I think saying it's 25% warmer gets across the meaning even if pedantically it is pure cobblers.

The television and radio weather reports here will say that tomorrow is 10 deg warmer than today (Fahrenheit), and the populace understands what to wear tomorrow.

Maybe. If it's -55C out and it's going to be 10 degrees warmer, that just means your todger will last an extra second before falling off through frostbite. OTOH, it it's 43C then +10 degrees means you''re probably going to die. Really, we just want to know if it's going to be a bit warmer, lots warmer, the same, colder, etc. But giving it numbers - any ones will do - makes it much more 'official' and specific, as if the computer says.

I am not disagreeing with you on a technical level, but much of the world doesn't understand how to even spell teknikal.

I grew up in northern Minnesota.  On cold days, the radio would predict “30 degrees below zero” (Fahrenheit, of course) and we all knew what that meant for parka and hat purposes.  We didn’t need “wind-chill” estimates to know it was cold.
Differences between medium temperatures, e.g. 10 deg between 60 and 70^o F, are comprehensible.
Jack London quoted Alaskan rules of thumb, e.g. “spit freezes before hitting the ground”, to evaluate risks.
I turn 75 next week.  I have never seen anyone refer to % changes in temperature for weather or other purposes using C or F, and rarely for technical temperature in K.