Well, to a non-American like myself the Fahrenheit scale feels weird and arbitrary (and yes, I understand the "human scale" argument, just don't buy it") while the Celsius scale makes sense.

Or is it just more familiar? We all like what we're accustomed to. Is there some reason Celsius 'makes sense'? Why does it make sense that 22 is comfortable but 40 is miserably hot? It doesn't make any more sense than 72 and 100.

The familiar part has changed for Canadians.

I grew up during the transition period I was too young to know Fahrenheit, I had just entered school when the switch was occurring.

My parents are pretty well used to Celsius now but still may occasionally use Fahrenheit such as some electronic device that allows F or C they choose F.

It's still a but messed up though since Canada borders the US. The old joke being weather reports that show the border with the US say Detroit shown as 90 and Windsor as 32 "See Marge as soon as you cross the border into Canada it goes dow to 32!"

But still you get people of all ages saying things like "300km/h wow that's nearly 200 mph" or "You weigh 70kg what's that like 150 pounds?". It's more due to history and the influence of the US on our culture, I bet more french speaking areas like Quebec, northern NB don't do that but who knows maybe they do.

I was born and raised in Florida. I now live in Sweden. Because of experience, I understand C pretty well between -15 and +10, and between +80 and +100 (Swedish sauna temperatures).

Between +20 and +50 I'm still more comfortable with 70F-120F. I also spent some time in Illinois, where I learned to understand the subzero range from -15F to 0F.

At LOX temperatures of 90K and below, I'm most comfortable in Kelvin. ;)

0°C - water turns into ice 100°C - water boils and turns into steam

-30°C and below - very cold weather +30°C and above - very hot weather

But the freeze/boil numbers are arbitrary as well only being true at sea-level. Where I live in Colorado, we're 1 mile (cough, 1.6 km) above sea level meaning those formulas are meaningless.

They aren't meaningless. Water boils at 95 C at your altitude - effectively the same temperature for all "common" usages. E.g. you'd still better not stick your hand into boiling water.

A <5% error for the zone that >99% of humans inhabit seems pretty practical to me. The freezing point is effectively identical.

iirc, water boils at 95C is around 5000m altitude, no one lives normally there - even La Paz is less than 4000.

Well its rather easy to make your own thermometer using the Celsius scale. Put quicksilver in a small glass tube and put it outside. When water starts to freeze mark the current position with 0. Drop it in boiling water and mark it with 100. Evenly distribute now 98 markings between the two existing ones and voila you have just made your own thermometer :)

Evenly distributing 98 markings is hard. Evenly distributing 64 is easy -- just halve the range six times. That's why the distance from freezing water to body temperature is 64 Fahrenheit degrees.

Except...

Freezing point of water: 32°F Average human body temperature (oral): 98.2°F

So that's 66.2°F, not 64°F.

Those two temperatures were the original reference points for the Fahrenheit scale, and body temperature was defined as 96°.

The current definition is relative to the melting point of ice (32°F), and the boiling point of water at standard atmospheric pressure (212°F, which is exactly 180°F higher than the freezing point). Except now body temperature is 98.2°F instead of 96°.

If you want a thermometer calibrated in Fahrenheit, you have exactly the same problem as you do with Celsius. You just have to put 178 markings between the two reference points, rather than 98.

I don't think its advisable to use my self-made quicksilver thermometer for any kind of body temperature measurement.

Which makes no difference because you're eyeballing it to start with. You can just as easily divide the 0-100 range in half a bunch of times. Your thermometer is not achieving <1 degree accuracy.

Of course if you have a tool to accurately measure halves, then you can also measure any other division you care to use.

All depends what you are used to - I would say 15C or so is "comfortably warm", 22C is definitely "warm" and 40C is utterly miserable with no wind but really nice with a breeze and in the shade....

People generally start sunbathing, wearing T-shirts and flip-flops, when the temperature goes above 10C here...

Could you say a bit more about why you feel it makes sense, perhaps with a chart like the one I wrote up?

I'll bite (and I grew up in one of the colder parts of Canada, so I'd call 0F a cold but not unreasonable winter day).

-30: Avoid going outside, if you do don't leave any exposed skin

-20: A cold winter day: toque, scarf, mittens, winter coat and boots are must have

-10: An average winter day: same load-out as above, but you'll be more comfortable

0: A mild winter day: snow possible, you may leave your winter jacket open, or skip the scarf

10: Early spring or late fall: you might be able to get away with a light jacket, but I'd leave that until 15

15: Long pants and light jacket

20: Leave the jacket home, maybe wear shorts or a t-shirt

25: Comfortable summer weather: shorts and t-shirt

30: Annoyingly hot summer weather, good for swimming

35 or higher: hole up somewhere with AC

LOL. I'm in Western Australia:

<10: Bloody cold, avoid going outside if possible. Rug up if you do.

15: Winter. Probably rain and wind.

20: Cool Autumn / Spring, you might need a jumper outside.

25: Vaguely warmish, t-shirt weather.

30: Getting warm, wear shorts and t-shirt.

35: Beach weather, avoid going outside for too long if you're in office attire and don't want sweat patches.

40+: Hot beach weather, stay out of the sun (or in the water) in the middle of the day.

Interesting...I'm in the UK, for me 24°C with a light breeze makes for a perfect Summer's day: warm enough to walk around without sweating buckets. 30°C is uncomfortably hot. I guess it's all relative to what you're used to (and what you're doing e.g. day at the beach or walking in the city).

Well, people don't for the most part look forward to 40 degree days - this is warmer than body temperature, so you have to engage in some form of cooling. Going to the beach is one form of cooling :) Avoid being outside, drink lots of fluids, AC if you can, anything, really. Construction workers generally stop work around 36, from memory. Conversely, people here will tell stories about the crazy British neighbour who is out doing the gardening in a singlet when it's only 10 degrees.

Here in Melbourne we'll have several 40-degree days each summer, and the only large city with more that I'm aware of is Phoenix. There are heaps of large cities that are warmer in general, but they don't peak quite as high. Record is 46.2, set a few years ago. Not a pleasant day, especially since it was after a few similar days...

Victoria, Australia here. About the same, except we can get three or four of those all in the one day. (The running joke is that if you don’t like the weather in Melbourne, wait.)

I’ve lived in both the low and highlands of Papua New Guinea and they have two types of weather: dry and wet. In the lowlands you do not go outside without wearing a hat as the sun will actively try and kill you. When a PNG friend visited us in Melbourne, it was a 30°C day, we were in shorts and t-shirts, he was in jeans and wearing a parka to keep warm.

Canadian here as well I see some people in shorts at 10C maybe even at 0C mostly "dudes" in their 20s and late teens. I even see bigger (fatter) guys in shorts in the middle of winter now it seems to be a thing.

After a cold winter 0C even -10C feels warm especially if there isn't any wind.

I don't know anyone who zips up their jacket unless it's -15C or -10C and windy.

It's the wind that gets you, -20C is nothing if there isn't any wind.

But really it also depends a lot on perspective since 10C in the spring feels very warm but 10C in the fall feels very cold. Maybe a lot of the descriptions depend on year round average temperatures, I can image people in southern California with a smaller temperature range would not feel as much of a difference as a Canadian would.

Not bad.

As a Brit:

-20: Inside a food freezer.

-10: Ski trousers weather. Doesn't happen where I live.

-5: There'll be snow, or at least frost. Hat & gloves weather. Happens one or two days a year.

0: Freezing, usefully. Full-length coat.

5: Pretty cold - not unusual for a winter day. Fridge temperature. Don't forget your jacket.

10: A warmer winter day, or typical early spring.

15: Spring/Autumn; temperature where I'd stop bothering with a jacket.

20: "room temperature" for science (though many actual rooms are a degree or two higher than this). Comfortable.

25: A sunny summer's day. Pleasantly warm.

30: Starting to get towards uncomfortably hot. Notwithstanding this year, I'd only expect a few days this warm in a year.

35: Doesn't happen where I live. You want to be in the pool or somewhere with AC when it's this hot.

40: On a person, serious fever.

60-70: Nice for a sauna. Warm drink (e.g. tea)

That's pretty much it, except 90C sauna is quite decent.

Hmm, I have the same intuitive feeling with Celsius. I guess it depend on which temperature system you grew up with.

As I suggested on a parallel reply, could you say a bit more about why you feel it makes sense, perhaps with a chart like the one I wrote up?

All natural extremes fit within +100 and -100 beautifully.

110 - sauna is too hot

100 - water boils in one atmosphere

50 - desert hot (record is 56.7 °C (134.1 °F) Death Valley, California, U.S.A. 10 July 1913)

30 - hot (for a finn)

25 - nice hot

20 - use more clothes below this

0 - water freezes, ice forms on roads, lakes etc.

-10 - cold but OK with good clothes

-20 - this starts to be so cold were you need to prepare yourself well if you want to spend more than an hour outside. Temperature inside freezer.

-30 - extreme colds start. You don't enjoy this even when prepared and using good clothes.

-40 - Winter War conditions

-50 - record breaking extremes

−89.2 °C (−128.6 °F) Vostok Station 21 July 1983

-110 Kriotherapy room for treating joint inflammation

F was designed for humans, while C is apparently designed for water. Half of your list talks about what water does, or are fairly meaningless numbers to humans. While pretty much every number on the GPs list is relevant to humans. That is the big difference. In day to day life, the average human doesn't really need to know what temp water changes state. They want to know how comfortable they will be.

I take it that you never tasted a cucumber partly frozen because your fridge went a bit overzealous...

We (and most things we eat, and all that white shitty stuff that blocks your commute in a winter day) are made mostly of water. The difference between -1C and 1C can be pretty significant, unlike the difference between -1F (fucking cold) and 1F (duh, fucking cold).

* Well, of course all of this is silly pissing match between groups of people who grew up with their favorite measure system, but anyway...

Humans are 80% Water.

This pretty good. I'd put in 40, so hot you forget it gets this hot (for northern europeans).

i should move to northern europe :\

40 (as a max) was a not overly bad summer day

For the British:

100 - Boiling water, cuppa tea

50 - Lawrence of Arabia

25 - The Great British Summer! Won't last, anyway

12.5 - Wear a jumper

6.25 - Wear a parka

0 - water freezes

-6.25 - Heathrow shuts down

-12.5 - Daily Mail reports weather causes cancer

-25 - Coldest reported weather in the UK

-50 - "I am just going outside and may be some time"

Well, I think it's just the same as you except 0 is the freezing point and not what seems like an an arbitrary 32 Farenheit for us.

Hmm, I just learned that below 10 degrees C it's cold, 20 is ok and 30 is hot. 0 is water freezing, 100 is water boiling.

Maybe I'm missing out on something but I never felt the need for a chart.

From where I live, in C:

50 Summer time

40 Average hot day

30 Average cool day

20 WTF is going on? Did hell freeze over?

It never goes below twenty.

50C summer time???

Where do you live? The sahara desert?

BTW you seem to have a greek name but you definitely don't live in Greece! We do have some hot weather here, but definitely not in the range you described.

Anything feels right when you are used to it, for the rest of the world Celsius feels right, you could say it's because you know anything starting by "minus" is cold as well as any single integer (0-9). Some sort of familiarity-bias.

Fahrenheit has a couple nice properties. One is that he picked two conveniently available reference points — melting ice and body temperature — and put them 2⁶ degrees apart, which made it practical to label thermometers by bisection.

The other is that (after a very slight adjustment) freezing and boiling water are 180° apart, which is a highly composite number.

Body temperature isn't constant. Even excluding abnormal cases (eg fever), normal body temperature not only is within a range, but even this range depends on the body part you use for your readings.

Instead with Celsius at 0°C water freezes, at 100°C water boils. Thus Celsius makes more sense because the main range (0-100) covers both very cold and very hot temperatures from a human perspective.

Furthemore it has a straightforward conversion to Kelvin.

Consider, in 1725, reliably dividing a measured distance into 100 segments. Repeated bisection is practical.

I find that extremely difficult to believe without some sort of external verification. Consider that the Rømer scale from 1701 had 60 degrees between freezing brine and the boiling point of water.

Any technology which can find 60 divisions can find 100 divisions.

For that matter, Pierre Vernier invented his calipers nearly a century earlier, and William Gascoigne the micrometer in the mid-1600s, so I doubt that 100 divisions was a hard thing by 1725.

(To be certain, 64 divisions is easier than 100, but I believe it was reliably possible to do the latter in the early 1700s.)

I thought the purpose of the OP was to justify using Fahrenheit in 2014, not in 1725. :)

But even in 1725 I think that people who could build a thermometer would probably be able to grade it properly.

As humans we built the Antikythera some 2500 years ago, and I'm supposed to believe dividing something in 100 parts in 1725 was hard?

Move to a different climate and your mapping may change a lot. 30 degrees centigrade below zero can feel cozier than 10 degrees above, if the latter comes with rain and wind gales, and 30 degrees above can feel hotter than 40 degrees above if the former comes with high humidity.

I know people who lived in Siberia, but complain about the cold weather in the Netherlands at times when it isn't even freezing.

Celsius is metric at its weakest only in the sense that Fahrenheit is the imperial system at its not-worst. Everyday temperatures don't vary enough to create the need for multiple units, like what happened with distances, so there's not random conversion factors strewn around. You never care about millidegrees, and you never see kilodegrees.

Incidentally, the same "beautiful bands" appear in Celsius. They're just separated by 5s instead of 10s. It's actually kind of hard to make a temperature scale that's bad for everyday usage, assuming you put zero between or near freezing vs boiling.

You're obviously not from Minnesota. 50s is shorts and a t shirt weather.

Celsius...

0: Freezing point of water

100: Boiling point of water

A lot more precise than the rationale you proposed.

> 100: Boiling point of water

Nope! Boiling point of water at 760 Torr. In contrast, Fahrenheit uses a frigorific 1:1:1 mixture of water, ice and ammonium chloride as the second reference point, which does not depend on ambient pressure.

But which conversely no one in day to day life ever uses or is likely to come across, whereas the vast majority of the world's population still lives as close as it can to sealevel.

>1:1:1 mixture of water, ice and ammonium chloride

mmm, my proffered drink! I always keep one handy in the kitchen.

I stand corrected.

Fahrenheit is absolutely bizarre outside US, scaling is so arbitrary it's not funny.

It seems to me that the most common general use of these units is to describe ambient outdoor temperatures, and while I admit that I haven't traveled widely, I'd be astonished to discover that most countries outside the US have physical environments more reminiscent of, say, Mercury or Venus, than what we in the US parochially consider to be Earth's standard range of atmospheric temperatures.

I suppose that's why I consider Fahrenheit to be a very appropriate scale - since it centers nicely on the range of temperatures one might actually experience on Earth, naively assuming that America is representative of the rest of the planet - and isn't quite so arbitrary as one indexed to the temperature of boiling water, a phenomenon which a person in America - and, presumptuously, the rest of the Earth - isn't likely to encounter without having intentionally induced it.

But perhaps -15 and 40 really are more sensible positions for roughly opposite extremes, and we really are the odd ones for using 0 and 100.

Could you explain what you mean? I think my comment does a pretty good job of demonstrating that the scalig is nicely suited to the temperatures a human being experiences over the course of each year.

The main problem comes with negative Celsius temperature (basically 32F) as freezing point is quite important traffic wise. Driving on ice/snow is quite a change and basically occurs around 33F-32F, in Celsius it that's zero, much easier to remember.

Then you have 100F which is supposed to be the normal body temperature, but actually it means having fever.

There is only one case you might wish to use a tenth of Celsius in every day life and that's reporting body temperature around 37C-39C.

Like I've said, scaling is weird, it doesn't make much sense when firstly introduced. I have a friend who graduated MIT (PhD in aerospace engineering), living in the US for over 10 years and he is comfortable doing calculations and every day estimations in imperial units, except for temperature.

If you had to use decimals in Celsius for everyday measurements then you might have a point, but since you generally don't it really doesn't make a difference.

There's easily enough precision in the Celsius scale for everyday measurements, so it all comes down to what you're used to.