That is not even wrong, however subjective your definition of accurate.
Anders Celsius did, for a start, and with good reason, I would argue.
And as a random fact I just stumbled across, Anders Celsius’ thermometer was calibrated with a value of 0 for the boiling point of water and 100 for the freezing point. In 1745, a year after Celsius’s death, the scale was reversed by Carl Linnaeus to facilitate more practical measurement.
There was a thread that contained a long discussion on the pros and cons of the Fahrenheit and Celsius scales. Am too lazy to find it at the moment, but I believe the conclusion was that the Fahrenheit scale was more convenient for weather, since (for most locations, most of the time) the temperature is quantified using a positive integer between 0 and 100. No negative signs, no decimal places, and just two digits (with the obvious exception of the end points).
I remember that thread, did not read it till the end, though: I got repetitive and did not convince me. No big deal, keep your favourite unit if you like it, I’ll keep mine. My beef was actually with the claim that one temperature scala is more accurate than another by virtue of having more subdivisions. That is not what the word accurate means to me.
Another random fact about temperature: Anders Celsius’ original thermometer still works and can be seen in the Museum of the Uppsala Unniversity, the so called Gustavianum. It looks like most modern thermometers: A glass capilar with a mercury deposit fixed on a wooden plank with a scale.
Not to hijack this thread, but a valid argument can be made that the Fahrenheit scale is more accurate than the Celsius scale if a given measurement is always rounded to the nearest integer. This assumes the measurement itself is very accurate.
[bolding italics mine]
This looks rather like an argument in favour of Celsius, except perhaps in places with long cold winters. Negative Celsius are the exception in the places I (subjective!) usually am.
[third bolding italics mine]
You cannot argue at the same time that the advantage of Fahrenheit is that they have no decimal places and impose the rounding of the Celsius, thus impeding the decimal places that would make it more… no, accurate is not the word, that was precisely my point, but whatever.
I mean, yes, you can argue like that, but it is not coherent.
Nitpick:
You can always make an argument, whether it is valid or not should not de decided a priori by the one presenting the argument.
That is not an argument, it is not even contradiction! Rather the opposite, even.
I’m sure there are people in the world who will wear a sweater if the temperature is 67 degrees but not if it’s 68 degrees, but I suspect they are firmly in the minority.
But how often does something get rounded to the nearest integer? You can bet that the clerk isn’t going to round your $10.40 laundry detergent down to a tenner. And the stock market would be chaos if suddenly they started reporting “Pfizer opened at 53-ish, which is less than yesterday’s 53-ish.”
I’d enjoy the greater accuracy of using a decimal point when discussing temperature. So (even though ah’m a ray-id blooded Amur’can) I’d love to switch to Celsius. I’d get used to it 14.2 days before I died…
And yes, I’ll bet I would notice the difference between 18.2 and 18.3 most-of-the-world degrees.
It’s technical name was pulse dialing. Which is why phones (at least used to?) have a little switch marked T-P for Tone or Pulse dialing back when some places still didn’t support tone dialing.
This also makes an appearance in the novel Red Dragon: Hannibal Lector’s demand to speak confidentially with his lawyer has to be respected; so they give him privacy in a room with a phone with no dial and the connection to his lawyer’s number dialed from the facility switchboard. Lector promptly hangs up and prong-dials the number he wants.
Fahrenheit degrees being smaller units than Celsius ones does not make them more accurate, it makes them more precise. How finely divided units are is a measure of their precision. Accuracy is whether or not the value measured with a device is truly what the device says it is.
When I was in college, the dorm room telephones (each room had it’s own phone) were on their own exchange, and only the last four numbers were needed to dial between rooms. That’s not the interesting part, though. For phones that were on this exchange, if a line was busy, you could “cut in” if you pulsed the switch-hook, as described above. The easy way was to let the rotary do most of the work. That is, if the number you wanted was 7654, you could dial 7653 and hit the switch-hook just as the dial stopped after the 3. If the other line was in use, you would be added to the call (and I don’t think there was any notification for those already on the line).
I only really only used this trick once. I had a buddy who had a habit of taking his phone off the hook if he was taking a nap. The phones did not give that annoying sound that was later adopted to alert you if the phone was off the hook, it only gave a dial tone, which stopped after a minute or so. Anyway, we had made plans for the evening that needed changing, but his phone was busy. So I waited 10 minutes and called again. Busy. I figured he was asleep with the phone off the hook, so I tried the trick and sure enough, I got on the line, but nobody was talking. So, started whistling and yelling into the phone to wake him up. It worked, but he wasn’t too happy about being awakened. As the change in plans was to his benefit, so he got over it quick enough.
Nymphing refers to a type of fly fishing, and not to any sort of girls-gone-wild scenario. 
Fly fishing refers to a type of fishing with fake insects, and not any sort of aircraft.
I made the same mistake about fly girls.
You’re incorrect if the data are rounded, as I stated above.
Let’s say you are the operator of a NOAA weather station. Part of your job is to report air temperature. So you use a very accurate platinum resistance thermometer (PRT) sensor to measure it. The sensor + digital readout has a resolution of 0.01 K and an uncertainty of ±0.01 K. So you might measure 286.45 K, for example. Or 294.32 K. The weather bureau wants you to provide them the average daily temperature in Fahrenheit or Celsius, and the numbers must be rounded to integers, e.g. 74 °F or 23 °C.
So here’s the question: from an accuracy standpoint, does it matter if you give them the temperatures in Fahrenheit or Celsius? Well, if the numbers are to be rounded to the nearest integer, then Fahrenheit will be more accurate than Celsius.
If you still don’t believe me, I ran a simulation in Excel. Here are 16 random temperature measurements. As you can see, the error of the Celsius data is greater than the error of the Fahrenheit data:
I then ran this simulation ten times. Here are the results. Hopefully it’s obvious that, if a temperature measurement is rounded, Celsius has more error than Fahrenheit.
Of course an argument can be made that, even though the Fahrenheit data is more accurate than the Celsius data, it doesn’t matter when reporting weather. I won’t argue that one. Or that the uncertainty of a consumer-grade temperature readout will be much higher than ±0.01 K, and thus it probably doesn’t matter a whole lot if it’s °F or °C. And again, no argument there. My point was simply that Fahrenheit is definitely more accurate than Celsius when the data are rounded and the inherent uncertainty of the measurement is low.
I can’t believe the NOAA would demand that a weather station give them “rounded to the nearest integer” temperature readings. Do you have any proof that they actually do that? Why would they go out of their way to ask their stations to send imprecise data?
.
eta: If you’re just having fun and saying “Well, Fahrenheit is better than Celsius… but only in this one obscure case where you have to round everything to a whole number”, then I won’t take it seriously, and I retract my questions.
But if you’re trying to make a larger point that Fahrenheit is more accurate, and hence superior, than Celsius… that’s weird.
No. I was simply using it in my example to prove a larger point.
How far out should they go, though?
Whole-integer °F is more precise than whole-integer °C.
°C is more accurate than °F if you give it in tenths of a degree.
But if you’re going to do that, than °F.1, .2, .3 is more accurate than °C.1, .2, .3.
If you’re striving for precision, though, you want to be as precise as your instruments can measure, and it doesn’t really matter which system you use. I don’t know that the NOAA is striving for precision, but even if they are, they may be working with a poorly designed database that only accepts whole integers, or mercury thermometers obvserved by eye where readings more precise than a whole degree aren’t available.
If you’re conveying temperature to a general audience, even if humans can sense the difference within a single Celsius degree, there are few circumstances where that would make a practical difference.
I prefer °F because I grew up with it and thus it’s more intuitive. My husband prefers °C for the same reason. We’re both right (or both wrong, depending on how you look at it).
Perhaps everybody knows this but I learn with interest that male giraffes sniff and taste a female’s urine to ascertain her fertility. Subject to the male’s satisfaction he follows the female around until she gives in.
With some piffling changes to societal attitudes I think this could catch on higher up the food chain.
You are correct, that was precisely my point, thank you for expressing it clearer than I could. Crafter_Man’s answer is absurd: why would somebody operating in a NOAA weather station who uses “a very accurate platinum resistance thermometer (PRT) sensor to measure” the outside air temperature be forced to round the transmitted data to Celsius or Farenheit integers? That is arbitrary, the weather bureau would not be so stupid even if it’s director was appointed by Trump!
And I don’t need a simulation to know that when you round to integers Farhrenheit you get a smaller margin of error than with Celsius compared to the 0.01 precision of you fancy platinum resistance thermometer (PRT) sensor, I can even tell you without a simulation that if you allowed decimal Celsius, those would be more precise, if in turn you allowed decimal °F, those would be more precise again until you allowed 0.01°C at which point… etc. etc.
Considering that the difference between °C and °K is 0,15 degrees (the 273 are irrelevant for error margins when converting whole degrees with no decimals back and forth, I hope you agree), if you first rounded the K to whole degrees and then converted to °C the error would be smaller than the 0.28 K you calculated. And looking at the table you simulated I find it improbable that in every single case the margin of error for °F is smaller than for °C, by sheer probabilistic chance at least one set of data should be closer to “accurate precise reality” in °C with a sample of 10 simulations with 16 readings in each if one set has a margin of error of 0.05% and the other of 0.09% as you claim.
So, that was a nice convoluted long sentence. Enough vented! Feel much better already, thank you.
Not only will I tell you a random fact, I will describe the stumble across.
We were out walking today and, looking at molehills by the path, it occurred to me that you never see them being made. Which raises two questions:
In attempting to resolve these questions, in this wiki page I came across a surprising statement:
William Buckland, known for eating every animal he could, opined that mole meat tasted vile.
At which point, who cares when moles sleep?
The remarkable academic William Buckland -
- together with equally remarkable son Frank, basically ate any and every animal they could lay their hands on.
William popularized an offbeat diet he dubbed zoophagy, which basically meant that the minister ate any creature he could get his hands on. Bear, crocodile, and hedgehog were all regular parts of the family diet. Unsuspecting guests learned the hard way that their host didn’t always bother to identify the main course by name before everyone started digging in. Still, at least one of William’s friends appreciated these bizarre meals. “I have always regretted [the] day,” wrote critic John Ruskin, “… on which I missed a delicate toast of mice.”…
…Of course, William’s adventurous appetite rubbed off on Frank. Nowhere was this fact more apparent than at the Royal Zoological Garden (today’s London Zoo). When a display animal died, Frank was usually on call to perform an autopsy. As he was dissecting, he gave the staff explicit instructions to save any and all remains that seemed appetizing. There was just one rule of thumb: “If they look good to eat, they are cooked; if they stink, they are buried.”
This system worked well. Over time, Frank checked off such entrées as viper, roast giraffe, bison, and a “whole roast ostrich.”
You would have thought that someone would publish a complete list of what they ate, together with tasting notes, on the internet; but I have failed to find one. If anyone succeeds, please post a link.
BTW#1: Frank Buckland
BTW#2: moles are not nocturnal.
This misconception is usually the result of people looking out their window in the morning and seeing fresh mole hills. In fact, moles are not necessarily more or less active at any time during the day or night. Current research suggests that moles sleep and work in 4-hour shifts. They are more active during quiet periods, such as early morning or late in the evening. When they feel vibrations in the ground, as created from people or pets walking, they will be more likely to cease their digging.
j