Can We Change Our Measurements?

[H3LL]

Bit of fun...

All our measurements have come to be in rather a round-a-bout manner and older, less accurate, measures have been shoe-horned into modern, precise methods of measuring.

This post was prompted by looking at the definition of two SI measurements; the metre and the second.

A metre = The distance travelled by light through a vacuum in 1/299,792,458th of a secondA second = 9,192,631,770 vibrations of a caesium atom measured by an atomic clockWhy not round them up to 1/300,000,000th of a second and 10,000,000,000 vibrations respectively? Second first, of course.

The metre I see no problem with, as it is purely a linear measurement and can be adjusted with ease.

The second a little more of a problem as it, by definition, must be a division of a year.

However, as we have; a sidereal year, a tropical year, a between two March equinoxes year, a between two September equinoxes year, a between two June solstices year, and a between two December solstices year - all of different lengths, coupled to the fact that Earths orbit varies each year, a few vibrations here or there shouldn't matter.

Changes have been made to all our measurements regularly, so we know that it can be done.

Would it help in any meaningful way?

Are there any other measurements that could do with an overhaul?

Decimal time/calendar springs to mind.

Thoughts, observations and suggestions please. Even though it is unlikely to happen....or is it?



ETA: Amused to find I could state "Second first of course" and make sense...Probably.
.

[Metullus]

I'm for it if it helps me get chicks. Otherwise, why bother?

[H3LL]

I'm for it if it helps me get chicks. Otherwise, why bother?

Sorry...No good...Important parts would be in ever-so-slightly smaller measurements...I think...But the lies would be bigger.

.

[Metullus]

I thought that size doesn't matter. That's what I was told anyway.

[OnlyTellsTruths]

What I never fail to forget is how they use light to measue atomic vibrations to determine time, but then use that time determination to measue light which is what they used to measue time. Confused?

[Dave_46]

<snip>
Changes have been made to all our measurements regularly, so we know that it can be done.
<snip>

Would it not be more correct to call them refinements, rather than changes. The "changes" are to increase accuracy rather than to change the measurement.

Dave

[El Greco]

The measurements were defined so that they would fit the practical use of the units. If a minute had 73 or 100 seconds we would have counted how many caesium vibrations that "second" would be. Your question is like asking why can't we perform plastic surgery on a woman's breasts so that her measurement would be exactly 80cm instead of 79.8 that it is now. The meter is used to measure the breasts, and not the other way around. :D

[Cuddles]

Changes have been made to all our measurements regularly, so we know that it can be done.

The thing is, they haven't actually been changed much, especially not recently since we've been capable to measuing with this kind of accuracy. The reason they have such apparently silly definition is precisely to avoid actually having to change anything. The biggest problem is really computers. If the length of a second changed, pretty much every single piece of hardware and software would need replacing. Think of the problems simply needing 4 digits in the year caused. Just imagine how much harder it would be to actually change the nature of time itself.

Technical difficulties aside, there are two other points. Firstly, look how well we're doing at converting to metric. Consdering that even countries like Britain which are officially supposed to be metric still work in miles, pounds and feet, what's the chance of anyone changing to yet another new system, especially when it's even harder to do so?

Secondly, what's the point? So the second is defined as a funny number of vibrations. So what? Bear in mind that those are exact definitions, they aren't silly irrational numbers rounded off. That's exactly as easy to measure as a nice round number, so what difference does it make? Why waste time changing something that will cause major disruption, but won't actually have any benefits?

The second a little more of a problem as it, by definition, must be a division of a year.

That's a rather odd statement. You've already given the definition of a second, and it has nothing to do with years. In fact, the second certainly isn't a division of a year, which is why we have to add leap seconds every now and then.

What I never fail to forget is how they use light to measue atomic vibrations to determine time, but then use that time determination to measue light which is what they used to measue time. Confused?

They actually measure frequency, not time. While light is used to make the measurement, the velocity of light is irrelevant. Essentially, what is measured is the peaks in the light wave. Imagine looking at waves in the sea, it is very easy to count how many peaks pass you, even though you might have no idea how fast they are going or how far they travel. Measuing how many peaks in light waves pass you is a little trickier, but it is essentially still just a counting exercise. To determine how long a second is, all that is done is start a stopwatch, count until 9,192,631,770 peaks have passed and then stop the stopwatch.

Interestingly, the definition of a second may actually change soon, since newer atomic clocks using other elements, notably strontium, could have accuracies thousands of times better than the current caesium fountain clocks.

[H3LL]

Interesting comments so far.


That's a rather odd statement. You've already given the definition of a second, and it has nothing to do with years.

This very point was a bit of a puzzle. What was the measure of a second before the caesium vibrations were used? If it was so accurate and reliable before that the difference of a few vibrations were important why change? If not, why not round up? It is a very tiny difference.

As to why? No better reason than it's easier to remember 10 billion than 9 billion, 192 million, 6 whatever blah blah. :confused: ;)

.

[DavidS]

That's a rather odd statement. You've already given the definition of a second, and it has nothing to do with years. In fact, the second certainly isn't a division of a year, which is why we have to add leap seconds every now and then.
This very point was a bit of a puzzle. What was the measure of a second before the caesium vibrations were used? If it was so accurate and reliable before that the difference of a few vibrations were important why change?
Actually, it wasn't all that long ago that the Second was defined in terms of the year. A particular year, in fact (leap seconds are as much to compensate for variation in length of the astronomical year than that a year isn't an integral number of seconds, IIRC). As standards go, however, it was clumsy and indirect because nobody can go back to 1900 to compare a measure directly to the standard. The cesium standard is something someone can set up in a lab for direct comparison to their timepiece.
If not, why not round up? It is a very tiny difference.
Are you really brave enough to tell everybody how much precision they need for their applications? Where do you draw the line - a few percent? Parts per million? Billion? Trillion?

The idea was not to change the amount of time we call one second, but to relate that predefined amount of time to a stable and reproducible standard. Once that relation is defined, anybody who can compare their timepiece to the standard can achieve whatever precision their facilities -- and the standard itself -- can achieve.

[Ziggurat]

Would it help in any meaningful way?

No. As Cuddles pointed out, an integer is an integer. I'll add that 1010 is only a round number in base 10. Furthermore, as already mentioned, we're probably going to change the standard in the future. Supposing we had rounded off to 1010 vibrations of Cesium, we'd now have something non-round with our new element. Do we round off again? That would be expensive to make such a large change. And if we change it again in the future, do we keep rounding off, making large changes in our definition each time? Bad idea. Whereas if we try to maintain the length of a second and just keep it to the closest integer each time, we're only altering it by about 1 part in 1010, so changes in the definition don't really affect clocks calibrated to the old standard (since pretty much all non-atomic clocks have worse accuracy than that).

Lastly, of course, most clocks are not atomic clocks. They may be calibrated with atomic clocks, but since they are not themselves atomic clocks, using round numbers on the atomic clocks is really of no use in the operation of non-atomic clocks.

[RenaissanceBiker]

I think we should change to a base 12 number system.

[shadron]

Actually, it wasn't all that long ago that the Second was defined in terms of the year. A particular year, in fact (leap seconds are as much to compensate for variation in length of the astronomical year than that a year isn't an integral number of seconds, IIRC). As standards go, however, it was clumsy and indirect because nobody can go back to 1900 to compare a measure directly to the standard. The cesium standard is something someone can set up in a lab for direct comparison to their timepiece..

Actually, I think the time standard, before the atomic clocks came into use, was based on the length of a day, as measured by the passing of a particular star across a meridian as observed by a specialized telescope. That would make it the sidereal day, not the solar day. I think the accuracy at that time was measured to 5 or 6 decimal places. We're talking 1940's and 50's here.

From the NIST: "This and similar work elsewhere in the world eventually led in 1967 to a an international redefinition of the second as "the duration of 9 192 631 770 periods of the radiation corresponding to the transition between the two hyperfine levels of the ground state of the cesium-133 atom. Improvements in NIST Frequency Standards over the next five decades proceeded at a rapid rate (accuracy improvement of better than an order of magnitude every ten years) culminating in today's standard, NIST-F1, with an accuracy of less than one second in 30 million years (less than 1 part in 1015)."

Notice the current accuracy is far above the standard (which was defined so by astronomical and pendulum standards of the day), which is an integer number of oscillations, and has not changed since it was defined. It hasn't changed, but we can measure the second to higher accuracy with every improvement in the technology.

(http://tf.nist.gov/general/museum/847history.htm)

[OnlyTellsTruths]

They actually measure frequency, not time. Ok.

While light is used to make the measurement, the velocity of light is irrelevant. Essentially, what is measured is the peaks in the light wave. OK?

Imagine looking at waves in the sea, it is very easy to count how many peaks pass you, even though you might have no idea how fast they are going or how far they travel. Measuing how many peaks in light waves pass you is a little trickier, but it is essentially still just a counting exercise.

Now imagine instead of measuring the waves in the sea with light (vision), try measuring them with another set of waves in the sea. Doesn't the analogy break down a bit?

To determine how long a second is, all that is done is start a stopwatch, count until 9,192,631,770 peaks have passed and then stop the stopwatch.

Oh is that all.

[H3LL]

Thanks for the funny and sensible stuff. Fascinating.

I think we should change to a base 12 number system.

:D Very sensible. Much more flexible. Would we get extra fingers into the bargain.

As long as we don't go mad and revert back to the Sumarian sexagesimal numbers. :eek:

So, you lot don't like the idea of rounding-up so that I can remember.

Can I mess with the calendar instead?

;)