Time Measures

[Resha Caner]

SI defines a "second" as 9,192,631,770 cycles of the radiation produced by the transition between two levels of the cesium 133 atom.

Obviously that measure wasn't chosen for its ease and accessibility. As I understand it, cesium was chosen because of its stability, i.e. the repeatability of using that as a measure. If I'm wrong, please correct me.

I have two questions, then. First, if ease and accessibility was not a concern, but rather maximum repeatability of the measure, is cesium better for some reason than using the speed of light to define "second".

Next, do we know why cesium is so predictable and the decay of other radioactive elements is not as predictable?

 

[durangodawood]

Interesting questions.

 

[Radagast]

SI defines a "second" as 9,192,631,770 cycles of the radiation produced by the transition between two levels of the cesium 133 atom.

Obviously that measure wasn't chosen for its ease and accessibility. As I understand it, cesium was chosen because of its stability, i.e. the repeatability of using that as a measure. If I'm wrong, please correct me.

It was chosen because of the existing technology of accurate caesium-based "atomic clocks."

is cesium better for some reason than using the speed of light to define "second".

Having defined the second that way, the speed of light is used to define the metre.

Next, do we know why cesium is so predictable and the decay of other radioactive elements is not as predictable?

Nothing to do with radioactivity. This is talking about 9.192631770 GHz radio waves (actually, microwaves) emitted by atoms.

 

[Resha Caner]

It was chosen because of the existing technology of accurate caesium-based "atomic clocks."

So it was chosen for pragmatic reasons. OK. But that just kinda kicks the can down the road. Why was the existing technology of atomic clocks based on cesium?

Having defined the second that way, the speed of light is used to define the metre.

Sure. I understand you have to start somewhere. But I'm asking if the speed of light is a more dependable parameter than the transition of cesium. e.g. if using cesium means we'll typically err +/- x% in measuring a second, would using light produce an error greater or less than x%?

Nothing to do with radioactivity. This is talking about 9.192631770 GHz radio waves (actually, microwaves) emitted by atoms.

OK. But doesn't the transition between levels cause the cesium atom to emit something? A photon or something? And isn't it the frequency of that emission that is detected?

Regardless, I'm asking: why cesium? Does it have a quality superior to lithium or some other element?

 

[essentialsaltes]

SI defines a "second" as 9,192,631,770 cycles of the radiation produced by the transition between two levels of the cesium 133 atom.

Obviously that measure wasn't chosen for its ease and accessibility. As I understand it, cesium was chosen because of its stability, i.e. the repeatability of using that as a measure. If I'm wrong, please correct me.

Not stability or repeatability, exactly, but the precision and accuracy with which it can be measured. How many decimal points we can accurately measure out to.

I have two questions, then. First, if ease and accessibility was not a concern, but rather maximum repeatability of the measure, is cesium better for some reason than using the speed of light to define "second".

Yes. Using the speed of light as a constant, in order to turn that into a unit of time, we need a precisely defined unit of distance. Or, we could use a precisely defined unit of time, in order to turn that into a unit of distance. With fast electronics, we're actually better at measuring time than measuring distance. So that's how we do it.

Next, do we know why cesium is so predictable and the decay of other radioactive elements is not as predictable?

As someone else noted, it's not a radioactive decay, but an electromagnetic transition.

As an alkali metal, cesium has a single outer shell electron, making it simple in that sense. (Rubidium, another alkali metal, is also commonly used for atomic clocks).

Cesium also only has one naturally occurring isotope, so different isotopes don't confuse the frequencies.

 

[essentialsaltes]

Sure. I understand you have to start somewhere. But I'm asking if the speed of light is a more dependable parameter than the transition of cesium.

The speed of light is so dependable, it's constant. In modern units, the speed of light is exactly 299,792,458 m/s.

It turns out we can measure time better than distance. So we have a definition for the second based on cesium, and then use that and the fixed speed of light to define the meter.

 

[Radagast]

So it was chosen for pragmatic reasons. OK. But that just kinda kicks the can down the road. Why was the existing technology of atomic clocks based on cesium?

See here.

Sure. I understand you have to start somewhere. But I'm asking if the speed of light is a more dependable parameter than the transition of cesium.

The speed of light could work, if we had a sufficiently accurate definition of the metre that didn't also use the speed of light. But caesium clocks work really well in practice.

OK. But doesn't the transition between levels cause the cesium atom to emit something? A photon or something? And isn't it the frequency of that emission that is detected?

Correct. A microwave photon, at 9.192631770 GHz.

 

[Radagast]

The speed of light is so dependable, it's constant. In modern units, the speed of light is exactly 299,792,458 m/s.

Since we've set it that way by the definition of the metre, we now will never know if it changes slightly...

It turns out we can measure time better than distance. So we have a definition for the second based on cesium, and then use that and the fixed speed of light to define the meter.

Good point.

 

[essentialsaltes]

Since we've set it that way by the definition of the metre, we now will never know if it changes slightly...

Presumably, we'd find that we have to keep making our meter sticks longer and longer (or shorter and shorter).

 

[Resha Caner]

Not stability or repeatability, exactly, but the precision and accuracy with which it can be measured. How many decimal points we can accurately measure out to.

Well, we're depending on repeatability at some point. If measuring cesium gave us 1 second one time and 1.5 seconds another time, we probably wouldn't use it. I realize the differences wouldn't actually be that large, but you get the idea. I was wondering if, in that sense, any element would serve equally well, and if there is some other reason for choosing cesium.

Cesium also only has one naturally occurring isotope, so different isotopes don't confuse the frequencies.

And maybe this is at least part of the reason cesium was chosen. It makes sense.

Using the speed of light as a constant, in order to turn that into a unit of time, we need a precisely defined unit of distance.

Yes, I get that. So, the speed of light is accepted as a constant. I'll try again. Is the frequency of the photon emission caused by the transition in cesium also thought to be a constant, or can it vary?

As someone else noted, it's not a radioactive decay, but an electromagnetic transition.

And what exactly is an electromagnetic transition?

With fast electronics, we're actually better at measuring time than measuring distance.

That's interesting. Why is that?

Since we've set it that way by the definition of the metre, we now will never know if it changes slightly.

Presumably, we'd find that we have to keep making our meter sticks longer and longer (or shorter and shorter).

Yes, I've wondered the same thing. Since time & distance measurements use those phenomena as their reference, would we ever know if they changed? I can't puzzle out exactly how we would know they've changed.

 

[FrumiousBandersnatch]

Well, we're depending on repeatability at some point. If measuring cesium gave us 1 second one time and 1.5 seconds another time, we probably wouldn't use it. I realize the differences wouldn't actually be that large, but you get the idea. I was wondering if, in that sense, any element would serve equally well, and if there is some other reason for choosing cesium.

They all have pretty reliable transitions; it's more a question of ease of use and frequency - the higher the frequency the better. Rubidium has also been used, but Strontium has far higher frequency (about 48,000 times higher!), so makes a far more precise clock, but is much more difficult to manage as a clock.

Since time & distance measurements use those phenomena as their reference, would we ever know if they changed? I can't puzzle out exactly how we would know they've changed.

In principle, you could compare with other related constants. Whether you could detect a difference would depend on what was actually varying (i.e. whether it was common to both constants), and the limits of experimental resolution. If the frequency had detectable cosmological-scale implications, you could compare projections of the current value back in time with the observed effects at increasing cosmological distances (i.e. looking back in time).

 

[Almost there]

I'm no fan of cesium. If I leave it in the milk too long it gets soggy.

 

[Radagast]

I'll try again. Is the frequency of the photon emission caused by the transition in cesium also thought to be a constant

Of course.

And what exactly is an electromagnetic transition?

An outer electron of a caesium atom absorbs a photon of a specific frequency and jumps up an energy level.

In practice, we use this phenomenon to adjust the frequency of a more traditional electronic oscillator:

You will notice that all the components of that device are things that we're good at making (quartz oscillators, microwave emitters, magnets, etc.).

 

[Radagast]

Rubidium has also been used, but Strontium has far higher frequency (about 48,000 times higher!), so makes a far more precise clock, but is much more difficult to manage as a clock.

We want that column of the periodic table of course, so that there will be just one outer electron, which reduces interference from other electrons.

 

[Resha Caner]

In principle, you could compare with other related constants. Whether you could detect a difference would depend on what was actually varying (i.e. whether it was common to both constants), and the limits of experimental resolution. If the frequency had detectable cosmological-scale implications, you could compare projections of the current value back in time with the observed effects at increasing cosmological distances (i.e. looking back in time).

Comparing across different cosmological distances is an interesting idea, but it's not as if we're actually travelling back in time. We're just observing the effects from past times. So, those effects could have changed along with everything else by the time they reach us.

I suppose we don't really know. For example, is the speed of light the same everywhere because all photons were produced by the same event or because some "law" governs their behavior everywhere? i.e. if a photon near Polaris changed, would that be because its behavior is independent of a photon near Vega ... or would it mean both photons would experience the same change because they all obey the same laws?

 

[essentialsaltes]

For example, is the speed of light the same everywhere because all photons were produced by the same event or because some "law" governs their behavior everywhere?

More the latter. Photons have no rest mass. Any particle with no rest mass will travel at the speed of light (in vacuum). This is essentially a result of relativity.

 

[Resha Caner]

More the latter. Photons have no rest mass. Any particle with no rest mass will travel at the speed of light (in vacuum). This is essentially a result of relativity.

OK. What other particles have no rest mass? And what distinguishes them from photons?

I did think about the vacuum reference WRT cesium, but too many questions were bombarding my brain at once. It seems we're always referring to speed in a vacuum even though a perfect vacuum is a rare thing and the speed of light changes in other mediums. So, isn't the cesium transition affected by the medium as well?

 

[essentialsaltes]

Well, we're depending on repeatability at some point. If measuring cesium gave us 1 second one time and 1.5 seconds another time, we probably wouldn't use it. I realize the differences wouldn't actually be that large, but you get the idea. I was wondering if, in that sense, any element would serve equally well, and if there is some other reason for choosing cesium.

So if we repeat the measurements, we will get a range of answers. This is the basis of our uncertainty in the measurement, which I'm recasting as accuracy and precision. If we get 45,44,46,45,46,45,44,45,45, then it's 45 +/-1. If we get 45.0001, 45.0000, 44.9999, and so on, then it's 45.0000 +/- .0001

In principle, the measured values would fall into something like a Bell curve. The narrower the Bell curve, the more precise the measurements can be.

I have read that this particular hyperfine transition in cesium has a particularly narrow Bell curve. The resonance has a high Q value.

Yes, I get that. So, the speed of light is accepted as a constant. I'll try again. Is the frequency of the photon emission caused by the transition in cesium also thought to be a constant, or can it vary?

It should be constant, the frequency is tied to the energy levels of the different quantum states, which presumably are fixed by Planck's constant, etc.

And what exactly is an electromagnetic transition?

These are the electrons jumping up and down energy levels in the atom. When bathed in electromagnetic waves, some electrons absorb energy and jump to a higher level. When they jump back down to the ground state, they release that energy in the form of an electromagnetic wave of a precise frequency. Basically, it's much the same as a neon light. You excite the neon with energy, and the release of that energy comes out in the particular color we associate with neon lights. Neon lights with neon in them are reddish. Mercury lamps are bluish. Sodium lamps are yellow.

That's interesting. Why is that?

Physically measuring distance with a yardstick, or even finely tuned calipers is pretty limited in precision. How finely can you draw lines on the yardstick? But chopping up time is easy with even home computer processing that runs in GHz, so that individual cycles are billionths of a second. I think most careful distance measurements are probably really time measurements. You bounce light off the other end of the thing, measure the time, divide by two, and multiply by the speed of light.

Yes, I've wondered the same thing. Since time & distance measurements use those phenomena as their reference, would we ever know if they changed? I can't puzzle out exactly how we would know they've changed.

If they both changed the same way at the same time, we'd never know. But it might be fair to ask whether anything actually changed.

 

[Chesterton]

SI defines a "second" as 9,192,631,770 cycles of the radiation produced by the transition between two levels of the cesium 133 atom.

Maybe you'd have fewer questions if you got your time measurements from somewhere other than Sports Illustrated.

 

[FrumiousBandersnatch]

... is the speed of light the same everywhere because all photons were produced by the same event or because some "law" governs their behavior everywhere? i.e. if a photon near Polaris changed, would that be because its behavior is independent of a photon near Vega ... or would it mean both photons would experience the same change because they all obey the same laws?

It came about from the idea that the laws of physics should be the same in all frames, which means the laws of electromagnetism should be the same, which means the calculated values associated with them should be the same, including the speed of light.