Should the Astronomical Unit be replaced by the Light Second/Light Minute Etc.?

[BadBrain]

Now that other solar systems have been discovered, as well as binary and trinary star systems, is the helio-geocentric Astronomical Unit still useful? Wouldn't it be clearer to use the terms light second, light minute. light day, light year, etc., instead? Obviously, these terms are still based on earthbound astronomical measurements for their time unit designation, but, seeing I can think of no rational basis for a universal time unit, it seems that clarity of presentation should prevail in deciding this matter.

 

[mathman]

I have never seen A.U. used except in discussions about our solar system.

 

[Drakkith]

I'm not sure I understand where you are coming form on this BadBrain. The AU is a set distance that can easily be converted to any other unit you want it to be. How could it get any clearer? As for a "universal time unit", we already have one, the second. Why? Because we already use it and we simply have to choose an arbitrary length of time for our unit, so why not use one we are familiar with?

 

[BadBrain]

I'm not sure I understand where you are coming form on this BadBrain. The AU is a set distance that can easily be converted to any other unit you want it to be. How could it get any clearer? As for a "universal time unit", we already have one, the second. Why? Because we already use it and we simply have to choose an arbitrary length of time for our unit, so why not use one we are familiar with?

In terms of a universal unit of time, I was hoping for something non-arbitrary. (Actually, the second is NOT arbitrary, but is the result of subdividing the observed Terrestrial mean sidereal day.) I was thinking in terms of the shake, which is the amount of time between generations of neutrons within a critical mass of a fissile substance. As I don't know whether this is dependent upon the specific fissile substance involved, I suggest Uranium, the heaviest element yet observed in nature. But this is such a tiny unit of time that the light-shake would, in turn, be such a tiny distance that the exponential value of light-shakes required to describe astronomical distances would be difficult for scientists to use and difficult for laymen to understand.

 

[Drakkith]

The second has been standardized 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 caesium 133 atom.

And by arbitrary, I just meant that we had to have SOME unit, so why not choose the one we already use.

 

[BadBrain]

The second has been standardized 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 caesium 133 atom.

And by arbitrary, I just meant that we had to have SOME unit, so why not choose the one we already use.

The unit we already use is, as I've stated above, a subdivision of the mean sidereal day, as observed from Earth, therefor, it is not universal. The shake has some hope of greater universality, as my vague guess would be that, within a critical mass of an element (actually, the U-235 isotope, itself having been observed as naturally occurring), local environmental variables would be overwhelmed by the physical effects emanating from within the interior of the uncontrolled nuclear fission chain-reaction, thus isolating those reactions from the physical effects of the surrounding environment.

 

[Drakkith]

The duration of 9,192,631,770 periods of the radiation corresponding to the transition between the two hyperfine levels of he ground state of the caesium 133 atom is as universal as you are going to get. It is also far far more accurate than this "shake".

 

[BadBrain]

The duration of 9,192,631,770 periods of the radiation corresponding to the transition between the two hyperfine levels of he ground state of the caesium 133 atom is as universal as you are going to get. It is also far far more accurate than this "shake".

You're correct! You're dealing in quanta of energy here, whereas my "shake" deals in transit time through local medium, however defined.

However, your definition merely redefines the second, a unit of time dependent upon subdivisions of the Terrestrial Mean Sidereal Day, according to quantum mechanics. Why choose 9,192,631,770 periods? Why not choose 1 period of radiation of from Cs 133 as one's base time unit for the measurement of time distance?

This doesn't solve the problem of workability and intelligibility which I've described above, but at least it's universal.

 

[Drakkith]

I don't see any problem of workability and intelligibility. What are you referring to?

 

[BadBrain]

I don't see any problem of workability and intelligibility. What are you referring to?

I am referring to the period of radiation due to electronic transition between the two hyperfine levels of the ground state of the Cs-133 atom. The number 9,192,631,770 periods is a number based upon Terrestrial observations of near-Earth astronomical phenomena. The number 1, with relation to this periodicity, is a matter of quantum mechanics, which is independent of Terrestrial observations of near-Earth astronomical phenomena, and is therefor more universal than the mean sidereal Terrestrial second.

If you don't see any problem of workability with my proposed system, then you must be working with equipment far superior to mine, for which fact you have my sincere congratulations.

 

[Vorde]

I think the key here is that there are only a handful of non-arbitrary measures of length and time (these are the various Planck measurements).

Assuming you aren't using those utterly unfeasible units of measurement, any unit or system of units you will be using is 100% arbitrary, so it makes no sense not to use units that we are familiar with.

 

[BadBrain]

I think the key here is that there are only a handful of non-arbitrary measures of length and time (these are the various Planck measurements).

Assuming you aren't using those utterly unfeasible units of measurement, any unit or system of units you will be using is 100% arbitrary, so it makes no sense not to use units that we are familiar with.

You are, of course, correct, which is the point I was trying to make, as well as the point upon which I was surrendering my initial argument.

But, a less-than-universal measurement of time, and measurements of distance based upon the transit of light through a vacuum over a measurement of time dependent upon Terrestrial observations of local astronomical activity, just leaves me feeling so UNSATISFIED!

 

[Vorde]

Satisfy yourself by knowing how frustrated you'd be if you asked for the time and got back a number in Planck time's.

 

[Chronos]

How often do you see distance expressed in AU's? It's just too tiny to be useful save in the occasional pop article. It merely serves to convey a sense of the incomprehensible vastness of the universe.

 

[Drakkith]

You are, of course, correct, which is the point I was trying to make, as well as the point upon which I was surrendering my initial argument.

But, a less-than-universal measurement of time, and measurements of distance based upon the transit of light through a vacuum over a measurement of time dependent upon Terrestrial observations of local astronomical activity, just leaves me feeling so UNSATISFIED!

Why? What is wrong with the second? What's the difference between it and any other measurement of time? Nothing other than the duration of the second is specific, and another unit simply won't be the same duration. The actual duration of the unit we use is pretty much not an issue as long as whatever we use is specifically defined and measured. The same goes for AU. And if you have a problem with that, why not abolish light-years as well? A year is measured as the time it takes the Earth to go around the Sun. We could go even further and just get rid of the parsec, the meter, and more. But, what would we use then? Why feel unsatisfied with our current units of measurement if it doesn't even matter?

 

[twofish-quant]

Now that other solar systems have been discovered, as well as binary and trinary star systems, is the helio-geocentric Astronomical Unit still useful? Wouldn't it be clearer to use the terms light second, light minute. light day, light year, etc., instead?

There's a problem, which is that if you use SI units, the conversion factors to and from celestial mechanical units have more uncertainty than the observations which are available. We can do celestial mechanics to something like nine significant digits, whereas G is known only to six.

Obviously, these terms are still based on earthbound astronomical measurements for their time unit designation, but, seeing I can think of no rational basis for a universal time unit, it seems that clarity of presentation should prevail in deciding this matter.

If you use light minutes, you just can't do the calculations to the right precisions. If you do things in AU, then the mass of the sun is 1.0000000000000 (exactly). If you do things in SI units, then we only know the mass of the sun to six digits.

 

[BadBrain]

There's a problem, which is that if you use SI units, the conversion factors to and from celestial mechanical units have more uncertainty than the observations which are available. We can do celestial mechanics to something like nine significant digits, whereas G is known only to six.



If you use light minutes, you just can't do the calculations to the right precisions. If you do things in AU, then the mass of the sun is 1.0000000000000 (exactly). If you do things in SI units, then we only know the mass of the sun to six digits.

I refer you to the answer I gave above, from which I here quote, according to my STILL EVOLVING idea (SHEESH):

"I am referring to the period of radiation due to electronic transition between the two hyperfine levels of the ground state of the Cs-133 atom. The number 9,192,631,770 periods is a number based upon Terrestrial observations of near-Earth astronomical phenomena. The number 1, with relation to this periodicity, is a matter of quantum mechanics, which is independent of Terrestrial observations of near-Earth astronomical phenomena, and is therefor more universal than the mean sidereal Terrestrial second.

"If you don't see any problem of workability with my proposed system, then you must be working with equipment far superior to mine, for which fact you have my sincere congratulations."

Please understand that my ideas on this subject are still evolving, and respond in the light of that understanding.

 

[Drakkith]

Please understand that my ideas on this subject are still evolving, and respond in the light of that understanding.

There's no need to get defensive about it. I don't see twofish's statement as condescending or anything like that, merely informative.

 

[Nik_2213]

Given that 1 AU is almost equal to eight light minutes, we do have a 'scientific' conversion factor...

IMHO, keeping the AU is handy because of the convenience when dealing with Sol-types and 'goldilocks zones'. It's a nice, round number for hand-waving.

It's the same argument over LY vs Parsecs, really. One embraces a solar system out to the Oort cloud, is handy for long-period and common-motion binaries beyond easy AU reckoning. The other is a convenient yardstick for neighbouring systems. IMHO, LY are more convenient when you look at relative positions of the neighbours, but that's just my preference.

Kilo-parsecs and their mega-parsec associates come into their own for galactic astronomy.

Just don't get kilometres and nautical miles confused...

 

[Dickfore]

Well, the astronomical unit is essential in parallax distance measurements of other stars. Namely, as the Earth revolves around the Sun, the apparent position of a star traces an elliptical trajectory on the night sky. If the angular size is $2\theta$, then the distance to the star d is:
$$\tan \left( \frac{\pi}{180 \times 3600} \theta(") \right) \approx \frac{\pi}{180 \times 3600} \theta(") = \frac{r}{d} \Rightarrow d = r \frac{206265}{\theta(")}$$
where the small angle approximation was used.

If r is chosen as a unit of length (1 A.U.), then this formula naturally gives the distance in those units. Thus, a distance of 1 parallax-second (parsec) corresponds to 206265 A.U.

Of course, if you want to convert these distances into metric units, you must measure the radius of Earth's orbit by other methods.

 

[twofish-quant]

IMHO, keeping the AU is handy because of the convenience when dealing with Sol-types and 'goldilocks zones'. It's a nice, round number for hand-waving.

It's also pretty essential for any sort of precision solar system astronomy. The good thing about the AU is that it is "local." If you try to convert from something local to something universal, you end up with a lot of conversion factors whose accuracy is uncertain, and that degrades your results.

It's the same argument over LY vs Parsecs, really.

And you have similar issues. With a parsec, you can take a measurement of a star, and then instantly convert it into a distance, without going putting in any conversion factors whose numbers are uncertain. When you do light years, then you have to do conversions whose accuracy is not certain, and that introduces a lot of issues.

One funny thing is that when you talk with observational astronomers in the United States, they talk in terms of inches. The reason for that is that if you have build an instrument, you will go through all sorts of heck trying to get metric screws. And yes, this causes big, big problems (i.e. Mars Climate Orbiter failing).

 

[Drakkith]

What conversions are you referring to? Why are they inaccurate?

 

[D H]

There's a problem, which is that if you use SI units, the conversion factors to and from celestial mechanical units have more uncertainty than the observations which are available. We can do celestial mechanics to something like nine significant digits, whereas G is known only to six.

That's overstating things with respect to G. G is known to a tad less than four significant digits, not six. The 2010 CODATA value for G is (6.67384±0.00080)x10^-11m³kg^-1s^-2, or a relative uncertainty of one part in 1.2x10⁴.

One way around this is to use the product G*M, which is known to almost ten significant digits for the Sun. Another way around this is to use IAU units, which essentially whitewashes away the problem of the lack of precision in G. In IAU units, the gaussian gravitational constant is a defined constant. Almost all of the uncertainty in the value of the AU is attributable to the uncertainty in the solar gravitational parameter, multiplied by 3/2.

 

[tiny-tim]

i think the basic unit of time should be called the tiny-tim