PDA

View Full Version : NewScientist :'Astronomical unit' may need to be redefined



Bozola
2008-Feb-06, 11:55 PM
New Scientist (http://space.newscientist.com/article/dn13286-astronomical-unit-may-need-to-be-redefined.html)

I love this kind of science. It's a reminder, to me, that everything is all interrelated.

Lil Grasshoppah
2008-Feb-07, 01:53 AM
I am philosophically opposed to the astronomical unit anyway. It is, in its way, as archaic as the Imperial units. It is based on something that has meaning only in this one speck of the whole universe and even that is variable - it must be averaged even over a year and, as the New Scientist article points out, it is truly only valid at this moment in time. Go too far back or forward in time and you would arrive at a different value.

Keep the AU if we must (it can be a useful way of discussing distances within the solar system), but base it on something less malleable, like the distance traveled by light in 8(ish) minutes. That would allow it to retain its usefulness without becoming a source of confusion for our descendants born on worlds orbiting other stars.

Ara Pacis
2008-Feb-07, 01:55 AM
Another redefinition? I don't think they've recovered from the last one.

I just ignore AUs. I'm working on a story and I just use mega-klicks (1,000,000km) as a unit of distance measurment. While, I'm here, I'll say that I use Exaton as a measure of mass for planetary comparisons.

Bozola
2008-Feb-07, 02:15 AM
I am philosophically opposed to the astronomical unit anyway. It is, in its way, as archaic as the Imperial units. It is based on something that has meaning only in this one speck of the whole universe and even that is variable - it must be averaged even over a year and, as the New Scientist article points out, it is truly only valid at this moment in time. Go too far back or forward in time and you would arrive at a different value.

Keep the AU if we must (it can be a useful way of discussing distances within the solar system), but base it on something less malleable, like the distance traveled by light in 8(ish) minutes. That would allow it to retain its usefulness without becoming a source of confusion for our descendants born on worlds orbiting other stars.


Agreed. What I like about it is that it's an "in-your-face" reminder that the solar system is a dynamic place and there's a lot of really nifty things always going on...and it just another way to appreciate the beauty of E=mc^2.

Jerry
2008-Feb-07, 02:56 AM
It would be nice if astronomers would simply use metric units. There is enough useful information that we should not forever be learning different distance systems and converting parsecs, lightyears, AUs, miles, rods, fingers, cupits, hands, feet and knots.

Simplify.

Anthrage
2008-Feb-07, 03:30 AM
Unfortunately I believe a minute, or any temporal measurement, is just as 'provincial' and unreliable as that which the AU is based on. Especially the quantity 8, again based on something as abitrary are our planet's current location.

As we as a species broaden our horizons, and as a conseqence, our possibilities, there are many such measurements, values and designations that will change. That must change. In fact, this process has already begun - the reclassification of Pluto, seen by many as a non-event at best and a farce at worst, is an example of the requirements posed on us all by our advancement and that of our knowledge.

This requirement will increase as we move forward, through time and certainly through space. Much has been considered about what calendar will be used for example, even what division of a 'day' will be adopted, as we establish colonies on planets in our solar system, where day/night cycles and seasonal periods will be vastly different from that of earth. While our circadian rhythm is a powerful dictator, it is - even now - somewhat flexible, and will no doubt be even more so as we go on, and outward, from our current home.

So I do not see any temporal foundation as being useful too far into the future - and hopefully, we humans have a very long future ahead of us. Some spatial constant perhaps may be a better choice. Something with a chance of being udnerstood, and perhaps even shared, by others, would probably be best.

neilzero
2008-Feb-07, 03:38 AM
Luckily a light year is approximately 10 to 16 th meters = ten million times a billion. It would get awkward for most of us to think about 7 trillion terra meters = 7 times ten to the 24 meters to a galaxy that is 700 million light years away. Neil

Noclevername
2008-Feb-07, 04:58 AM
Luckily a light year is approximately 10 to 16 th meters = ten million times a billion. It would get awkward for most of us to think about 7 trillion terra meters = 7 times ten to the 24 meters to a galaxy that is 700 million light years away. Neil

Awkward for those of us not used to that terminology, sure. But for those who grow up with that, converting "light-years" to real numbers is awkward.

JohnOwens
2008-Feb-07, 05:07 AM
I just ignore AUs. I'm working on a story and I just use mega-klicks (1,000,000km) as a unit of distance measurment. While, I'm here, I'll say that I use Exaton as a measure of mass for planetary comparisons.
Well, while I agree about AUs, there's definitely a place for the standard Solar mass and Earth mass. After all, the only way we have to measure planetary masses is by their gravitational effects on other bodies. So, what we really end up knowing is the product of their mass and the gravitational constant. Since we can get remarkable precision in tracking the orbits, and hence in that product, but we can only nail down the gravitational constant to about four decimals, any orbit-based claim that a planet or star weighs X number of kg/tonnes/whatever else inherently assumes a certain value of G, which you don't know unless it's explicitly stated, in which case you might as well be giving that product of mass and G (also known by mu, the Greek letter).

Ara Pacis
2008-Feb-07, 06:06 AM
Well, while I agree about AUs, there's definitely a place for the standard Solar mass and Earth mass. After all, the only way we have to measure planetary masses is by their gravitational effects on other bodies. So, what we really end up knowing is the product of their mass and the gravitational constant. Since we can get remarkable precision in tracking the orbits, and hence in that product, but we can only nail down the gravitational constant to about four decimals, any orbit-based claim that a planet or star weighs X number of kg/tonnes/whatever else inherently assumes a certain value of G, which you don't know unless it's explicitly stated, in which case you might as well be giving that product of mass and G (also known by mu, the Greek letter).

But the definition and units would be precise, even if the measurements are approximated.

AndreasJ
2008-Feb-07, 10:56 AM
I don't see how the meter, originally based on the polar circumference of the Earth, is any less provincial than the AU?

NEOWatcher
2008-Feb-07, 01:07 PM
How common is the usage of AU in the astronomical circles?
Is a parsec directly based on AU? Is parsec very common in relation to LY?

I've always thought that it was more of a popular measurement rather than a scientific one which would mean precision is not necessary.
In other words, I always thought the experts used whatever units are better for thier usage, while relating it to the general populus is just converted to something they can understand.

Dave Mitsky
2008-Feb-07, 01:12 PM
I don't see how the meter, originally based on the polar circumference of the Earth, is any less provincial than the AU?

I agree.

Besides, it's a handy distance scale coincidence that there are similar numbers of AUs in a light year as inches in a mile.

Dave Mitsky

ngc3314
2008-Feb-07, 01:45 PM
How common is the usage of AU in the astronomical circles?
Is a parsec directly based on AU? Is parsec very common in relation to LY?


Sure. At a distance of one parsec, an AU subtends an angle of one second of arc (arbitrary, blame the ancient Mesopotamians). That means that one parsec equals, by definition, 3600*180/pi = 206264.8 AU. For some reason, parsecs (and their larger cousins Mpc and Gpc) are much more commonly used in the research literature than light-travel-time distances. As if the definition of parsec were any less Earth-based than light-years.

Trivia question I seem to bring up every couple of years - without doing the numerical calculation, for which major planet in our system would the locally-based definitions of parsec and light-year be most nearly equal?

NEOWatcher
2008-Feb-07, 01:52 PM
Sure. At a distance of one parsec, an AU subtends an angle of one second of arc...
That's why I asked. Has parsec been scientifically defined in that way, or has it been locked into a more stable definition?

Celestial Mechanic
2008-Feb-07, 01:54 PM
How common is the usage of AU in the astronomical circles?
Quite common in Solar System studies.

Is a parsec directly based on AU? Is parsec very common in relation to LY?
"Parsec" is short for "parallax (arc) second" and is the distance at which an object will have an annual parallax on Earth of one second of arc. This distance turns out to be:

(1 AU)/(1 parsec) = arctan(1 arcsecond) or pi/648000 approximately, or

1 parsec = 206,265 AU.


I've always thought that it was more of a popular measurement rather than a scientific one which would mean precision is not necessary.
Actually, it is a professionally-used unit where high precision is unattainable.

In other words, I always thought the experts used whatever units are better for their usage, while relating it to the general populace is just converted to something they can understand.
Actually, in the Solar System the AU is a much better choice than millions of kilometers. 5.2 AU means 5.2 times the semimajor axis of the Earth's orbit, it means Jupiter (in this example) is 5.2 times farther away from the Sun than Earth. Not something easily gleaned from saying Jupiter is 780 million km from the Sun.

NEOWatcher
2008-Feb-07, 02:02 PM
Quite common in Solar System studies.
So; If I'm consolidating things in my mind correctly.
The units used are usually based on the magnitude of objects being described and are nearly equally "scientific".
AU for solar system objects.
Parsec for outside of the solar system for observational purposes.
LY for non-observational purposes.



(1 AU)/(1 parsec) = arctan(1 arcsecond) or pi/648000 approximately, or
1 parsec = 206,265 AU.
Yes; but is that the official designation?

If so, then redefining AU makes sense to lock it down, but from observational perspectives, it makes sense to keep it fluctuating so the angular measurements are consistant.

Click Ticker
2008-Feb-07, 03:40 PM
How different is AU from Light Year?

Isn't a light year dependent on how long it takes the earth to revolve around the sun? Doesn't that have meaning only in this one speck of the universe at this time?

NEOWatcher
2008-Feb-07, 03:48 PM
How different is AU from Light Year?

Isn't a light year dependent on how long it takes the earth to revolve around the sun? Doesn't that have meaning only in this one speck of the universe at this time?
I think that's Lite year... the one without Feb 29.

If we nitpick, yes it is, but the definition has been officially redefined to use QM properties to lock it into a more stable definition.

Hornblower
2008-Feb-07, 04:04 PM
Let's get back to the topic of the OP, which is the NewScientist article whose link appears in the OP.

In my opinion the writer and the scientists he quoted have not made a compelling argument for redefining the AU. The writer gave an awkwardly worded description of a definition he attributed to the IAU, but those words do not appear in the IAU document to which he linked. That document merely gives a number of meters to six significant figures, with no information about the line of thought that was used to justify the choice.

It does not matter for orbital calculation purposes whether the line of thought behind the choice of the unit of measure was good, bad or indifferent, provided the unit is a constant. If we try to calculate orbital motion with hairsplitting precision but ignore the Sun's mass loss, the results will diverge from the actual motion regardless of which unit of measure we plug into the equations of motion.

NEOWatcher
2008-Feb-07, 04:14 PM
...It does not matter for orbital calculation purposes whether the line of thought behind the choice of the unit of measure was good, bad or indifferent, provided the unit is a constant...
I interpret the article as saying it is not a constant, and shouldn't be used as one. I gather that from his mention of defining "k".

It sounds like he's stressing the fact that it's a ratio and not a constant.

Hornblower
2008-Feb-07, 05:55 PM
I interpret the article as saying it is not a constant, and shouldn't be used as one. I gather that from his mention of defining "k".

It sounds like he's stressing the fact that it's a ratio and not a constant.

I now interpret the article as being very poorly written. I don't know whether to laugh or cry.

Upon further Googling on the definition of the astronomical unit, I have concluded that the quantity 2(pi)/k indeed is a constant, specifically the Gaussian year. That is the value of the siderial year as calculated by Carl Friedrich Gauss from rigorous observation. It is somewhat off the present accepted value for reasons Gauss could not have anticipated, but for better or worse the IAU adopted it as a constant for the purpose of defining the astronomical unit. It is independent of present and future changes in the actual solar mass. It is commonly expressed in days, with the day defined as 86,400 SI seconds. That is close to the actual mean solar day, which is not constant.

The writer and the scientists he quoted did not give us any explicit mathematical demonstration of what they thought a misuse of what they thought might be a varying "constant" would do to a calculation. I think they simply had some blind spots, an occupational hazard for specialists in any profession, and saw a bogeyman where there was no real problem.

laurele
2008-Feb-07, 06:04 PM
"I now interpret the article as being very poorly written. I don't know whether to laugh or cry."

As a writer, I feel the IAU needs to pay far more attention to issues of wording and language when making or redetermining definitions. A definition that is poorly written, unclear, sloppy, etc. can do far more harm than good.

Also, if it's the IAU that continues as the body making such determinations, they need to restructure the way such decisions are made so deliberations and votes are inclusive of a wide variety of viewpoints. A repeat of the last incident in which 96 percent of its members were excluded from the vote because they were not personally present is not advisable.

parallaxicality
2008-Feb-07, 06:17 PM
Perhaps we need to develop new measurements based on Planck scales?

tdvance
2008-Feb-07, 06:25 PM
Perhaps we need to develop new measurements based on Planck scales?

but...but...that's universe-centric. It wouldn't hold in the universe in which Spock has a beard....

parallaxicality
2008-Feb-07, 06:41 PM
How intriguing. A mile is almost exactly 10^37 Planck lengths. How's that for metric?

Ara Pacis
2008-Feb-07, 11:12 PM
Actually, in the Solar System the AU is a much better choice than millions of kilometers. 5.2 AU means 5.2 times the semimajor axis of the Earth's orbit, it means Jupiter (in this example) is 5.2 times farther away from the Sun than Earth. Not something easily gleaned from saying Jupiter is 780 million km from the Sun.

For sure, division is such a pain.

Anthrage
2008-Feb-08, 03:29 AM
Perhaps we need to develop new measurements based on Planck scales?

I was wondering when someone would get around to addressing a point that I am surpised everyone else seems to have given a pass...that being the eventual, if not urgent need, for measurements that are not in some way dependant on the earth, the solar system or any other human-centric ideas.

I was thinking of a system which used something such as the Chandrasekhar Limit, or the Tolman-Oppenheimer-Volkoff limit, in conjunction with a universal constant - say the distance at which 2 masses with no relative velocity of either of those two limits no longer have gravitational influence over eachother; or something similar.

That is perhaps a poor example, but you get the idea. :)

Hornblower
2008-Feb-08, 01:13 PM
I was wondering when someone would get around to addressing a point that I am surpised everyone else seems to have given a pass...that being the eventual, if not urgent need, for measurements that are not in some way dependant on the earth, the solar system or any other human-centric ideas.

I was thinking of a system which used something such as the Chandrasekhar Limit, or the Tolman-Oppenheimer-Volkoff limit,1. How would you evaluate such quantities with any certainty?

2. What merit would they have for describing things on the scale of our planetary system?

in conjunction with a universal constant - say the distance at which 2 masses with no relative velocity of either of those two limits no longer have gravitational influence over eachother; or something similar.How do we know there is such a beast, let alone measure it with any certainty?


That is perhaps a poor example, but you get the idea. :)
No, I do not get the idea.

In principle it is straightforward to calculate units of measure based on familiar objects that are known to be stable and are measurable with great precision. The challenge is in getting a multitude of fallible human beings to agree on standard choices.

The strength of the Sun's gravitational field has been calculated with only a very small uncertainty by observation of orbiting objects, and the radius of an ideal one-Gaussian-year orbit has been calculated from that quantity. We need only to specify the epoch of that calculation to allow for the fact that the solar mass is gradually decreasing. I would choose the year 1976, when the IAU adopted the standard, take that calculated radius as a constant and stick with it for all calculations. As an alternative we could express the distances in meters and tolerate the awkwardly large numbers. That makes no fundamental difference in the accuracy of the calculations. It merely gets in the way of thought-process clarity in visualizing the system. Using a unit of length on the order of magnitude of Earth's orbit results in small numbers, and we can come close to doing Kepler's third law in our heads.

Argos
2008-Feb-08, 01:28 PM
NewScientist :'Astronomical unit' may need to be redefined

So, it seems the Parsec needs to be redefined too.

Halcyon Dayz
2008-Feb-08, 08:23 PM
Redefine it as 150,000,000,000 m.

Done.

Celestial Mechanic
2008-Feb-09, 04:56 AM
A preprint of the article may be found here (http://arxiv.org/abs/0801.3807). The main problem I have with the article is that it contains a lot of speculation about mass loss due to axions. If I recall correctly, axions are some sort of special solution to the field equations of QCD (quantum chromodynamics), the strong force. The author can only give a range of possible values for this as-yet-undetected source of solar mass loss. I'd prefer to wait for 1: detection of axions, and 2: accurate measurement of this loss mechanism before tossing our current definition of the AU.

Hornblower
2008-Feb-09, 02:39 PM
A preprint of the article may be found here (http://arxiv.org/abs/0801.3807). The main problem I have with the article is that it contains a lot of speculation about mass loss due to axions. If I recall correctly, axions are some sort of special solution to the field equations of QCD (quantum chromodynamics), the strong force. The author can only give a range of possible values for this as-yet-undetected source of solar mass loss. I'd prefer to wait for 1: detection of axions, and 2: accurate measurement of this loss mechanism before tossing our current definition of the AU.
I see no need to wait for anything. Dr. Noerdlinger's opinion concurs with mine. By using his first option, we can make the standard value for the AU truly constant by specifying the epoch at which it is calculated, and then agree to use that constant for all subsequent calculations of orbital elements. Alternatively we could do the calculations in meters. Then we can use sharpshooting radar or laser ranging to look for evidence of possible axion effects.

If the IAU policy makers cannot agree on such a trivial exercise in a timely manner, I cheerfully would wash out their mouths with a greasy sample of the soap I made in high school chemistry lab.

laurele
2008-Feb-09, 06:08 PM
Luckily a light year is approximately 10 to 16 th meters = ten million times a billion. It would get awkward for most of us to think about 7 trillion terra meters = 7 times ten to the 24 meters to a galaxy that is 700 million light years away. Neil

Isn't the entire concept of a light-year also earth centered? After all, it's based on our concept of a year, which is one revolution of the earth around the sun.

Jerry
2008-Feb-10, 12:55 AM
Luckily a light year is approximately 10 to 16 th meters = ten million times a billion. It would get awkward for most of us to think about 7 trillion terra meters = 7 times ten to the 24 meters to a galaxy that is 700 million light years away. Neil

Why awkward? That is what exponents are for - use them, they are handy. I Ackward is converting horsepower per acre stone to kilowatt joules per electron volt erg.