Talk:Astronomical unit

The article mentions the improved measurement of the speed of light in 1983. The way I know it, is that the ability to measure time, or frequency, improved, but the ability to accurately measure distance did not. So, defining c increases the ability to measure distance. But also, defining c means that it is now measured infinitely accurately, by definition. Or, the other way to look at it, you can't measure the speed of light anymore. Gah4 (talk) 09:51, 14 April 2025 (UTC)

I mean in SI the defintion of a metre, the base unit of distance is proportional to the distance light covers in 1 / c seconds and the transitions of a cesium atom state for a certain amount of oscillations (i.e., a second).

So basically the more accurate you measure the speed of light and the second the better your definition of a metre (i.e., distance) Quintofin (talk) 18:59, 18 June 2025 (UTC)

The speed of light is defined to be exactly 299,792,458 metres per second. No measurements are needed. Johnjbarton (talk) 19:45, 18 June 2025 (UTC)

Well yes, but that would be a circular definition.

The speed of light is measured and then the metre is defined to be the speed of light, which then if you rearrange it defines the speed of light again.

The metre is 1 / c (times the cesium ground state oscillations of a certain number making up a second). Then, if you measure the speed of light more accurately, you define the metre more accuratly.

https://en.wikipedia.org/wiki/Speed_of_light#Measurement Quintofin (talk) 20:49, 18 June 2025 (UTC)

The IAU recently restructured its website, and PDF files of some publications are now made available on Google Drive, for example some resolutions referenced in the article. The original links to the files were removed.
Here is the link to the current page listing resolutions:
https://www.iau.org/Iau/Publications/List-of-Resolutions
The direct link to Resolution B2 adopted at IAU General Assembly 2012 is here:
https://drive.google.com/file/d/1FBKjaswlP-rTs6z6V4WygXsbdHKVHYcP/view
The direct link cannot be included in the article because it will receive a tag "citing a blog or free web host" indicating a source that is not reliable. The workaround that can be used for now is to cite a link to an archived version of the publication instead.
See also https://www.iau.org/Iau/Iau/About/Archive.aspx?hkey=8e870620-d676-447e-b436-805189f20a99
Uxh (talk) 10:09, 20 April 2025 (UTC)

A discussion at Template talk:Val#Spacing around ± mentions an example of how {{val}} is used in this article. Anyone interested in what spacing occurs before/after ± might like to join in there. Johnuniq (talk) 02:59, 24 May 2025 (UTC)

The measurement of the distance to the Sun by Eratosthenes is sourced to an ancient Greek author. This measurement is not mentioned in the Hughes history. I do not consider this source to be reliable. Johnjbarton (talk) 00:41, 25 July 2025 (UTC)

The following Wikimedia Commons file used on this page or its Wikidata item has been nominated for deletion:

• Astronomical unit SVG.svg

Participate in the deletion discussion at the nomination page. —Community Tech bot (talk) 03:23, 14 October 2025 (UTC)

Any discussion of the recent change in SIGFIGs? I think I agree with it, but it would be nice if it was discussed. Gah4 (talk) 11:26, 5 February 2026 (UTC)

Are you referring to my recent revert of an anonymous edit? Johnjbarton (talk) 17:14, 5 February 2026 (UTC)

Yes. One is that I don't understand what {{convert}} does with sigfig=100, but it doesn't seem to give 100. In general, I am against excessive sigfigs, though for quantities that do have useful precision, I like to see it. Gah4 (talk) 09:49, 6 February 2026 (UTC)

Convert changes large sigfig values to be no more than 14 which experiment a long time ago showed matched what {{#expr}} does. Johnuniq (talk) 10:03, 6 February 2026 (UTC)

The standard is exact so in principle one could use infinite significant figures. The way one accomplishes this in the 21st century is to use the exact value in computer programs with appropriate precision arithmetic. No one doing precise work will type in a converted value with many significant figures.

In spirit the exact value reflects experimental results at 12 or maybe 13 significant figures. (It meaning and thus accuracy are not close to that I think). So I think 12 is a good value for the list. Alternatively we can delete the conversion table absent a source. Johnjbarton (talk) 17:35, 6 February 2026 (UTC)

And even more, there is no WP requirement that sigfigs reported agree with the source. But yes, users should compute the appropriate approximation needed, in unit systems that don't have a nice conversion. Gah4 (talk) 23:55, 6 February 2026 (UTC)

Experts writing computer programs will enter the exact value, or to as many significant figures as the precision of the chosen variable type will hold. If the value is represented as a floating point, there will be some more imprecision when the decimal value is converted by the computer to binary. If stored as an integer, the decimal to binary conversion will not affect the precision. Jc3s5h (talk) 17:41, 7 February 2026 (UTC)