From msdemlei at ari.uni-heidelberg.de  Tue Jun 10 11:01:23 2025
From: msdemlei at ari.uni-heidelberg.de (Markus Demleitner)
Date: Tue, 10 Jun 2025 11:01:23 +0200
Subject: Logarithmic abundances in VOUnits
In-Reply-To: <3CD03105-24B5-48AB-AE3C-3A8855022F28@googlemail.com>
References: <3CD03105-24B5-48AB-AE3C-3A8855022F28@googlemail.com>
Message-ID: <20250610090123.n4epyql5crhaxzyu@victor>

Dear Alberto,

On Thu, May 22, 2025 at 02:58:22PM +0200, alberto micol via semantics wrote:
> I have a user that is asking which units should be used for logarithmic abundances relative to the sun;
> He is suggesting:
>
> 	log(Sun)
>
> At first I was surprise, but reading the VOUnit standard (*) I
> think he is right.  Do you confirm?

Yes, to my own surprise it's there.  And, well, I have to say it *is*
noticeably better than the "dex" we're trying hard to avoid.

Very frankly, I have either missed it before or forgot about it
myself.  We've had various discussions about the abundances in the
meantime, for instance here:

https://github.com/ivoa-std/VOUnits/issues/37

Given what's in VOUnit now, I think the slightly ugly log(Sun) is
what we ought to recommend instead of dex if that's what you've been
doing.  And with that decide that we won't use the opportunity to
tell mass from particle number; log(kg/kg) vs. log(ct/ct) feels a bit
brittle anyway, and I wouldn't want to quarrel with people insisting
that the former should be written as log(g/g).

Thanks,

          Markus