[daip] VCAL in SETJY

[Eric Greisen]

On 07/08/2016 10:51 AM, Andy Biggs wrote:
> Hi.
>
> I have VLBI observations of an HI absorber that was initially discovered
> with the GBT. If I compare the two spectra, they look very similar, but
> they appear to be offset in velocity by ~300m/s. The attachment shows
> the spectra as well as their difference.
>
> Now, the vertical scaling is arbitrary as there's a bright source of
> unknown flux density in the GBT beam that's diluting the apparent
> optical depth. The way I've scaled the spectra is as if there's an
> additional component missed by the VLBI array (certainly possible) but a
> velocity offset remains a possibility. I therefore want to be sure that
> the velocity scales are correct. I'm talking to GBT people, but I'd also
> like to check that what I did in AIPS with the VLBI data is correct,
> especially as I'm not an experienced spectral-line VLBI observer.
>
> Basically, I used SETJY with its VCAL option to calculate the velocity
> at the reference pixel and then ran CVEL. That's it really - seems very
> straightforward. The reference pixel is also close to the line so there
> should be no issues with the velocity increment being a function of
> frequency in the optical definition. This all sounds fine?
>
> What I've also done is to try and calculate the geocentric to
> barycentric velocity conversion using some IDL code from the astrolib
> library and compare this to the AIPS value. The latter isn't reported,
> but I can calculate it as I know what the velocity at the reference
> pixel is in the geocentric frame.
>
> In the IDL code, there's an input that I don't entirely understand
> ("input epoch of mean equinox of dvelh and dvelb") where dvel are the
> heliocentric and barycentric velocities. I assume though that it should
> be set to the date of the observation and doing that causes the AIPS and
> IDL values to differ by only 5m/s which is of course excellent.
>
> So, AIPS seems to be doing the right thing, but it would be great if you
> could reassure me that what I'm doing sounds okay.


I would guess that the GBT code uses the radio velocity convention 
rather than the optical.  This makes a surprising difference even at 
smallish velocities.  What you are doing sounds fine.

Eric Greisen