An interesting puzzle

[Marcus D. Leech]

I've attached the plot of one year of our H1 Doppler data, along with an
adjusted VLSR correction curve, and the corrected data.

In addition, I've attached the data file that produced this plot.

So, the puzzle?  This is just over 1 years worth of data.  The curve
doesn't *quite* return to whence it started after 366 days.  I can't
quite figure this out.

If it were measurement errors, then the phase of the (computed, by
model) VLSR correction curve, and the measured data would drift
significantly.  They don''t.

Earth orbit, of course, is not QUITE circular, but the results should
still be strictly *periodic* with a period of 365.25 days.  Granted the
solar system is moving as well,
  at 20km/sec or so.  But I don't think this will change much of
anything of the course of a year--after all we're looking at sources
that are eye-watering distances away.

So, any ideas?

Cheers
Marcus

[fasleitung3]

I have plotted the data in such a way that the critical part can be
seen more closely. This by plotting it twice, one original and the
other by shifting it by 365.25 days. Then the ends seem to meet within
measurement accuracy.
There is, however, an overall deviation between calculated shift and
actual shift with an oszillatory behaviour over the year. Which VLSR
model have you been using?
Best regards,
Wolfgang

> --

[Marcus D. Leech]

On 2026-02-21 16:47, 'fasleitung3' via Society of Amateur Radio

Astronomers wrote:
> I have plotted the data in such a way that the critical part can be
> seen more closely. This by plotting it twice, one original and the
> other by shifting it by 365.25 days. Then the ends seem to meet within
> measurement accuracy.
> There is, however, an overall deviation between calculated shift and
> actual shift with an oszillatory behaviour over the year. Which VLSR
> model have you been using?
> Best regards,
> Wolfgang
>

Indeed,  I think that the oscillatory behavior of the corrected line is
due to the VLSR model being computed for the center of our FOV,
  but the spectral component we're measuring isn't necessarily right at
the center of the field.   The model is from AstroPy.

[Marcus D. Leech]

On 2026-02-21 16:47, 'fasleitung3' via Society of Amateur Radio
Astronomers wrote:

> I have plotted the data in such a way that the critical part can be
> seen more closely. This by plotting it twice, one original and the
> other by shifting it by 365.25 days. Then the ends seem to meet within
> measurement accuracy.
> There is, however, an overall deviation between calculated shift and
> actual shift with an oszillatory behaviour over the year. Which VLSR
> model have you been using?
> Best regards,
> Wolfgang
>

Thanks.   Once i looked at the "plotting against itself" result, it
became obvious.  The quasi-SIN wave is effectively
  drifting through your sampling window but isn't necessarily
phase-aligned with it.