[vsnet-chat 3648] Re: Overobserving

[Chris Lloyd <cl@ast.star.rl.ac.uk>]

For what it's worth I do not believe that over observing by experienced
observers is a problem, nor do I think that it will lead to an increase in
noise in the mean light curve. Depending on how accurate the observations
are, they may show a step-like light curve or reveal possible
irregularities or nonlinearities in the comparison stars, neither of which
is necessarily wrong. With under-sampled data these sort of things pass
unnoticed all the time. I suppose there is a possible danger that the
observer may be fooled into thinking that the star has not changed
significantly when it has, but this is a question of experience, and the
same observer would 'probably' make less reliable, under sampled
observations. 

The light curves of Miras are not exactly repeatable and contain
subtleties that are poorly studied. Although LPV light curves may appear
to be well sampled when observations from many observers are put together,
in reality the details are lost because of personal bias. This is a
convenient hold-all for all the differences between observers mentioned in
a previous posting. Also when looking at high frequency components in these
light curves, folding the data over some interval is a poor substitute for
well sampled data. 

As John mentioned I have what I believe is a non-controversial routine for
merging visual observations from a number of sources. While this cleans up
the light curve significantly, by removing some of the systematic
differences between observers, it cannot put back the data that it's there,
because of undersampling. 

Regards,

Chris Lloyd

VSNET Home Page

Return to Daisaku Nogami

---

vsnet-adm@kusastro.kyoto-u.ac.jp