[vsnet-chat 3641] Re: Overobserving

["Stan Walker" <astroman@voyager.co.nz>]

Greetings, All,

The present discussion which presents a large number of viewpoints is
interesting. I've been analysing visual observations of 80 Miras for the
RASNZ VSS and a rather small number of SR stars as well. So let's look at it
from this viewpoint.

Initially the stars can be divided into two categories - those with a single
period which comprise Miras in the 200+ day range and those with multiple
periods, Miras with periods of under 200 days, semi-regulars and some other
variables called LPVs. It's difficult to overobserve these latter stars.

These days analyses are invariably done with computers and sheer volume of
data is not a problem. Periods are best determined by fitting most of the
light curve - and in these cases the observations near maxima and minima
carry low weight. They do indicate amplitude changes but these need to be
large to be believable. But there are many well documented cases of such
changes.

I would differ from Jan Hers and suggest that the perfect observational
sample is where one observer has observed a Mira star several times a night
for every night of the year. The dataset will be homogeneous and the
inherent inaccuracies of visual observing will be reduced. There will also
be enough measures, 50-100 in the best part of the cycle, to allow a good
epoch to be found. But subjective effects come in as the observer will be
influenced by his remembered earlier observations.

L2 Puppis is a semi-regular star which illustrates the problems nicely. The
main period is about 140 days but there is a strong secondary period of ~100
days. There are some others as well. If you analyse any one observer's
measures (DFT, FFT or whatever) these two periodicities are clearly visible.
But if the analysis includes measures from 10 observers the secondary period
becomes difficult to identify. Once the measures are incorporated into the
old 10 day means presentation the secondary period is no longer seen.

I note Taichi Kato's comments about Mira itself. The VSOLJ is not the only
organisation where this star is measured by hundreds of observers. The
resulting light curve is almost unusable. So it's a simple task to pick the
six most prolific observers and throw out the rest. This shows a bias toward
people who observe more frequently which rewards those who stick to a small
group of stars and do the job effectively. I've always thought that the VS
directors should make more efective use of their observers by encouraging a
better spread of measures - say by allocating stars to people. Not done? But
with CCD measures of CVs the NZ group cooperate to ensure there's no
unnecessary duplication of effort.

A reasonable proportion of Miras have companions which may interact in some
manner. Mira B does this but it's probably hard to see much visually. And
there must be dozens of others with WD companions which are unrecognised.
Many will have unevolved companions. The Mira stage of variability seems to
be only the most stable of the various red star light variations. Even here,
though, there are some interesting features which could be resolved if the
visual measures were reliable enough to determine the epochs of maximum to a
day or two. They are not at the moment good enough to do this and won't be
if we rely on measures from many people. But three or four good observers
per star making measures at interval of 2-3 days could produce a much more
useful dataset.

With all the less stable red variables with reasonable amplitudes the more
measures the better. We've tried UBV measures to get data on stars such as
L2 Puppis and other SR stars but the task is daunting. So until a group can
assemble CCD measures of a large number of these stars every night the
visual measures are the best that's available.

In reality it's the CVs that are overobserved by visual
observers. Not during quiescence but during outbursts. The moment a star is
bright everone and his dog is measuring it six times a night. There are
well-known features during outbursts but it's useless to try to document
these visually. So this just becomes 'astronomy for entertainment'. This
wasted time would be much better spent on making a few Mira, SR and LPV
measures. Insofar as the CVs are concerned the visual observers are a superb
'early warning' group which documents the type and frequency of outbursts
(and allows the opportune use of pep and CCD devices) but can do little
more.

My interest in Miras in this particular case was searching for possible
period changes. There was little indication of that but there is no doubt
that they alternate between closely separated (1-2%) two main periods at
intervals. The stars where this could be seen best were those where there is
a long history of observing. About 100 Miras south of -30 latitude come into
that category. But even here there were large gaps in the data and it is
quite clear that many stars are not well enough observed.

One wonders at the wisdom of publishing hundreds of charts of Mira stars,
some very faint, and then just waiting to see what is observed. Unless
enough measures are made then observations of the least favoured stars are
just random data and are completely wasted. Probably 50% of the Mira
measures come into this category which is quite disappointing.

Regards,
Stan


----- Original Message -----
From: Peter Kroll <pk@stw.tu-ilmenau.de>
To: VSNET Chat <vsnet-chat@kusastro.kyoto-u.ac.jp>
Sent: Friday, September 29, 2000 8:34 AM
Subject: [vsnet-chat 3640] Overobserving



>

VSNET Home Page

Return to Daisaku Nogami

---

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