[vsnet-chat 4315] Re: Omega CMa comparison stars

[Thom Gandet <tgandet@mindspring.com>]

Sebastian,

Sebastian Otero wrote:
> 3 Pup:
> 
>    HR 2922 = HD60863    7h35m23s  -28º22'10" (2000)
>                         V=4.639  B-V=-0.102  (from Tycho-1)  B8V
> 
>                     One of the main problems in visual estimation is
> the choice of stars of very different brightness to make the
> estimate. In fact, HD 60863 (p Puppis) is very faint compared with
> omega CMa.

    The nominal magnitude for HR 2922 is V=4.63; the nominal
magnitude for kappa CMa is V=3.96 (with an historical range of 3.78-
3.97 V).  Many reliable variable star visual sequences have larger
magnitude differences than that.  11 Pup is fainter than HR 2922,
by about 0.2 mag; 11 Pup is 50% farther from kappa CMa than either
k Pup or 3 Pup, and, thanks to Taicho's digging, we now know it
to be variable.

    11 Pup's B-V is about +0.7, and delta Col's B-V is about +0.8.
Both  are too red and too far from Omega CMa to use as a
comparison star, especially when there are other stars similar in
color that are closer.
 
> ** ... k Puppis doesn't have a Flamsteed number as the others
> (e Puppis = 11 Pup; l Puppis = 3 Pup) It is a very nice pair of
> nearly equal stars. They are HD 61555 and HD 61556.

    Thanks for the HD number.  How did you find it?  It's interesting
that SIMBAD doesn't recognize either "k Pup" or "k01 Pup"; "k01 Pup"
is listed as one of the SIMBAD identifiers for HD 61555!  Very
maddening.  Were the Roman letter names assigned by Herschel,
I wonder?)

> All of these stars MUST SURELY BE EXCLUDED from any photometric
> study of the star but are usable for visual purposes.
> 
> I insist: all we can win using perfectly constant stars, we lose
> it with brightness difference or using a star that it is too far
> away from the field.

    I agree.  But I submit that differential extinction becomes
important when the zenith distance between the two stars becomes
greater than about 15º at hour angles around 3h 30m (see posting re
Theta CrB).  Color difference effects may also contribute to
systematic and random errors.  If visual estimates are to be made
more precisely, then these factors have to be taken into account
in reductions and comparison stars very close to the variable
should be used.  I think that much more precise visual photometry
is possible, as your observations demonstrated.

    The Epoch Photometry for HD 61555 shows a total range of
4.02 to 3.73, or 0.29 magnitude.  The faintest magnitude measured
is near the end of the mission, and nearly as many "bad" 
measurements occur at "maximum" as at "minimum".  The Hipparcos
photometry should not be taken as an absolute indication of
variability or non-variability.  There are at least two examples
I can cite where the Hipparcos magnitude is wildly different from
the Bright Star value (and from other, more recent, PEP).

I wrote:
> >>> ...an Ae/Be star may have...Small "outbursts"...and may not be
> noticed; e.g., if the comparison brightens by 0.08 magnitude and 
> [is] not noticed as such, the variable will be reported as having
> faded by 0.08 mag.  Ae/Be stars should not be used as comparison
> stars especially as comparison stars for other Ae/Be stars.  (Sort
> of like using one Mira variable as a comparison star for another
> Mira.)
> 
> ** As expressed before it is a great problem for PEP. Visually
> we can deal with that.

   All of these type stars, including supergiants like 3 Pup
(as Taichi and Petr Hermanec have also pointed out here), are
> subject to outburst.  The size of those outbursts are not
> predictable.  If these stars are to be used as comparison stars,
then adequate check stars must be used to ensure accurate and
precise estimates.  I think visual estimation can be done more
accurately than 0.08 magnitude - as you have done - but all
possible sources of systematic and random errors have to be
eliminated or minimized as much as possible can.

> It is important to say that one should be careful in using them,
> but I think care must be taken in every estimate we make (see
> vsnet- chat 4199 and 4233).  Every time I estimate a variable,
> it's like estimating its comparisons since I take the job of
> checking if everything is "okay".

   Yes, I agree that's the best way to do visual photometry.  I
follow the same practice of estimating all the comparison stars
with respect to each other.  This is the practice is photoelectric
photometry (and I include CCD photometry there), and visual
sequences are generally designed to make that possible.

> BUT I know that this may lead to a confussion and that people
> are not obligged to check everything every time they observe,
> so generally speaking it is important to support the notion of
> using non-suspected stars for comparisons.  But there are cases
> when we lose accuracy when we are actually trying to achieve it !!

   Perhaps we ought to agree to disagree about this.  I think
accuracy and precision *are both* lost by using suspected variables
as comparison stars.  The observations are certainly open to some
doubt if one or more of the comparison stars may have undergone a
flare, or have flickered, or some other short-lived change
occurred that may be potentially large but go unnoticed; the
problem is, you don't KNOW what has happened.  I think the possibly
is great enough to try to eliminate it completely by using only
known non-variables.

   As Petr Hermanec pointed, variability in R of about 0.5
magnitude should disqualify the star from use as a comparison
star alone, even if no variability >0.05 magnitude in V is
known.  The variability in R indicates the likelihood of
an outburst (or, in other types of stars, fading) in V at some
- unpredictable -  time.

Cordially,
Thom Gandet


-- 
************************************************************************
                      Lizard Hollow Observatory
                      Thomas L. Gandet, Director
                            PO Box 77021                   
                      Tucson, AZ 85703-7021  USA
************************************************************************

VSNET Home Page

Return to Daisaku Nogami

---

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