[vsnet-chat 1747] Re: BVRI Photometry

["William S G Walker" <astroman@voyager.co.nz>]

Arne, Dan & Taicho,

Answers to the latest questions. I like the multiple faint star method but
this still means determining BVRI values for a number of 'substandards' on
the image. We used to do this in the pm tube days but I was hoping to avoid
this. It's one avenue I could consider even though it means more reduction
time. I've rewritten my software to use three comparison stars, which is
often as many usefully bright other stars present on an image. I was
suspicious of the linearity of the CCD but everyone seems agreed that this
is good. We've always avoided faint comparisons in the past as they merely
add to the noise but I'll try this route as well. Perhaps two to five
fainter stars on the plate could be used as 'checks' initially, which would
make the reductions much faster, and then the variable adjusted as
necessary once reliable values are found for these stars. 

One limitation is that with the Meade 8 the fields are centred using the
finder only so the bright variable is always close to the centre.
Offsetting to include other stars is not as easy and results in many false
starts so I try to avoid this. But there are usually some other usable
images present.

Actually, Arne has answered Dan's question - if I'm measuring in two
filters for the Cepheids (B & V) mainly - why worry about the R & I? Let's
go back in time. In 1970 we were asked to measure eclipse of Z Cha. So we
dutifully measured eclipses in a single colour for 15 minutes then spent
the intervening 95 minutes measuring Miras in UBV. Thus the detection of
superhumps on this star (and to some extent SU UMa stars in general) was
delayed for 4 years! So let's invest a small bit of extra time and collect
information which may be useful in the future. I think that some of the
Cape people are or were measuring some cepheids in BVRI but not for the
same purpose.

In the past we've collected UBV light curves of Cepheids but only used the
B and V light curves as the U tends to be a bit noisier (lower signal
strength, sea level problems, etc.) It's easy enough to derive a time
correction for the B curve. So with twice the number of points the annual
epoch should be better. We use our own measures over some years to produce
a standard light curve then fit seasonal measures to this for the yearly
epochs. In a few cases there are adequate B-V measures in the literature
for a standard light curve but many of the early measures are poor and one
or two observers didn't standardise to the UBV system.

Szabados mentioned that the R & I curves are less reliable for timing. If
so, then we've (there are several of us who measure Cepheids in NZ) merely
produced some colour data for posterity. If they are reliable, then we can
use the R, or even the R & I, curves to produce more reliable epochs. The
response level of the R filter is very attractive and the B is down a bit
so it would be good to be able to use R instead of  or in addition to B.

Finally, if I'm going to do BVRI photometry, the more exposure I get - and
can pass on to my collaborators who are keeping up the measures of CVs
while I do this - the better. One other goal is to enable people like Bruce
Sumner to get reliable V sequences for CVs in the southern sky as I'm a
very keen addict? of CVs. I naively set out to do this a year ago but
didn't realise how complicated it was if you were not BVRI experienced.
Arne's doing a good job here but I don't think that he can get the far
southern objects. 

Thanks for the additional comments. They've been very helpful. It's
completely different from our early days when, in 1969, we bought a couple
of EMI 9502 tubes, the local university built and lent us part of the
equipment, the remainder we managed ourselves, and set off to learn about
UBV photometry. Only to be assured it was 'impossible at sea level'. I
think it took three years before we really felt confident about the whole
thing. So if there are any other useful ideas don't hesitate to toss them
at me. 

Regards to all,
Stan




----------
> From: aah@nofs.navy.mil
> To: vsnet-chat@kusastro.kyoto-u.ac.jp
> Subject: [vsnet-chat 1746] Re: BVRI Photometry
> Date: Wednesday, March 31, 1999 7:45 AM
> 
> Kato-san makes a good point:  using comparison stars that
> are really faint can lead to systematic errors if, for example,
> the sky background is not treated properly.  Like him,
> I recommend that you use brighter comparison stars if at
> all possible (and if you have lots of choice, I select
> comparison stars that are equal in brightness and color
> to the variable).  However, if you are in a situation like
> Stan, where there are no bright comparison stars in the
> CCD field of view, then I would much rather use the
> faint-star ensemble method for getting differential
> photometry rather than all-sky photometry with comparison
> stars in other areas of the sky, if you are
> trying to get high accuracy.
> 
> An interesting side-point:  since you would normally
> expose your frames so that the variable is close to
> saturation to give as much signal/noise as possible in
> the fainter comparisons, you actually get better results
> near the variable's minimum brightness, where the exposures
> are longer and therefore you have more signal/noise in
> the comparisons.  Likewise, when the variable is bright,
> you often have trouble adjusting the exposure with
> changing seeing such that the variable is not overexposed.
> This method is not for the weak-of-heart!
> 
> Dan asked if you could just observe in one filter since
> the object is to check for period changes.  As Brian
> mentioned, Cepheids do have different times-of-maxima
> depending on wavelength.  This is a small effect and
> is not important if you are using, say, a CCD with a V filter
> to compare with previous PEP V-filter data.  So the answer
> is 'yes': one filter, especially V, would be sufficient.
> However, observing with a second filter once you have spent the
> time moving to a field is a small penalty, and then you
> have sufficient information to report data on the
> standard Johnson-Cousins system.  Also, most of the bright
> Cepheids have little or no R&I data (they were observed back in
> the PEP UBV days), and getting the red-filter light curves
> can have some useful scientific benefits.  Stan has a filter
> wheel and standard BVRI filters, so I would recommend more
> than one filter for his project.
> 
> Arne

VSNET Home Page

Return to Daisaku Nogami

---

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