[Aavso-photometry] GJ 436 update and upcoming transit times

Tue Jun 19 10:34:45 EDT 2007

I have checked the flatfield using NGC7790 and the method described by
Arne. And sure enough, I found systematics in the data. It looks like
there is a gradient from the center towards the edges. There is some
vignetting in this telescope, and it doesn't look like the flat
corrects it completely.

After applying corrections for both position in the frame and color, I
get the following:

Comp Avg. GJ436 V-mag
_106 10.73
_107 10.70
_116 10.69
_114 10.72
_127a 10.72

Avg. = 10.71, Std. = 0.016, compared to 10.62 and 0.053 uncorrected.
The baseline V-mag (not including the dip in the light curve) is at
10.706 , compared to Arne's 10.702.

I should also mention that ensemble photometry improved the light
curve compared to the one I posted in my report (I will post a new
revision shortly). By eyeball measure I get a transit duration of
about an hour, and a depth of about 0.08 mag.

Analysing this data took a lot of work, but it has been very
educational. Thanks for your help.

I have submitted the data to the AAVO database.

Geir Klingenberg

On 6/16/07, Arne Henden <arne at aavso.org> wrote:
> In my mind, this is one of the biggest problems with service-run
> robotic telescopes.  You cannot either get sufficient time, nor pay
> for sufficient time, to fully characterize your system.  You have to
> rely on the provider of the service.
>
> Proper flatfielding is a "must", and when I see differential photometry
> numbers like what Geir indicates, I get suspicious of flatfielding.  Many
> amateur telescopes suffer vignetting, and scattered light mimics
> vignetting, so just looking at a flat does not tell you much, and looking
> at a processed image looks nice and flat even in the presence of
> scattered light.  The photometry suffers greatly, however.
>
> I've told the list before how to check for scattered light problems,
> both internal checking of the telescope as well as on-sky tests
> with large calibrated regions like M67 and NGC7790 or raster
> scans of your own regions.  These tests really need to be performed
> before you do much serious photometry.
>
> As also mentioned earlier, you can work even in the presence of scattered
> light and poor flats if you accurately reposition your field on each
> visit/image,
> and apply a zeropoint offset to bring your photometry into line with others
> (correcting the improper flat).  You have to take sufficient data to compare
> with others, and overlap with them, but it can be done.  Much better to
> get rid of any such systematic problems at their source.
>
> This holds for almost any robotic telescope service, including some professional
> ones.  We've had to recently do some heavy image processing to create
> master sky flats for a telescope that could only supply master flats taken
> 6 months ago.  Even then, we now have to look closely at the data and see
> if any scattered light issues remain.  Even going to a national facility, you
> either have to rely on the site to have characterized the telescope, or
> burn precious time of your own to double-check. It is *so* much easier to
> have your own telescope and have it fully characterized!
>
> It may have been easier to get good transformation coefficients in the GJ436
> region because the target star is so red, giving you a wide color range for
> the determination.  Using Gj436 itself is not necessarily bad, as it appears
> to be constant and even the transits only dim it by 0.6percent.  However, this
> is a "secondary standards" calibration of the field, and you should generally
> obtain your coefficients from Landolt standards if possible to guard against
> any systematic effects.   For most differntial use, getting precise coefficients
> that do not contain systematics is probably not that important; I just do lots
> of all-sky calibration work where others rely on my photometry, and so have
> to be more careful than the typical researcher.
> Arne
>
> On 6/16/07, Geir Klingenberg <geir.klingenberg at gmail.com> wrote:
> > Sorry for not posting the data yet.
> >
> > The report was made before the campaign was announced. Following the
> > suggestions in the announcement I have used ensemble techniques on the
> > data, and I have tried to transform. But I can't seem to get accurate
> > V-mag estimates for GJ 436, and I am running out of ideas as to what
> > might be the problem. Except the obvious, of course, which is me :-)
> >
> > >From Alert Notice 350 GJ436 has a baseline V-mag of 10.702.
> > Untransformed, I get the following:
> >
> > Comp  GJ436 V-mag
> > _106    10.64
> > _107    10.53
> > _114    10.64
> > _116    10.53
> > _127a   10.56
> >
> > Average 10.58, std 0.056. That is about 0.12 mags away, which seems
> > pretty much to me. These results are consistent over three different
> > nights.
> >
> > Transforming does not help much either. With the transformation
> > coefficients that I have established on several occasions - using both
> > NGC7790, M67 and some Landolt fields - the difference compared with
> > the untransformed ones is negligible.
> >
> > You find info about the transformation coefficients here:
> >
> > http://www.remote-astronomical-society.org/Tables%20and%20Files/Transformation%20Coefficients%20for%20ARE01.pdf
> >
> > (I used a method based on differential photometry, but I have checked
> > that the method in the Henden book gives similar results). I use the
> > following equation to calculate the transformed V-mag, using the known
> > B-V of GJ 436:
> >
> > Vt = Vc + (vt – vc) + Tv * ( (Bt - Vt) – (Bc - Vc) )
> >
> > I also got some B and V frames a couple of days ago to find the color
> > myself, without any significant change in the results.
> >
> > Finally I used these B and V images to obtain transformation
> > coefficients from the stars in the GJ436 field, including GJ436
> > itself, and that gave me the following coefficients: Tbv = 0.957, Tv =
> > 0.09. That helped and gives an avg. V mag of 10.65. But using GJ436 in
> > the solution is kind of cheating, I guess.
> >
> > I have noted earlier that I get the best result by finding
> > transformation coefficients using the comp stars in the field of the
> > variable. Is this common?
> >
> > Also, could it be a flat issue? We can not make our own flats for this
> > robotic telescope, instead we use the one that is supplied. It is not
> > of highest quality when it comes to photometry, but I didn't think it
> > was *that* bad.
> >
> > If anyone want to have a look at some of the images I will be glad to
> > send them to you.
> >
> > Please let me know if you see any calculation errors or similar.
> >
> > Best regards,
> > Geir Klingenberg
> >
> > On 6/15/07, Aaron Price <aaronp at aavso.org> wrote:
> > >  We have our first set of data, reported by Bill Goff (GFB). I've placed a light curve
> > > of his data on the web page: http://www.aavso.org/news/gj436.shtml .  There seems to
> > > be some disagreement about Geir's reported transit on May 31. The data has not been
> > > submitted to the AAVSO, but his report is also available via that web page.
> > >
> > >  Here are some predicted transits (UT) for the next week. As you can see, they are
> > > fast (62 min.)!
> > >
> > >  Begin                                    End
> > >  JD               Mon. Day  Hour Min
> > >  2454267.54 2007  6    16    0   53      2454267.59 2007  6 16  2  3
> > >  2454270.18 2007  6    18    16  20      2454270.23 2007  6 18 17 30
> > >  2454272.82 2007  6    21    7   47      2454272.87 2007  6 21  8 57
> > >  2454275.47 2007  6    23    23  14      2454275.52 2007  6 24  0 24
> > >  2454278.11 2007  6    26    14  41      2454278.16 2007  6 26 15 52
> > >
> > > >From http://www.ucolick.org/%7elaugh/GJ436____b.transits.txt
> > >
> > >  But remember we need data all the time, not just during transit windows.
> > >
> > >
> > > Aaron
> > > _______________________________________________
> > >
> > > Aavso-photometry mailing list
> > > Aavso-photometry at mira.aavso.org
> > > http://www.aavso.org/mailman/listinfo/aavso-photometry
> > >
> >
> > _______________________________________________
> >
> > Aavso-photometry mailing list
> > Aavso-photometry at mira.aavso.org
> > http://www.aavso.org/mailman/listinfo/aavso-photometry
> >
>