S Dor and RT Cru: BSM Projects?

RT Cru is a bit faint for quality BSM photometry.  I usually quote 3<V<10 for good results.  You can go as bright as V=2, but with lots of images per filter to reduce scintillation.  You can go as faint as V=12 with the current 100sec exposure limit, or fainter if you stack.  There are two main problems on the faint end.  First, the mounts are low-end and don't track well beyond about 80-100 seconds.  We usually recommend taking many such exposures and stacking with image tracking.  Second, the pixels are 4-5arcsec, so faint companions get merged into the blurry image.  An example of that is my 2017 SAS paper, "How Faint Can You Go":

http://www.socastrosci.org/Publications.html

page 75, where I show a 640-second stack of M67 with a BSM system, which in uncrowded areas, reached about V=16.8 as a faint limit, which means you could do decent photometry at about V=14.

What this means for your two targets is that RT Cru is marginal.  You could probably do V and Ic, but not B, with a BSM, and then only if the target was isolated (no companions within, say, 15arcsec of the target).  For this target, I'd recommend about 1000 total seconds for B, 300 seconds V, 150 seconds Rc and about 60 seconds Ic, as it is quite red.  This could be implemented as 10x100sec B, 5x60sec V, 5x30sec Rc, and 3x20sec Ic.  After initial images are taken, you can adjust exposure times and number of exposures to stack.

S Dor highlights one of the side benefits of the BSM network.  In the past, when I was running the network, I'd put targets of opportunity in the queue.  If they were interesting stars, they often stayed in the queue long after the original request.  For S Dor, there was an Alert Notice 453 in February, 2012, for supporting observations of this LBV for a multiwavelength campaign.  It was placed in the queue for BSM_S and BSM_Berry, with BVRI imagesets.  Both telescopes have been monitoring it ever since, with about 150 imagesets per year.  That data has not been analyzed, except that a few of the early year's data are available through the Epoch Photometry Database:

https://www.aavso.org/aavsonet-epoch-photometry-database

If you want to take ownership of this data, and at least for the future, submit the observations to the AID, I think that would be great!  An example of the images can be found at

http://images.aavsonet.aavso.org/bsm_berry/bsm_berry_190104_images/inde…

look at images 336-339.  Of course, the LMC makes a very pretty field!  The exposures here are 60/30/20/15 for B/V/R/I.

Arne

Hi Arne

I would be happy to “take ownership” of the S Dor data.

Presumably, before I can do any analysis, you will need to arrange for the calibrated images to be piped from BSM_S and BSM_Berry to my account (YPFA) on VPhot?

I’m familiar enough with VPhot, but a bit rusty on the overall process. I’m assuming that VPhot is used in the normal way to analyse the images and then VPhot’s output is pumped through George’s TA program before submitting to the AID?

Hi Paul,

You can analyze images using whatever software you like.  Many users prefer VPHOT because of its simplicity and diagnostic tools.  If you want to use it, I'll transfer a few nights of recent data to get started.

For those that might like to "adopt" other stars, I'll prepare a list of those that have been monitored, but that do not have an owner yet.

Arne

Thanks Arne

If you can send  a few nights of data to VPhot, that would ge good.

Hi Paul,

I've placed 4 January nights for S Dor into your VPHOT account.  Work with them, and see what you get.  You are now the "owner" of S Dor for future observations.

Arne

Hi Arne and Paul,

I am observing S Dor and RT Cru amongst many other stars since years in snapshot mode on a nightly basis in different filter bands.

You can find the data in the AAVSO database. If Paul would like to analyse the individual frames I can make them available via dropbox as there are many as you can imagine.

I am collaborating with professional groups on those stars and one paper on R71 came out recently.

https://arxiv.org/pdf/1709.00160.pdf

Josch

Hi Paul,

if you are only interested in doing phtotometry on images which have never been analysed photometrically then there is no need to do so on my data. However, if you intend to use the photometric data to get physical information about the stars then it is another pair of shoes.

In the end my contimuous coverage should end up in a scientific research paper as there was an alert some time (actually several years) ago by a professional astronomer.

I hope it was not just for fun :-(

Josch.

First, I want to congratulate Josch for such a nice long-term time series of S Dor at V and Ic!  This curve shows the value of quality CCD photometry - you see many low-amplitude bumps and wiggles, and can see the long-term trend of increasing brightness for this star, now reaching nearly its brightest values.  There are a handful of discrepant points that should be examined and deleted if incorrect, but the majority follow a nice curve.

As for Paul's questions:

1.  I’m familiar enough with VPhot, but a bit rusty on the overall process. I’m assuming that it would be OK to analyse these images with VPhot in the normal way, and then transform the VPhot output using TransformApplier, before submitting it  to the AAVSO? Is this correct?

Yes, this method works.  I think you can also transform single pairs of images in VPHOT, but since we're doing BVRI, TA makes more sense.

2.  If so, it seems to me that I should choose only one comparison star (as well as a check star), because the TA program, as far as I can tell, does not handle ensemble photometry. Do I have this correct and, if so, what is the tradeoff in choosing a single comp star, as opposed to an ensemble?

A single comp star is fine.  Each technique has its good and bad features.  Single comparison stars are easier to use and ensure that your data can be compared with others, as long as they know which comp star you used.  Ensemble techniques essentially provide a "master" comparison star formed by the average of many stars, and therefore can have higher signal/noise and can remove systematic issues, such as a nearby bad pixel impacting the photometry of a single comp star, or a comp star with incorrect standard magnitudes.  They are also more complex.  If it has been a while, I'd select the single comp for now and then consider going to the more complex ensemble method somewhere down the line.

3.  Should the comparison and check star (once chosen) be used for all 4 filters, or do the different filters call for different comparison stars?

If you are transforming, you need the same comp and check stars, at least with TA.

4.  Assuming that the same comparison star is used for all 4 filters, should an attempt be made to keep this same comp star for all of the many observation sets captured over an extended time period?

I think you can do this with S Dor, since its variability range is relatively small.  You pick a star that won't be saturated with the longest expected exposure.  For the BSM monitoring, we have used the same exposure times since the beginning, so that won't be a problem.  For some stars, such as Mira (LPV) variables, things get more complex.  First, the brightness as a function of wavelength can be extreme; from B to Ic, the star might have 7 magnitudes of change, and so you need different comp stars in the red than in the blue.  Likewise, from minimum to maximum, you might have 7 magnitudes of variation, and so you might need different comp stars at minimum than at maximum.  Miras are tough with CCDs!

5.  Finally, should the aperture, once chosen, be retained across all 4 filters? Should an attempt be made to keep the same aperture over an extended period.

It is essential that you use the same aperture for all stars in a given image.  It is not essential to retain that aperture for all filters and all time.  The only difficulty in switching apertures between images is if there is a nearby contaminating star that is included in the aperture used for one filter, but not for another filter.  Then there will be a systematic bias in the color of the star.  So if you can, you leave the aperture selection alone, but if there are reasons for switching the aperture size (say, poorer focus at Ic than at V, or poorer seeing on one night), then don't hesitate in doing so.

Arne

Arne

Thanks for your replies - all very helpful.

Following on from your responses to Q2 - Q4 above, I have one final question - I am inclined to choose the two stars 000-BJM-419 and 000-BLG-343 (both mag 9.6) for check and comparison stars. They seem to comply (more or less) with all of the criteria in the CCD Photometry Handbook (p.45): not too red and not too blue etc, and also lie roughly midway between the extemes of variation of S Dor. What do you think of the choice? (I would like to get this as "correct' as possible from the get-go).

And would one of these be better as a check star than the other?

Regards: Paul

Arne / Josch

Belatedly, I thought of looking at Josch's observations via the Light Curve Generator to see which stars he has been using - it looks as though he has chosen 000-BJM-421 as the check star and 000-BJM-420 as the comp star. Maybe this is a better choice than mine?

Perhaps Josch could provide some insight into why he made these choices?

Paul

S Dor itself has (B-V) approximately 0.2mag.  You would normally choose a comparison star that has similar color, so transformation errors are decreased.  Based on that, 000-BJM-419 (the 96 comp) would be a good choice.  In fact, most of the 9-10mag comps in this field are from Mati Morel's SNAPSHOT sequences, and so have been used by southern observers for quite a while.  000-BJM-420 (the 101 comp) has (B-V)=0.89, and so is a little redder than I'd like for the target.  However, you also have to look at distance from the target.  101 is much closer, and may be why Josch chose it (I don't think he has the 3 degree field of a BSM).

If the 96 is not saturated in the BSM images, I'd probably choose it, with the 101 a chose second.  In fact, maybe use the 101 as the check star.  Then researchers can re-analyze your results with either star.

I didn't see 000-BLG-343 on the chart.  You might give me the chart ID that you used, so that I can pull up an identical chart and find it!

Arne

Arne

Looking at a fuller Photometry Table (X23939KK) and associated Chart X23939KJ, it appears that none of the candidate comparison stars that we have recently discussed have values for R_c and I_c.

Does this mean that the values don't exist at all or could it be that the Sequence Team just has not entered them. Can the Team provide them if asked??

If they cannot be provided then, looking down the Photometry Table, I don't see values for R_c and I_c until we get to BMP-588 (104). After that, they become a bit more frequent. However, none of the possible comp stars are close to the target, and they are increasingly fainter thanb S Dor's variation range.

What to do here?

Paul

Hi Paul,

You were using a 2-degree chart, which is bigger than the one I was using.  The second 96 star probably is on the edge of that chart, which is why it appears in the table but not the chart.

As for Rc/Ic: I've calibrated that field and have Rc/Ic magnitudes for all of the comp stars.  I'll see if I can't update the database for the Sequence Team so that they can fill in the values.

Arne

Hi Paul,

Tim Crawford of the Sequence Team was nice enough to update the magnitudes for most of the 9-11mag stars in the S Dor field.

Choosing Josch's stars is fine.  Check the signal/noise of those stars when analyzing the images, and make sure that the S/N is above 100.  If not, then choose one of the brighter stars, such as the 96 or 98 entries, as your comp star, but continue to use one of Josch's stars for the check star so that a researcher has continuity.  Good luck!

Arne