Resently I received some images of R Mon in NGC2261 since ASTAP could not measure this star. The star is now measurable in ASTAP with a special option. But I noticed different programs measure a different magnitude. Using AstroimageJ I get the following R Mon magnitudes using 000-BBL-632 as comparison star:
AstroImageJ, Aperture 3 px, R Mon magnitude 11.235
AstroImageJ, Aperture 4 px, R Mon magnitude 11.083
AstroImageJ, Aperture 5 px, R Mon magnitude 10.931.
ASTAP, aperture 1.6 x HFD (1.6 x 4.2px), magnitude 11.147
ASTAP, aperture 1.8 x HFD (1.8 x 4.2px), magnitude 11.106
ASTAP, aperture 2.0 x HFD (1.8 x 4.2px), magnitude 11.043
I did also some observations of RMon and noticed a significant magnitude difference compared with other observers. The measured magnitude seems very dependent on the selected aperture radius. I assume this is caused by the flat star profile of R Mon
What would be the correct approach (aperture) for measuring such a star inside nebulosity? See attached cropped FITS image and AstroImageJ screenshot.
Since the nebula is resolved and the nebula is variable in brightness over time it is a very difficult object to observe with typical synthetic aperture photometry. Basically because the inner ring depending on size gets different amounts of light from both the star and nebula. The âskyâ ring also gets contaminated by the nebula. Different seeing conditions will spread the light differently from hour to hour, night to night.
PSF fitting is a better choice than synthetic aperture photometry when the star is involved in nebulosity or say a SN in a resolved galaxy. Try MAOPhot which is being developed by user petefleurant here on the AAVSO Forum and last time I used it worked quite well.
Even the peak pixel of R Mon contains nebular flux added on top. The PSF fitter will faithfully fit the shape, but the amplitude it solves for is inflated because the baseline itâs fitting against already contains nebular light. PSF fitting solves the shape confusion problem but not the flux contamination problem at the source position itself.
Probably a very sharp image would help the find the peak value of R Mon. For aperture photometry a very small aperture will then help.
You could also assume that total flux of NGC2261 nebula represents the magnitude of R Mon.
This sort of problem has long been an issue, for instance, in getting consistent photometry of active galactic nuclei (Seyferts, blazars, quasars). About the only thing that works (and it doesnât really, except in favorable cases) is to use a consistent measuring aperture always, and take enough data over many years so that you can look at the systematics. That measuring aperture is going to be somewhat larger than you might like, since you have to accommodate nights with soft seeing etc. Thus changes will be muted by the strong underlying nebulosity, but lacking a 10-m telescope with always-perfect seeing (in space) you simply do the best you can.
Thereâs plenty of work in the literature, for instance, on the âoriginalâ blazar BL Lac, to see how folks have dealt with this sort of thing. In that case, it is recommended to use an aperture of about 15" diameter, and folks looking at the data are able to subtract out the contribution of the underlying galaxy, which is not variable. The problem with R Mon is that the nebula is variable, so things are not straightforward.
A very large aperture like 15" fixed in number of pixels.
At focus, the stars in my images have a HFD or FWHM of about 4". So your suggesting an aperture of 3 or 4 times the HFD/FWHM value that will measure at least all star flux independent of the seeing? It also means you have to fix the aperture in pixels and not express it in HFD or FWHM values since it should be independent of the seeing.
Iâm suggesting not using some factor larger than fwhm, but instead some fixed value in arcseconds always no matter what the seeing is (or what instrument is involved). Some trials on multiple nights will probably be required to decide on this, and donât think you wonât have to re-reduce the data at some point as better methods come along.
The fundamental problem is that the seeing and focus has more influence on the comparison star then on R Mon. R Mon is like a cloud and the comparison star is more like a point source.
R Mon inside the NGC2261 cloud is a like Chinese lantern. The R Mon star light is spread across a much larger area then the comparison star(s). So seeing or focus has not much influence on the surface brightness of R Mon and therefore the flux measured inside an aperture.
However any aperture enclosing only the center of the comparsion star will pick less flux at poor seeing since the star profile will be more flat.
The only way to decouple the seeing and focus for this unequal behaviour would be to choose an aperture enclosing all or almost all flux of the comparison star for any seeing and focus. That would require an aperture of about 2 or 2.5 times the worst seeing. So something like 10 or 12â fixed in pixels values.
But the diameter of NGC2261 obscuring R Mon could be light years in diameter. So any variability in R Mon could take month or years to travel trough the cloud. If so, then maybe two different apertures would be the solution. One large aperture for full enclosure of the comparison star and a tiny one (one pixel) for R Mon to measure the surface brightness. Then apply a surface correction on the R Mon brightness measurement to get a magnitude value.
So what are AAVSO observers doing for photometry methods for say the Orion nebula variables, T Tau stars, deep in the Orion nebula.
You could always move away from this star and leave it to the professionals and TESS! Just realize that some stars arenât meant to be observed by amateur astronomers with amateur level equipment.
According to the wiki page for R Mon the actual star isnât visible but is buried within the nebula. True IR imaging may be required to âseeâ the actual star. Maybe with a deep red like Cron-Cousinâs Ic filter might yield better more useful results. Shown on the wiki page is a V LC from ASAS data but in another place TESS is credited. How does yours compare? The AAVSO LC seems not to track and with significant bias.
Isnât there an AAVSO group observing these early T-Tauri and related proto-stars? Anyone care to comment? Anyone at all?
R Mon ââŚseems not to track adnd with significant biasâ. Thatâs because people are not doing things in a consistent fashion.
Large image-scale with excellent seeing certainly helps with consistency for the various embedded objects. TESS wonât work at all since the pixels are huge and you need to encompass several to get a consistent measurement. Plus itâs more-or-less unfiltered, so results are difficult to interpret apart from simple periodicities.
I follow about 60 Taurus stars, mostly just in V, and for the most part they are not greatly embedded. HL Tau is a case where the thing present in the visible is not a star, though it certainly looks like one. V1023 Tau (what professional tend to call âHubble 4â) has some nebulosity, but with the short exposures involved, it is hardly visible. I havenât looked at Orion variables, though perhaps Josch Hambsch can comment. I would think that as long as the sky annulus is representative of the underlying nebulosity, things should be OK.
I would like to conclude by determining which photometric measuring method would be most practical using the tools available. Since the profile of R Mon resembles a cloud, while the comparison star is a point source with a Gaussian profile, the only way to exclude the effects of seeing and focus is to use a large aperture that encloses all the flux of the comparison star. For the sample image in my first posting, this requires an aperture with a radius of 7 pixels.
Furthermore, in AstroImageJ you have to switch off âCentroid aperturesâ to allow the aperture to be placed correctly on R Mon. By doing so, I obtain a measured magnitude of about 10.7 for R Mon. This procedure must be applied consistently. Is this a sound methodology?
Since Brian has mentioned my name, I reply to this thread. I observe R Mon since a couple of years in four photometric filter (BVRI). Here are my observations in the AAVSO database.
I use standard 6 pixel for the inner aperture, which translates to about 7 arcsec.
I see the variation in all the filters and hence believe that my observations are decent.
Josch
Thanks Josch. You graphs look good. What is the typical FWHM or HFD of the imaged stars?
Here I did an experiment by applying a Gaussian blur on an image to see the effect of poor seeing on the photometry of R Mon. Seeing has an effect because the comparison star has a sharper profile. The conclusion is that seeing has no significant effect on the measured magnitude as soon the aperture is larger then 1.7 times HFD/FWHM. Measurements where done with AstroImageJ. So if you keep the aperture large enough en always the same it looks all good.
Yes of course there is a group doing just that - the YSO section, and we have been following several stars for a few years now. R Mon may be a star best left to visual observers in fact. Several other similar objects exist such as RY Tau, V380 Ori, V373 Cep and V1331 Cyg (which we see pole-on). It is not as though R Mon and its nebula are distinct objects; they are basically one integrated object, hence the difficulty of determining the magnitude of one as if it were unrelated to the other.
