The VSP contains about 82000 comparison stars. About 2300 of these comparison stars have Sloan magnitudes. So doing photometry with Sloan filters is not very well supported. The Pan-STARRS ‘refcat2’ contains Sloan g,r,i photometry values with a pretty low uncertainty (typical around 0.02) . Are there any plans to include these values into the VSP or is the recommendation to use Johnson-Cousins filters only?
Han
Populating the VSP with Sloan magnitudes sounds like an excellent proposal to me. It seems a fair bet that the majority of the data importation could be fully automated by cross-matching the relevant catalogues.
I use Sloan r for time-series photometry for some scientific purposes, such as for determining asteroidal rotation periods and for exoplanetary transits.
Paul
I haven’t looked to any Pan-STARRS team copyright but even so, I assume getting permission from the Pan-STARRS organization to use it must be easy.
Since I work regularly/process large star databases, I could help by preparing a list with existing VSP AUID identifiers and the three Sloan magnitudes plus some more star data. This by finding cross matching star positions (and similar magnitudes) between the VSP and Pan-STARRS database within maybe 3 arcseconds and taking in account the proper motion star motion. Then it should be the same star.
This cross matching will keep a computer busy for a day or more but is not so difficult for me to implement. This AUID list should make it easy to import in the VSP.
Would the AAVSO staff be interested in such a list and willing to import it in the VSP? If so tell me.
Han
Han: Please, please, go for it! You and AAVSO.
I could also do the cross matching but it sounds like you are already geared up for it. My code is now 3 years out of date and needs some love and care.
Paul
I have send an email to the AAVSO standard email address with the proposal of creating the list. I assume the first review of the proposal is by the staff astronomer.
To be clear this is the catalogue I’m talking about:
Links to the catalogue:
or
The catalogue is licensed under a Deed - Attribution 4.0 International - Creative Commons
so can be used if cited.
ATLAS-REFCAT2 is an all-sky reference catalog containing nearly one billion stars down to apparent magnitude m ~19. The catalog includes PanSTARRS DR1, ATLAS Pathfinder, ATLAS re-flattened APASS, SkyMapper DR1, APASS DR9, Tycho-2, and the Yale Bright Star Catalog. Gaia DR2 serves as the source of the astrometric solution for ATLAS-REFCAT2, with typical systematic errors of < 5 mmag RMS, although this can be as much as 20 mmag near the Galactic plane. The ATLAS Pathfinder telescope was used to collect g,r,i photometry for stars brighter than the 14th magnitude bright limit of PanSTARRS, and to extend the reference system below -30 declination. You can refer to the full collection of ATLAS-REFCAT2 data products in papers or webpages using the DOI MAST DOI.
Hi all,
We can generate sequences with Sloan values, no problem. If a request asks for them we can supply them. I think the intention is to ultimately make this the default format but we still need choices. Seqplot outputs 3 different files to upload to the comp star database, including a file with Sloan values. So it’s not a problem - just ask!
Michael Poxon, Norwich, UK
sequence team
Hi Michael,
I understand from you, that sequences and Seqplot is they way to update the VSP/comp database and this is done by the sequence team.
My question is, if I prepare with some effort a large list with possibly up to 82000 AUID indentifiers and there corresponding Sloan values, will the AAVSO be happy with this list and will it be used? I like to know before I put the effort in.
Can you or other of the sequence team add my large list to the comp star database and does adding such a large list require additional approval/checking by staff or a committee?
Han
To be clear, I’m talking about updating the existing AUID’s with missing Sloan values.
For traceability I will put in the list besides existing AUID, the Sloan magnitudes and corresponding uncertainties the following:
Han
Hello Han,
this needs some discussion within our Science Team. We will contact you about this.
Cheers,
Sebastian
Hello Han,
Very interesting discussion !
I have a question : why do not use the Gaia DR3 ICRS positions, instead of DR2 ? Some errors should have been corrected, and the precision could be better.
Will you include all the available Sloan mag for each star ? Currently, many stars do not have magnitude R or U, for exemple. (of course, Sloan will not replace Johnson Cousins, but this could be better than no referenced mag at all)
Working with VizieR, I have found that a radius searching area comprise between 0.4 and 1 arcsec works well. Such a small radius is preferable to larger radius, for resolving ambiguities in the case of multiple stars or stars located in dense areas.
Waiting impatiently for this new data set.
Cheers,
Christophe
The ATLAS-REFCAT2 contains the Gaia DR2 catalogue positions. So the first step would be to cross match the DR2 star positions with the comp-database star positions. That would allow to link the ATLAS-REFCAT2 to the comp-database and add the Sloan magnitudes to the comp. Yes a delta of one arc second could be enough but I assume in some cases the error could be larger.
A comparison of the match magnitudes would also be advisable in addition to the positional match to prevent a match error. So a conversion from Sloan magnitudes should give a match with comp Johnson-Cousins magnitudes.
The celestial positions of the comp-database stars are pretty accurate once you include proper motion. I have no idea where they based on, but you could use the Gaia DR2 position and proper motions as well. The epoch has to be shifted from 2015.5 to 2000 to make the compatible.
I do not think Gaia DR3 will give much improvement for the intended use. But it will require a similar effort to add DR3 positions. Down to magnitude 16 you will have search through about 60 million stars. For the moment I think it is better to use the original positions and possible use DR2 as a first step. The calculated distance errors could flag any positional discrepancy.
Han
A lot of this discussion is academic because in order to get anything into the comp star database you need access privileges, and to be honest (and friendly!) I can’t see this being granted at the moment.
Seqplot is typically - though not exclusively - used to create new sequences for stars that previously did not have any. Of course if you have requests for Sloan mags for existing variables these can be added, no problem.
I do not want or need access privileges to add. The comp-database can be read via the API and I have already created a local copy here. My suggestion is to add Sloan magnitudes to the existing AUID’s to support photometry with Sloan filters. To assist with that, I can create and provide a list which the cross-matched data. An AAVSO administrator has to take it from there.
I believe the Pan-STARRs filters have different characteristics from the amateur filters. Pretty much the same, but different. I typically collect my own references in PostgreSQL, then use Q3C to match positions. I extract position with sextractor for astrometry.net (local data) resolved images. It is possible within the PostgreSQL query to make decent estimates on variability of the PanSTARRs images and to help narrow down the error bar statements.
“I believe the Pan-STARRs filters have different characteristics from the amateur filters.” It is not just the filters, but the whole light-path including optical train, detectors, and atmosphere, so everyone’s sensitivity is somewhat different from everyone else’s. That is the reason calibrations are required.
\Brian
Possibly relevant to the discussion of g,r,i catalogues is a paper regarding the VPHAS+ survey, just appeared on astro-ph by Janet Drew and gang:
…which outlines the complex calibration of this survey along the southern galactic plane (an important stretch of sky as far as variable stars go). Looks as though the catalogue will cover the magnitude range about 13 < r < 19. The paper is accepted for MNRAS, but the files will appear at VizieR some time downstream.
\Brian
Are you saying that the ATLAS-REFCAT2 magnitudes are biased and that the reported uncertainties are larger than stated? Where can we find reports confirming this?
Previously it was stated the Sloan values are quite reliable, since it is based on new observations from the Pan-STARRS and smaller ATLAS telescopes as well as massaged and transformed data in existing catalogues.
Currently there are practically no Sloan magnitudes in the VSP available, so adding them will be an improvement.
Han
Sorry for causing confusion. I was not saying anything about the Pan-STARRS ‘refcat2’ photometry. I was saying that merely having some Sloan-type filters will not make your data conform to the Sloan system. You will always need a transformation, even if modest, because whatever your particular optical set-up is will inevitably differ from the scheme used for ‘refcat2’. This would be the case even if your filters were a perfect match since everything else in the light-path is different. So I wouldn’t worry about am/pro filter differences, per se, but work toward getting really consistent photometry.
\Brian
Hi All,
This has been a wonderful discussion and I appreciate the range of perspectives that have been shared. Given where we are, I thought it might be helpful to offer some broader context and observations.
Over the last three years, I’ve had countless conversations, both internally and externally, about the AAVSO’s photometric reference system. As it exists today, that system is effectively an amalgamation of no fewer than 26 different catalogs spanning ultraviolet through near-infrared wavelengths. Each of these catalogs brings its own idiosyncrasies in terms of photometric zero points, uncertainties, calibration philosophies, and filter systems. As you might imagine, this introduces complexities that accumulate over time.
About a year ago, I attended a Flux Calibration Workshop at STScI where one message came through loud and clear: many of the science cases professional astronomers are pursuing now, and those planned for the next decade, require 1% absolutely calibrated photometry. There are, of course, a many exceptions to this statement where 10% or even 50% error bars are acceptable (e.g. most eruptive events); however, lets assume that we should strive to meet this goal in the future.
Unfortunately, VSD was not built with a 1% requirement in mind. As it exists today, AAVSO’s photometric reference system is effectively an amalgamation of no fewer than 26 different catalogs spanning ultraviolet through near-infrared wavelengths. Each of these catalogs brings its own idiosyncrasies in terms of photometric zero points, uncertainties, calibration philosophies, and filter systems. On average, the uncertainties within a photometric system in VSD are at the 3-5% level, which is pretty darn good in my opinion. Unfortunately, I suspect uncertainties relative to an absolute scale are much greater.
With these things in mind, I’m also left wondering what will happen in the near future. I’m aware of ~5 photometric catalogs that are in the works that use SED modeling or spectro-photometry to anchor themselves to GAIA. For example, there has been considerable progress on APASS DR11 in the last few months and GAIA will be used to rank the accuracy of measurements before the data are averaged into the output catalog.
Complicating matters further, the AAVSO’s long-term plan will have us invest considerable resources into VSD in 2027 (VSX is occupying 2026).
Suffice it to say, I see considerable opportunity and also significant uncertainty with regard to the long-term evolution of VSD.
In January, I plan on putting together a working group to explore these topics (and more) in greater detail to advice AAVSO on how it should proceed. If you are interested in helping out, please send me an email or PM.
Kind regards,
Brian
My idea was to also cross-match the standard-field stars to the ATLAS-REFCAT2. So to provide also a list of the Sloan magnitudes for the standard-field stars. This will allow you to find the Sloan transformation factors of your own setup using (guarantied) non-variable stars.
Thank you for highlighting future developments and trends.
For my understanding, is this 1% of the magnitude value? So for a magnitude 10 star, the standard deviation uncertainties (1-sigma errors) should be equal or less then 0.1 magnitude?
If so the ATLAS-REFCAT2 as reference does meet this criteria easily.
Han