Hello Everyone,
I've been exploring the VSX database and the amazing diversity of variability types as I learn more about Variable Star Astronomy. I have a few small questions about type annotation as represented in the VSX database.
1. There are 118 VSX records with apparently 2 sub-types. For example, TTS/ROT/DIP (28), LMXB/BHXB/XN (25), ELL/RS/BY (10) being the most common cases of this. So for example, LMXB/BHXB/XN stars are low-mass X-Ray binaries (LMXB) with a black hole companion (BHXB) AND exhibiting occasional rapid increases in brightness (XN). Am I interpreting the multi-subtype notation correctly? I assume the subtype order makes no difference. (?)
1b. Is type UGZ//IW:|NL: for V0538 Mon (oid 19366) a typo (note the double slashes)? The info button brings up a blank page.
2. There are 15 stars that list both + (AND) and | (OR) notations. Seven of the 15 stars are of type LMXB|CV+EA. I assume that A|B+C means (A|B)+C and not A|(B+C). Is that correct? So, LMXB|CV+EA means that these 7 stars are definitely Algol-like eclipsing variables (EA). Does this mean that the companion star is either a black hole or neutron star (LMXB) or a cataclysmic variable of some unspecified type (CV)? Or perhaps the main star is the CV?
Thanks,
John Rachlin (RJOJ)
Lots of questions...
an I suggest some essential documents to start with?
https://www.aavso.org/sites/default/files/2%20-%20VariableStarTypeDesignationsInVSX2013_4.pdf
https://obswww.unige.ch/~mowlavi/Images/variability_supertree.pdf
https://www.aavso.org/sites/default/files/5%20-%20VSX%20Variable%20Types_Alphabetical_20160303_5.pdf
These documents are part of the basic documentation for the course “Classification of variable stars and light curves” which is given each year (see the CHOICES courses).
Hope this provides some answers to your questions.
JBD
Thank you JBD - that's very handy. I'll be watching for Choice courses in 2024. -jnr
Hi John
As a long-term user of the VSX database, I am curious as to how you went about "exploring the VSX database". I am similarly interested in exploring the VSX and doing various types of data mining.
My question concerns how you went about reading the entire database: Did you, for example, simply run all of the searches that are possible via the extended search facility on the opening VSX page, or were you able to somehow read the database in one single go, by some other (more efficient) means?
Regards
Paul (YPFA)
Hi Paul,
Some students of mine and I are developing an exploratory dashboard for the AAVSO VSX catalog as part of a collaborative research project (and as an educational experience for my Database Design and Advanced Programming students at Northeastern University.) For the prototype, I downloaded the vsx catalog from the CDS FTP site (https://cdsarc.cds.unistra.fr/viz-bin/cat/B/vsx) and wrote code to convert this single table into a normalized SQLite database from which I can then write SQL queries to identify stars with interesting variability types and that are accessible to my south-facing patio in the light-polluted skies east of Phoenix AZ. Some gotchas in the data processing include the fact that the VSX table makes the "interesting" choice to sometimes store an amplitude in the minimum magnitude column, and that the variability type should really be modeled as a many-to-many relationship between the variable star and the catalogue of AAVSO variability types. (Building the join table is what motivated my original question.) When properly structured, you can then ask questions such as "Show me all M-type stars above declination -20 with a maximum magnitude < 8, and which are currently at least 20 degrees above the horizon - sorted in descending order by period. (Conversion to horizontal coordinates obviously required some additional code!) Thus, my hope is that the dashboard will be useful to AAVSO observers looking to identify variables in their area of interest that are accessible to their available equipment. By integrating information about number of historical observations and/or active campaigns, we hope to help observers target potentially under-observed variables and thus make more impactful contributions to the science of Variable Star astronomy. And by embedding visualizations into the dashboard we can explore the full vsx data more holistically and highlight variables with outlier properties. I hope to have a prototype dashboard developed by the end of the semester and our plan is to make all of the code and dashboard open source and freely available. (That's free as in beer AND free as in freedom!) Stay tuned.
John
Hi John
That sounds both interesting and useful. Are you aware of the AAVSO's Target Tool (https://targettool.aavso.org) - it attempts to do similar things, but maybe not in as much detail as you are proposing? It is also tailored to some degree to a user's location but again, not in as much detail and not using horizontal coordinates ...
Paul
I am aware, thank you! I think our tools are complementary. We would like our dashboard to enable users to query the vsx catalog and visualize results while at the same time highlighting stars that are under-reported and/or part of an active observing campaign. We envision having some sort of a link-out to the target tool when a user identifies stars within their area of interest and that are accessible to their geographical location, LST, and available equipment and that are also part of an active campaign.
As an example use-case, in testing a dashboard prototype, I came across a star that would be accessible for my 10" dob and southwest location: SZ Aqr, an interesting (to me) LC-type variable with a magnitude range of 8.4 - 10.4 V [1]. Alas, only 15 observations between 1972-2009 have been submitted [2], probably due to the lack of comparison stars and it's southerly declination. The compstar team has now kindly remedied this in the last couple of days. But it occurs to me, if the dashboard could readily highlight stars of certain types, amplitudes, and periods that have no comparison stars (perhaps for specific FOVs) we could help the compstar team prioritize their important work - i.e., give priority to stars that, while not necessarily part of an active campaign, are potentially interesting and accessible to a wide range of observers.
According to the AAVSO definition: LC stars are "Irregular variable supergiants of late spectral types having amplitudes of about 1 mag. in V." It was the word "Irregular" that caught my interest! Also interesting is that one of the past observers of SZ Aqr was Gonzalo Beltran of Bolivia (observer code BZX) who repeatedly reported that the star was fainter than magnitude 12.0. That's interesting and surprising given the specified catalog range, isn't it? (Gonzalo was/is an experienced observer who had over 1000 observations in some years. [3])
Sources:
[1] VSX: https://www.aavso.org/vsx/index.php?view=detail.top&oid=871
[2] WebObs: https://app.aavso.org/webobs/results/?star=000-BCT-587&num_results=200
[3] Observer Awards 2010-2011: https://www.aavso.org/observer-awards%E2%80%942012
Hi John,
These stars have ASAS-3 and ASAS-SN observations so one can see how they behave in the long-term.
This one was more erratic (several periods going on) in the ASAS-3 era (2000-2009) but more recently it become highly periodic.
See the ASAS-SN phase plot with P= 107 d.
I revised the VSX entry accordingly.
About the <12 mag. observations, these might be the result of a misidentification so I will contact Gonzalo to check.
The star is never fainter than V= 10.6.
So, two things:
1) Lack of observations in the AID doesn't necessarily mean lack of observations for a given star in the sky surveys. Remember that we have ASAS-SN, ZTF, TESS, etc. out there. So checking those sources is also useful.
2) Classifications in VSX might be improved after all the available data are checked. We don't have enough manpower to check all the stars so observers' revisions to VSX information are always welcomed. There is VSX online documentation to learn more about the VSX usage (Manual, FAQ page, guidelines) that you will find at the bottom of any VSX page.
Cheers,
Sebastian
Hi John,
Your interpretation of the types is basically correct.
The order is kept the same way for consistency and because there is a hierarchy here.
E.g. LMXB/BHXB/XN: X-ray binaries might be high mass, intermediate mass or low mass. Any BHXB will be one of those three and then, both high mass or low mass can show large amplitude outbursts. One of the subtypes is describing the physical nature of the components and the other is describing its behaviour.
ELL/RS/BY is an incorrect notation, because a star can't be an RS and a BY at the same time. This is a giant vs. dwarf thing.
If the luminosity of the components is not known, a | should be used instead of a /. "/" is used for subtypes and "|" means "or". Thus ELL/RS|BY.
If any of the three classifications is uncertain, then it should be ELL|RS|BY.
That is probably the case right now in VSX: those 10 objects come from the BEST survey and the classification is very uncertain. They are not subtypes, they are just three possible classifications, so ELL|RS|BY is what should be given. Something to correct in VSX :)
Thanks for noting the typo in 1b. Corrected.
About 2), you got the subtype thing right, but in a CV it is not that the CV is the companion OR the main star. It takes two to make a CV. We are talking about a binary system that is eclipsing but we do not know if it is a CV or if it is a LMXB. CVs may show x-rays too.
I would replace the terms companion and main star by primary and secondary in this context. And in all these types of variables, the type applies to the two components. They wouldn't be variable on their own.
Cheers,
Sebastian
Sebastian - Thank you for the detailed response and especially for enlightening me about Cataclysmic Variables. (Hellier's book, "Cataclysmic Variable Stars: How and Why they vary" is on my reading list.) -John