This is my first post on this forum. Posting here was suggested by Sebastian Otero, after my first (quite imperfect) attempt to submit a variable star discovery. Since that time I have collected additional data and reanalyzed all the sessions data.
Although I have had experience observing asteroids (photometry, including light curve analysis of about 100 asteroids...), I have only imaged a few variable stars, and this is my first VSX discovery submittal.
I would appreciate any and all comments on the format and content of the submittal, or on the analysis itself.
The analysis was performed as follows:
MaxIm DL image capture (300s second exposures)
MaxIm DL calibration (bias, dark, flats, dark-flats) and alignment
Prism astrometry and batch photometry (for each session)... I chose a Reference star close to the variable and with a similar color to it rather than the Reference stars chosen by Prism (not consistent session to session, too close to the edge of the image, and with a range of colors...). Several stars of similar color were used to check the results.
Spreadsheets (Microsoft Works and Excel) were used to reformat the Prism output and adjust for HJD vs JD (using the British Astromonical Association (BAA) HJD calculator) and to prepare a VSTAR-compatible format.
AAVSO VSTAR was used to develop the phase plot for 5 sessions during July and August 2018.
Here is the VSX discovery information:
UCAC 563 - 10905 (I understand that the Gaia identification is preferred but I was reluctant to change the name of the submission)
Position: Gaia J2000 RA 20 13 53.277 Dec +22 24 18.809
Variable type: EW
Maximum magnitude: 15.5 (based on the VSTAR mean of the data at the maximum)
Bandpass: CV (as a description of the Baader UVIR filter 400-700 nm)
Minimum magnitude: 16.1 (based on the VSTAR mean of the data at the minimum)
Bandpass: CV (as a description of the Baader UVIR filter 400-700 nm)
Peroid: 0.365 (days) (VSTAR analysis... double the peroid suggested by VSTAR...)
Epoch: 2458335.857 (VSTAR analysis)
Supporting information: VSTAR Phase plot (attached in .png format)
Additional information (e.g. calibrated images, HJD vs. mag tables...) can be provided as necessary.
Again, any and all suggestions and comments to make the VSX submission better (or acceptable) will be very welcome,
This star is bright enough for ASAS-SN, so you can download their measurements using ASAS-SN Sky Patrol. Put Gaia coordinates and try last 3000 days. It will take a while.
I recommend doing this, as your timescale is so short and your period is accurate only to 3 decimers. Combining with ASAS-SN would allow you to measure to fifth or even sixth decimer for this star. The epoch is fine. V band measurements are more preferred, so your CV datapoints would be shifted to match ASAS-SN values, which are in V. And remember to add references, which you will find in FAQ section.
This star has a different secondary minimum, so you can add under remarks "Min II = XX.X V".
Congratulations on your discovery!
Which of these Gaia source IDs does the target correspoond to?
or neither of these?
I'm interested in obtaining the data from the Gaia archive web page:
You may be interested to kniow that Cliff Kotnik has developed a Gaia obs source plugin, currently under test: https://www.aavso.org/comment/61146#comment-61146
I thought I should comment on this in this topic.
In the Gaia DR2 archive, you can see which Gaia sources have light curve data by looking for gaiadr2.gaia_source.phot_variable_flag = "VARIABLE". This query
select count(*) from gaiadr2.gaia_source where phot_variable_flag = 'VARIABLE';
returns 550737. So out of 1.3 billion only 0.5 million have been flagged as potential variables. And only these potential variables have light curve data available for the VStar plugin I wrote to access. I think that the number of variables will increase with the next data release (2021?), but this is how it stands now. Anyway, check the phot_variable_flag column if you want to know if light curve data is available. Or you can run the VStar plugin and see if you get any data.
It might also be interesting to note that the number of potential variables in Gaia DR2 is about the same as the number of variables in VSX. However, a crossmatch I ran using celestial coordinates found between 105,000 and 120,000 of the sources in VSX match Gaia DR2 variables. I suppose that is good news (lots of potential new variables) and bad news (Gaia did not find many variables).
Thanks... The variable is GAIA DR2 Gaia ID R2 1830174423422381440.
I have attached an image (.jpg) indicating the star... North up, East left... FOV ~ 20 X 13 arcmin
That ID doesn't seem to have data associated with it.
After looking at your light curve, I think this start might be a Beta Lyrae type (EB) instead. It's a bit difficult to say for sure with the data you have, but it looks like the secondary eclipse is flat. Just a brief side note on the physical differences in case you are curious.
EW systems are actually two stars in contact, meaning they look a bit like a rotating peanut. This type of system will never have a flat bottomed eclipse because there isn't really a transit. In a transit, the smaller star crosses the bigger star and the light is roughly constant except when entering in front of (ingress) and exiting from (egress) the the bigger star. In contact systems, by their very nature the eclipse is sharp and will come to a point as they are essentially always in ingress or egress.
Anyway, I don't know that it's completely obvious in your case. The secondary eclipse does look flat to me but there is quite a bit of scatter so it is difficult to say for sure. I might bin the data in phase and see if it indeed looks more flat. It may not give you a satisfying or obvious answer (unfortunately these things often do not), but if you are interested in playing with the data some more it's worth a try.
Staff Astronomer, AAVSO
Thanks Pablo. I have collected data for three additional sessions. They are still being analyzed. I will pay particular attention to the shape of the secondary minimum.
With respect to your post I think there is some misunderstanding about the EW type (W UMa) binary stars.
The phenomenological classification used in GCVS is sometimes misleading. Subtype beta Lyr (EB) is so indefinable that is hardly should be a prototype of any subclass (Percy J., 2007). Depths of the minima depends not only on the difference in the photosphere temperatures of the components but also on their individual masses, the mass ratio of the pair, the fill-out factor (f), the inclination of the orbital plane (i), limb-darkening coefficients etc. In general it is necessary to carry out a comprehensive analysis of the eclipsing binary systems instead of using only the geometrical characteristics of their light curves.
The modern morphological classification of the eclipsing binary stars is based on the equipotential surfaces of the total gravitational field of the pair. These are surfaces where the sum of the rotational and gravitational energy per unit of mass is constant. The equipotential surface of the total gravitational field of the pair passing through the first Lagrangian point (L1) is called the critical Roche boundary and it determines the smallest possible dimensions of the components of the binary system. For each component in a binary system, the space covered by the Roche boundary is defined as Roche lobe. I would not want to go deeper into the theory but just to mention that according to morphological classification the eclipsing binary systems are divided into three types:
In EW type binary systems if the eclipses are total, the minima are deep and may contain flat areas. If eclipses are partial, the eclipse minima are shallower and contain no flat areas.
There are many, many, many contact EW type binary stars that undergo total eclipses and transits and depending on their mass ratio (q), difference in the components temperatures, the fill-out factor (f), the inclination (i), limb-darkening coefficients etc., they have one or both flat minima.
About a quarter of observed by me EW stars undergo total eclipses and transits and have one or both flat minima (see the attached pictures). Those who are interested can find some examples in our publications (more are in-print):
I may be causing this confusion because I orginally submitted the discovery to VSX as "UCAC 563 - 10905". In revising the submission, I was relautant to change the name of the submission... some databases don't alow the nane or key word to be changed. I presume that I will ultimately need to change to the Gaia DR2 ID or make a separate submission.
UCAC 563 - 10905, J201353+222419, and Gaia DR2 1830174423422381440 refer to the same star (at RA 20 13 53.277 Dec+22 24 18.806)... if I understand the Gaia DR2 "designation" (or "Source designation") correctly.
Any clarification would be appreciated.
First of all, be careful with the identifiers, you submitted the star as UCAC4 563-109605 not UCAC 563-10905. Also in your submission the catalogue name formats are not okay. Disregard UCAC3 and Gaia DR2. We don't want those long names in the database, noone will use them. I don't know why the Gaia DR2 issue came up. We use Gaia DR2 J2000.0 positions but not their identifiers.
As you suggested, I downloaded the ASAS-SN data (3000 days, V magnitudes only) and analyzed it with VSTAR. There is definately variabity in the ASAAS-SN data (about 15.4 V mag to 14.8.V mag... similar delta mag to my measurements). I have not (yet) been able to find a reliable peroid.with VSTAR. I will continue to work with VSTAR to try to refine the estimated peroid of 0.365 d.
With the ASAS-SN data I will be able to adjust my measurements to V magnitudes... and update the submission.
A New/Revised VSX submission has been made. It is "New" because the automated VSX administrator deleted the earlier submission. This submission (UCAC 563-109605) includes changes in response to VSX Forum comments.
Revisions based on VSX Forum comments:
- CV magnitudes converted to V based on calibration with ASAS-SN data over 3000 days
- Period estimate extended to 4th and 5th digit (days) based on adding datta to extend the observation period by aproximately a factor of 2
- ASAS-SN data checked for 3000 days. similar variability observed (14.8 V to 15.4 V)... no period established
- Position changed to Gaia DR2 J2000 position
- Additional data (see phase plot) continues to support EW vs. EB variable type
- New phase plot with the additional data
- New Star Chart (MaxIm DL image) for star identification.
Sincere thanks to all the forum members commenting. It clearly resulted in an improved submission.
Any additional comments would be welcome,
I have downlodaded data of UCAC4 563-109605 from ASAS-SN Sky Patrol and have tried to find its period with PDM. I have used only data from Camera, bc and bd. The best fitting period for your period would be 0.36510 d. PDM shows clear peak at this period but its phase plot is not so clear. I hope, that this PDM analysis and phase plot help your submission to VSX.
Thank you for your analysis. It is extremely useful. I ran 300 days of ASAS-SN data for the V filter and the bc camera. I was concerned with the very significant scatter in the data and the difficulty in identifying a period. Knowing the answer was approximately .365, I was able to get VSTAR to identify a period of .18255 d (half of the .3651 d). I will continue to try to extract what I can from ASAS-SN.
As suggested I have combined 3000 days of ASAS-SN data (the V filted data from cameras bc and bd) with the data from images I took between July and September 2018. The combined phase plot is attached. The results look quite consistent with Kiyoshi's analysis.
One thing that I have not figured out is how to display different data sets (ASAS-SN camera bc, ASAS-SN camera bd, and my data) differently (e.g. with different colors) in VSTAR. help would be appreciated.
Happy to help with this.
Have you used the ASAS-SN VStar plugin to get UCAC4 563-109605 data to which you then additively load your own dataset?
Thanks for the offer to help. I have used Canopus for asteriod light curves and phase analysis many many times, but this is my first experience with VSTAR.
I have been using Simple Format (HJD and V mag) Excel flies input to VSTAR by "New Star from File...". I have been additatively loading files. One file for ASAS-SN camera bc, One file for ASAS-SN camera bd, And 11 files for my 11 imaging sessions,
Hope this is clear enough.
You don't need to add all files separately. Just use one file with all your data and another file with all the ASAS-SN data (beware of g mags. since now ASAS-SN is making available V and g data, look at the last column and select the V data only).
Thanks, I can "simplify" the data files (one for ASAS-SN data and one for my data), but I still don't know how to show the data separately in the VSTAR plots. Is this a problem with using a "Simple Format", or is there a mechanism in VSTAR to differentiale among files?
... getting close but still needing help... even though I read the VSTAR manual... and the File Format and the Preference sestions several times...
Sorry, I posted this comment to a false place.
I wanted to post "List of pending submissions" this comment.
there is already an entry in the WISE_NewVar catalog, which matches to your new variable.
205034.37 472027.6 WISEJ205034.3+472029 EW W1=12.582 W2=12.676 amp=0.517 P=0.3853084
I hope, but you can submit your discovery to VSX.
I hope this is the right place to post this topic..:)
In the last two years, I've been blessed to work and be mentored by a professional astronomer Dr. Andy Odell. Among other things, Andy specialized in binary star systems and Image Reduction and Analysis Facility (IRAF). He helped me with a few projects...in particular, V963 Per. (Primary minima/secondary minima). I've become fairly confident using IRAF....which I'm still wondering if this is an asset or a curse!!! Unfortunately, Andy passed away a few months back and now I'm looking to be more active in aavso alerts/data.
If I'm looking at this correctly, webObs does not support IRAF? A work around? I can give more details on my data reduction methods if needed:)