This is my first thread about variable star research. I am an astronomer-amateur and I would like to present one of my latest variable star discoveries.
Generally, variable stars can be found via two methods. First, is to observe a random field in the sky and then check survey data for confirmation if we found anything that vary in brightness. This, what I do, is opposite - the second method. By checking multiple catalogs (Gaia, WISE, 2MASS and many other), I find new variables by looking at archive photometric data. ASAS-SN, ZTF, CRTS and many more, released their photometric observations of millions of stars. Too many to handle, also too many for a precise automatic check for new variables. That means, you can still discover new variable stars with your computer (and this is called data mining), so I'm using this to find targets of interest. For now, I published about 500 of these and more are coming shortly. And now, instead of choosing a random area, I prefer to spend my observation time on my variable that I find interesting. To prove that this method is much better, I would like to present one of recent findings - MGAB-V266.
MGAB-V266, as the name suggests is my 266th variable star, is an eclipsing hot subdwarf. A quite small star (0.2-0.3Rs), but very blue, bright and hot. I found something orbiting the star with period just 5.5 hours, showing high reflection effect in the light curve. That's a new eclipsing binary of HW Virginis type! One of only ~50 known so far. The target located in Lacerta is relatively bright. During maximum, it reaches 14.8 mag. Seems very faint (Pluto is even brighter!), but nowdays good findings brighter than 15 mag stars are rare. The discovery (actually a co-discovery) was published to VSX in August 2019. There was an identified variable star by ATLAS collaboration, but with a wrong period and pulsating type. That means, eclipses were not noted, so the co-discovery of an eclipsing companion is reasonable, but generally I don't publish such ones.
But what's so interesting about this star, that I decided to write as my first post? This star is probably a record-breaking eclipsing binary of all currently known. We can actually speculate this, because eclipsing companions around novae and eclipsing exoplanets around white dwarfs might have larger amplitudes. But, MGAB-V266 is one of these 'normal' eclipsing binaries. The companion causes the target to dim more than 6.5 magnitudes! During maximum (14.8-15.0 mag), the target reaches below 21.5 magnitude. And to date, we still couldn't reach the minimum! Only telescopes with aperture 0.5m and larger are able to give a try, because latest observation with 16" scope still ended with 'fainter than' measurement.
For now, we know about only three examples of eclipsing binaries with extreme amplitudes. NN Serpentis and Konkoly J064029.1+385652.2 (later: J0640). The first one is actually a white dwarf with an orbiting red dwarf. The second one is more similar to MGAB-V266. This is a hot subdwarf with a confirmed red dwarf companion. The amplitude is also deep (5-6 mag, but I am not sure if in g-band or in r-band). MGAB-V266 reaches deeper, so we can expect the target is cooler than J0640. Based on my analysis and from the paper, J0640 was found to be a M6V red dwarf and the amplitude in all grizy bands (from Pan-STARRS1 data) show similar reflection effect amplitude.
What's different about that in MGAB-V266? The reflection effect looks totally different. Of course secondary eclipses are visible (and that was expected, because this is not a white dwarf, but a hot subdwarf), but the amplitude changes dramatically in longer wavelengths. It is the smallest in g-band and strongest in y-band. This is a very good indication what type of companion could it be, but still haven't found any paper (I might start it shortly?). For now, I found three possibilities:
- the amplitude is higher in longer wavelengths, but it reverses soon (eg. g: 0.30, r: 0.40, i: 0.40, z: 0.35, y: 0.30) -> this indicates an obvious M-dwarf, or even late K-dwarf.
- the amplitude is the same in all wavelengths (like J0640) -> most likely a red dwarf
- the amplitude grows in longer wavelengths (eg. g: 0.30, r: 0.40, i: 0.45, z: 0.50, y: 0.55) -> a possible brown dwarf (???), but a late M-dwarf is very likely
While having found several compact binaries (eclipsing and non-eclipsing, where reflection effect is still present), only a small percent of them show the third option. MGAB-V266 is actually one of these! Both stars however (with J0640) present similar ingress and egress durations compared to total eclipsing time. Still, these stars are the only two that show total eclipses around HW Virginis stars. What's also very important, is that MGAB-V266 shows lower reflection effect than J0640. Does this mean that the hot subdwarf has lower luminosity and might be smaller, so a brown dwarf can be large enough for a total eclipse? That's only a good indication, but that's not enough to state with current data. That means, MGAB-V266 is definitely a good one for a paper.
This, what we can expect, the companion is at least a late M-dwarf (M8+), but I cannot 100% rule out a possible brown dwarf. This, what is currently needed, are radial velocity observations. And this is the hardest part for astronomy amateurs, because this require help from professionals. For now, we decided to keep eye on this system. Such objects often show eclipse timing variations caused by a nearby brown dwarf or even exoplanets! Will we find anything around? We will see. Observations with 8" telescope shows that our timing precision can be as good as a few seconds, which is enough to start searching around. Typically, a tertiary brown dwarf object causes O-C differences with amplitudes up to few minutes!
ZTF data show that there aren't any large variations in short timescale, but that's not surprising. Typically, periods might be a few years, so it will take some time before we finally realise that O-C signal is real. We (amateurs) have already started a campaign in search for O-C differences. We have hope that it will end with success, so we can publish a paper... like this one: https://arxiv.org/abs/0903.1357
What if not? Maybe another time! I have found several HW Virginis systems, as my colleagues (other amateurs) as well. But, they are not like J0640 or MGAB-V266, so eclipses have lower amplitudes. Actually, parameters of MGAB-V266 give us the largest chance, because relatively high brightness (14.8 mag) and extreme amplitudes give us the most precise eclipse timing measurements. And here's one more reason why it gives more opportunity than J0640: latest finding is 2 magnitudes brighter, so observations will be more precise (and even possible with small scopes).
How does the eclipse look like on your frames? Here's my timelapse from latest observation with 8" f/4 telescope and ASI1600MM-c camera. Each frame corresponds 60 seconds of exposure. Warning: 7MB file!
And the star can be found in VSX database through this link:
Any observations of this system are welcome! We are taking this O-C project very serious, so we observe this object regulary for more timing measurements.
Edited by libmar96, 08 October 2019 - 12:45 PM.