Forums / The Science / Data Analysis / The Evidences for Distinguishing Physical Variable and Eclipsing Variable
The Evidences for Distinguishing Physical Variable and Eclipsing Variable
Hi all,
I am an amateur interested in eps Aur lately.
At the present time, the mysterious secondary of eps Aur system is being discussed warmly. But in my opinion, we should consider in earnest whether the eps Aur is not a binary at all, but a pulsatile variable.
The eps Aur is classified as eclipsing variable but physical one mainly according to its profile of light curve, because one thinks that only eclipsing variable can have such a profile of light curve. But I think that maybe we have only a little understanding of physical variables, consequently, it is better for us to believe the observed data instead of the prior knowledge.
What are the strong evidences for distinguishing between the physical variable and the eclipsing one? The following is my opinions:
1. The data of line of sight velocity can give the strongest evidence of eclipsing variable model.
Fig.1 is a sketch of eclipsing variable (ignored ellipticity). Two companions rotate around their system centroid. Because the mass of the primary is very close to that of the secondary, the orbit diameter of the primary is nearly the same as that of the secondary. In addition, their orbits are nearly in a position of edge-on, thus, the line of sight velocity of the primary IMO can be obtained and should be consistent with the following law: The line of sight velocity reaches the extremums in 1990 and 2003 with the directions toward us and away from us respectively, while in 1996 and 2010 the line of sight velocity is close to zero (ignored the proper motion of system). In 2010, the primary should reach the farthest position from us (point d). This phase relationship is a strong evidence of the eclipsing binary. What a pity I just could not find a reference to such a proof, yet. Anyone? If the line of sight velocity measured is very small at all time and/or without the above phase relationship, the eclipsing variable model should be ruled out.
2. The interferometry can provide the strongest evidence for the eclipsing variable model. If eps Aur were an eclipsing variable, the swing of F star along its orbit could be observed by interferometry because F star has a large orbit diameter, too. If the swing of F star can't be observed, there are three reasons: a). It is a lot farther than that we have estimated; b). the mass of the primary is much larger than that of the secondary; c). Eps Aur is a physical variable, and without any secondary.
3. Except for the above two seasonal changes, any out-of-eclipse variations with a period of 27 years are the strong evidence of pulsatile variable, because the change of the primary should not be influenced remarkably by the orbit motion of the secondary. If possible, I suggest investigating whether the magnetic field reverses every 27 years.
4. Perhaps the red shift during ingress and the blue shift of egress have nothing to do with the eclipse. If it is a pulsatile variable with a period of 27 years, the spectrum changes can be explained well: When contracting, red shift occurs and brightness dims while expanding, blue shift happens and brightness increases. Because the speed of expansion is higher than that of contraction, the blue-shift is bigger than the red-shift, so the asymmetry can be well explained, too.
5. If the interferometry has an enough resolution, perhaps the diameter changes of eps Aur can be observed when contracting and expanding.
If observing in light of the method put forward by me, I predict that the observed results will prove favorable to the pulsatile variable model.
The above is only my preliminary supposing so that I' m not quite sure.
Liang
Hi Liang > But in my opinion, we should consider in earnest whether the eps Aur > is not a binary at all, but a pulsatile variable. Eps Aur is currently considered very reliably to be both: A) Its a quasi-periodic pulsator that shows several shifting periods in the 60-250d range with a total amplitude of up to ~ 0.2 mag in the V band (more in the bluer bands, less in the redder bands) when the quasi-peri- odicpulsations are pronounced or several of the periods are beating. B) And an eclipsing binary with a period of 27.1a and an amplitude of ~ 0.7 mag V. Well, its not a classical EB where one star eclipses an other star, but a giant star that is partially eclipsed by a large debris disk, so one
might want to introduce a new type to account for this. KH15D is an other of these beasts that some consider to be even a central binary star with a debris disk around them where just one component of the binary peeks out ofthe disk from time to time due to their orbital motion that is not in thesame plane as the debris disk (in this case the amplitude is > 4 mag). Its often good and OK to challenge existing models. But one should also do this reasonable and at best with some evidence or discrepant obser- vations that somethingmust be wrong with the currently acceptedmodel. You did not give anything like this. Science should not be based on belief or guessing. There are real proofs that eps Aur is actually a binary star and these cann't be doubted just by a feeling or guess without any evidence. Otherwise everybody would have to doubt everything as soon asthey think or do anything. IMO its already a sufficient indication that not a single known pulsator can oractually does show a light curve behaviour that resembles the total eclipses that are observed with eps Aur. Show me a pulsator light curve that can or actually does create such a light curve and is as strictly periodic every 27.1a and we can seriously talk about this. CS WR
Hi Wofgang,As far a I am aware, the origin of the out of eclipse variations has not been established with any high degree of certainty. While they may be due to pulsations of the F star, I have not seen any spectroscopic evidence published to support the pulsation hypothesis and the earlier CHARA interferometry work did not find variations in radius. http://arxiv.org/abs/0810.5382The F star is suspected to have circumstellar material of its own, responsible for H alpha emissions which vary over similar timescales, though not as far as I am aware in sync with the brightness varaitions. This material might have some role to play in the brightness variations. Robin
Liang,Good job and good questions! I've never thought of getting magnetograms on eps, but I'm not sure it's possible given the distance.I agree with your scepticism because it seems logical that there should be a simpler explanation for the star, but ultimately must agree with Robin and Wolfgang. There is a lot of literature on eps Aur and getting caught up with the current theory (or finding out why we have the current theory) is very time consuming. I actually spent the last year reading some of the "classic" works since 1821 when the star was first noticed to be variable. I doubted the disk-theory to begin with, but after time it appeared to be the most logical explanation for the observed phenomenon. I created a post on the "evolution" of the disk theory that you might want to read to get yourself caught up:http://www.citizensky.org/forum/history-and-evolution-disk-theory-epsilon-aurigaeI think Robin and Wolfgang covered the basics above, but I'd like to elaborate on a few things I've learned from my Ph.D. thesis (on eps Aur with Dr. "Bob" Stencel as my research adviser). I've read an extensive amount of literature on eps, but have yet to cover the ~400 articles (in English) on the topic. I have one point about the out-of-eclipse variations:After reading much literature on this topic I am not aware of anyone who has been able to pin down the cause of the out-of-eclipse (OOE) variations to any single effect. Gary Henson (in his 1989 Ph.D. thesis) stated that the OOE variations could be due to non-radial pulsations in the F-star, but he could not determine the mode of pulsation with much certainty.I have two points about interferometry:The first is a correction to Robin's post above. The non-pulsation paper was from PTI (the Palomar Testbed Interferometer at Mt. Palomar, CA), not CHARA (the Center for High Angular Resolution Astronomy, on Mt. Wilson, CA) using the wide-band visibility mode. As part of my Ph.D. thesis on eps Aur, I'm re-analysing the data using a more robust statistical method. The method that I used to analyse the data on that paper has much room for improvement, but it was the best we were new to interferometry then and it was the best we could do at the time. I will also look at the narrow-band visibility mode (think spectrally-dispersed interferometry). These results will come out in a future paper.The second point is about what interferometry can do and can't do. An interferometer is effectively a very high resolution telescope with a very limited field of view. Unless there is a reference point in the field of view (which there isn't with eps Aur), the left-right motion of the star can't be observed. One of my collaborators on the CHARA results paper (in press), Ming Zhao, has an excellent paper on Beta Lyrae, a well-known binary star system, in which he imaged both components and matched them to the orbital phase (see http://adsabs.harvard.edu/abs/2008ApJ...684L..95Z). In principle something like this could be done with eps Aur, but the problem is that the disk isn't that bright so it is drowned out by the F-star, thereby making these observations nearly impossible.And I have one point about distance estimates:The current distance estimate for eps Aur is 625 pc (parsec) derived from a parallax measurement from Hipparcos (see http://simbad.u-strasbg.fr/simbad/sim-id?Ident=eps+Aur) which gives us a value of 1.60 +- 1.16 mas (milli-arc-seconds). Note that with these error bars, eps Aur could be 625 +- 453 pc away. Other methods of determining distance put eps Aur at ~650 pc so the 625 pc from Hipparcos is preferred. With the present technology, a minor shift of 10 AU is not a quantity that can be observed at this distance.Keep asking questions!Brian
Thanks Robin, Wofgang and Brian for your kind comments. After read the paper (Struve et al. 1958) recommended by Robin, I had believe that the eclipsing variable model is authentic. The velocity curves presented in the paper is consistent with my Fig. 1 commendably. Hi, Brian, I would like know how long (angle) a star shift is the quantity that can be observed without a reference point. Under the conditions that the orbital separation of esp Aur is 18.1-19.6 AU and the mass of F star is 2.15 times smaller than that of B star, can the shift be observed ? Liang
Hi Liang,I am a professional scientist I am only an amateur astronomer too. There is evidence related to the points you raise that favours the binary model. 1. The radial velocity shows a 27 year variation.The Radial Velocity of Epsilon Aurigae. Struve, O.; Pillans, H.; Zebergs, V. Astrophysical Journal, vol. 128, p.287 http://ukads.nottingham.ac.uk/abs/1958ApJ...128..287S2.The CHARA interferometry data is still to be published but here are the hints Dr Bob has given about the resuts here on the forum http://www.citizensky.org/content/success-its-not-over-til-its-over4. The absorption lines that show red and blue shifts during eclipse are superimposed on a stationary F star spectrum indicating that there is an addition component present during eclipse.Epsilon Aurigae in eclipse. II - Optical absorption lines from the secondary Lambert, D. L.; Sawyer, S. R. Astronomical Society of the Pacific, Publications http://adsabs.harvard.edu/abs/1986PASP...98..389LEpsilon Aurigae. II - The shell spectrum Ferluga, S.; Mangiacapra, D. Astronomy and Astrophysics, vol. 243, no. 1, March 1991, p. 230-238 http://adsabs.harvard.edu/abs/1991A%26A...243..230FBest RegardsRobinwww.threehillsobservatory.co.uk