Forums / The Science / Data Analysis / Literature Review: A Ring Model for the Disk
Literature Review: A Ring Model for the Disk
As part of my Ph.D. dissertation, I has been conducting a comprehensive review of literature on Epsilon Aurigae. Each month I select one paper to review or discuss in the newsletter, adding relevant information from more recent literature when necessary.

As many observers are reporting that we are near or at totality, I thought I would take a few minutes to discuss what we can expect in the coming months and why visual, photometric, and spectroscopic measurements are very important in this time period. Probably the most exciting portion of totality is the mid-eclipse brightening that was reported in 1983 to be about 0.2 mags. This same brightening was also seen in the 1929 and 1956 eclipses, although to a lesser extent. Although this brightening is interesting, there are also subtle variations during totality that are really worth investigating.
Although there are a few papers that discuss these mid-eclipse variations, few address them in as much detail as Ferluga's 1990 work (see below for complete citation) in which he created theoretical light curves for the system. He compared his models with photometry from several observers and used the V and P-magnitudes reported in the 1956 and 1929 eclipse, respectively, to argue that the disk is changing.
His objective was to investigate the main structural features, not the physical composition, present in the disk. To do so he first starts off with a 100 x 100 pixel model of what the star would look like in a given 10-day interval. Then he moves the disk across the face of the F-star noting the net brightness at a given time. After changing some of the disk's parameters, he repeats this process until a satisfactory fit to the observed light curve is obtained.
In his paper he reports that he modeled the disk as a set of concentric rings, like the rings of Saturn with a large central opening to explain the mid-eclipse brightening. The rings were very opaque, whereas the region between the rings had a changeable opacity. He varied the width and placement of the opaque rings, and the disk's inclination and eventually arrived at the model shown in the figure to the right (note, the numbers listed are for the high-mass model, all radii should be scaled by 0.6 for the low-mass model) in which the inter-ring gaps are fully transparent.
With Ferluga's work, is the structure of the disk solved? I would say no. He investigated if his model could be applied to prior eclipses and found that there was no systematic trend that could account for the changes; hence the difference could not be accounted for by stating that the disk is precessing. He argues that the structure of the disk is itself changing.
Your photometric observations will give us a good idea on the large-scale changes in the disk's structure by modeling the disk in a similar fashion to what Ferluga did. Spectroscopic observations provide even more information, allowing us to probe the disk's composition (see Robin Leadbeater's results).
For more information, see:
Epsilon Aurigae. I - Multi-ring structure of the eclipsing body
Ferluga, S. A&A 1990
http://adsabs.harvard.edu/abs/1990A%26A...238..270F
Heinz,
I am not priivy to all of the details of the model, but it does appear to be quite simple (six parameters, four are radii and the other two are opacities for the regions that contain clearings. I think his ring model is reasonable, it needs to be tested this time around. Of course in 27 years a lot could have changed.
Ferluga has made some really great contributions to eps Aur (and astrophysics in general). He has another paper in which he subtracts the spectrum of the F-star and comes up with a "shell spectrum" for the disk... perhaps I'll review that paper next time.
Brian



Very interesting indeed. I think there is a saying like "computer models are doomed to succeed", and I guess it expresses the danger that if you just add enough degrees of freedom in your model, given enough computing power you can always find a point in parameter-space that will make the model fit the observation data very well. So I guess the quality of the fit is not enough to give credibility to this ring model. But it would be exciting if the model could make predictions for effects that could be measured but were not part of the observation data that was used in the model fitting (not the light curve), right? CS Heinz