I'm reaching out to the AAVSO EB community in order to find someone who is experienced in using Binary Maker 3. I've recently purchased it and I'm having difficulty getting up and running. Are there any online tutorials? Any help is greatly appreciated.

Chuck

CCHD

Hello Chuck

Twice I have used BM3. I know of no videos on line. I would suggest following the manual, and running one of the canned examples. I found it very difficult to get up and running, so I can emphamise with you. I am no expert, and mostly just hunted and pecked for a few days, and was able to figure it out.

Gary

Thanks Gary. Unfortunately, the manual is not clear on how to work from a light curve to model the EB when there's not much specific data about the two stars, e.g., Effective temps, limb darkening, etc

I'd suggest you start with your phase curve. Its shape will give you an idea of the type of binary system you are studying. The section of the manual on light curve analysis gives an overview. There is also an extensive library within BM3 of various types of EBs, so you can look at your own phase curve, estimate what type of EB you have, go to the library for that type, and have a look at some of the examples.

For effective temp, find the spectral type of the system. That will give you some idea. Also, if you have colour index data of the system you are studying, if the index varies througout the cycle, you may have significantly different effective temperatures of the two stars. If the index is nearly constant, the effective temps may not vary much. Note that it is the difference in effective temperature of the two stars that is important in BM3 modelling, not so much the specific temperatures.

For limb darkening, there are tables at the end of the manual.

I suggest you start with the section of the manual on light curve analysis, and in particular light curve analysis tips. It is also instructive to display the models in the BM3 library, and 'play' with the data to see to what extent changes in particular parameters affect the model. For example, the models are very sensitive to changes in inclination (in the Observer tab) but less sensitive to changes in effective temperature (unless the latter changes are quite substantial).

Once you have a model, and you have to tweak the parameters to improve the fit, use the LC Residuals window to close in on the 'best' model asymptotically. The number in the bottom right of the window is the RMS of the fit. Tweak the parameters to minimise the RMS.

Roy

Hello Charles,

Modelling EBs is a big topic. My experience level is not very high, but here are a few ideas.

You might check out papers by Wilson, e.g. IAPPP Comm. 55, pp 1-20 (1994), and PASP 106(703) p 921ff (1994). I especially recommend the former, where he was specifically aiming to provide how-to advice to a group of serious amateurs. The kinds of "rules" he lays out on page 16 are especially valuable. I found it useful to work out for myself why those rules were true. They provide a powerful starting point for making a model.

The book by Kallrath and Milone (2nd ed'n, 2009) is useful, but is so extensive as to be quite daunting. Their little section on "Estimating Initial Parameters" pp 198-202 is similar to Wilson's section mentioned above.

If you have any colour information (B-V) use emprical relations (e.g. Allen's Astrophysical Quantities section 15.3.1 to get temperatures, remembering that you actually have two stars and are observing their combined light.)

Chapter 11 of the BM3 manual is useful. See the section ' Light Curve Analysis "Tips" '. (well, the whole manual is useful and there is a lot of info there!)

The parameters you are solving for are NOT orthogonal. I.e. several different parameters will affect, e.g., the depth of eclipse -- and each of those also affects other lightcurve properties in different ways. Thus modelling is a bit tricky. It would be great if BM3 had an "undo" function, or a facility for comparing light curves from different parameter sets, but it doesn't. I find myself keeping careful notes about what aspect of the LC changed when I changed a certain parameter, and take screen grabs or actual data dumps of the light curve for compiling and plotting elsewhere.

With BM3 you can get >>> A <<< lightcurve out very easily and quickly, but when I am going to seriously try to model a lightcurve (and I'm talking about something that is still pretty crude by professional standards), I plan to settle in for several days of focused work. This is partly because of the way parameters interact: even with my notetaking etc, I really have to build up in my head, a sense of how they are interacting.

I hope this helps a little,

Gary Billings

Hello Charles,

As Gary mentioned in his post modelling the eclipsing binaries is a large topic.

During this adventure you should learn a lot and gradually you will get new knowledge. To begin with, you need to prioritize, create the right sequence of work and to answer the question: what I am expecting from modeling, what I expect to get and where I should stop, and where I could publish my results so to be helpful for the astronomical community?

For scientific results in the future you would use the PHOEBE code or Wilson-Devinney code but as the beginning Binary Maker works fairly well.

The main stages:

I. Photometric stage.

II. Obtaining other physical characteristic and preliminary stellar parameters of the target.

III. Light curve solution and obtaining the necessary stellar parameters.

IV. Estimating the absolute physical parameters of the eclipsing binary stars – primary and secondary components.

Very briefly, a sample workflow could look like this:

I. Photometric stage.

1. Obtaining the photometric light curve (LC) of the appropriate target (EW star) minimum in two passband (Johnson-Cousins or SLOAN system). It depends of the used equipment or your interest.

2. The photometric data should be corrected from JD to HJD

3. Preferably to estimate your own epoch in HJD (the primary minima of your LC). The convention is that the primary star is eclipsed during the deeper (primary) minimum.

4. Take from literature the orbital period of EB star (p in days).

5. If you are familiar with ToM (Time of Minima) and O-C (observed minus calculated) procedures and if you can obtain from literature measured minima during several years you can get the improved ephemeris of your target (epoch and period).

6. Using the initial epoch and period create the folded (phased) LC.

7. Save the folded LC in three columns: HJD – Magnitude – Magnitude Error (uncertainty) in the necessary format (.txt or .dat)

II. Obtaining other physical characteristic and preliminary stellar parameters of the target.

8. Finding the mean target temperature Tm (Teff mean): there are many, many catalogs and all sky surveys in different passbands that give opportunity to estimate the color index and the approximate temperature and spectral type of the target: (B-V), (J-Ks) and more.

9. Calculating of the initial temperature of the secondary star T2 using some empirical relations.

10. Calculating of the initial mass ratio q (q = M2/M1) using empirical relations as for example: (q=T2/T1)^1.7;

III. Light curve solution and obtaining the stellar parameters.

11. Take values of T2 and q from the preliminary solution.

12. The coefficients of limb darkening (LD), gravitation darkening and reflection correspond to the initial temperatures of the stellar components.

13. Linear LD law is appropriate at the beginning.

14. The guessed value of the orbital inclination i (deg) is suspected from the eclipse depths.

15. The fitting mode (detached, SD, contact, OC) is guessed from the light curve.

16. Initially only "primary level” should be varied until observed and synthetic curves roughly coincide. Unfortunately Binary Maker do not provide "the standard error of the estimate" i.e. standard deviation or “chi-square distribution” (χ2-distribution).

17. Vary consequently each of the parameters as the secondary temperature, orbital inclination, mass ratio, and potentials "omega" until synthetic curves coincide to original LC.

18. If the observed light curve is asymmetric and/or distorted add surface spots and vary their parameters..

During this stage use the following rules:

(a) The bigger widths of the eclipses could be reached by increasing of the relative radii;

(b) Simultaneous increase/decrease of the depths of the two minima may be provided by increase/decrease of the orbital inclination;

(c) The increase/decrease of the relative depth of the secondary minimum can be provided by increase/decrease of the temperature of the secondary star;

(d) Spots with longitudes around 90◦ and 270◦ change the light levels at quadrature (O'Connell effect) while spots with longitudes around 0◦ and 180◦ change the shape and depth of the corresponding eclipse. The longitude of the spot center corresponds to the maximum distortion of the light curve.

The obtained final values of your modeling will be the following parameters: T1 (primary star temperature), T2 (secondary star temperature), i (inclination), q (mass ratio), the potentials Omega1 and Omega 2, relative radii r1 and r2. Than you can calculate the final luminosity ratio l2/l1.

IV. Estimating the absolute physical parameters of the eclipsing binary stars – primary and secondary components.

Here you should jump into the lots of possibilities and different procedures (some empirical and statistical, some pure physical) which allow you to estimate the absolute physical parameters of the eclipsing binary star components which is the most important thing we want to achieve:

T1 and T2 (the components temperatures, Kelvins), M1, M2 (the component masses in solar masses), R1, R2 (the components radii in solar radii), L1, L2 (the stellar luminosities in solar luminosities), a (the orbital axis), d, pc (the distance to the EB in parsecs).

Hopefully this explanation will give you the right guidance on this long journey.

Best regards,

Velimir

P.S. Some of my publications dedicated to eclipsing binary stars that could be helpful you can find in my web site.