What Julian Date to use for Eclipsing Binary Observations

American Association of Variable Star Observers (AAVSO)
Wed, 11/05/2014 - 18:19

I just observed my first eclipsing binary (V376 And) and am ready to upload data.  My question is on the recommended version of the JD.  Since it is an eclipsiing binary it makes sense to use the Heliocentric Julian date, but with respect to what time standard?  TT or UT?  Or should I use the Barycentric Julian date?  My inclination is just to leave it in standard JD based on UT and let anyone who uses the data do the conversion.  Is there any AAVSO standard recommended for these types of observations?


American Association of Variable Star Observers (AAVSO)
time details

Good questions, Jim.  Always good to be thinking about the details...

I'll take a stab at this, and Gerry Samolyk (who publishes ToMs, and has more experience) can correct me where I am wrong!

Two aspects to this:  ToMs, and photometry.

When archiving your actual photometry in the AID, use the same recommendations as for other AAVSO photometry.  I.e. JD, or HJD (specify which), in UT.  The WebOBS file format specification doesn't actually say the data should be with respect to UT, but it also doesn't say anything about converting from UT (which is what your computer will be referenced to), to anything else.  

Further on this, data is generally referenced to UTC.  Use of TT is so rare that when Guinan and Ribas 2001 had to use it  (ApJ 546:L43-L47, Jan 1, 2001 in a paper re V471 Tau) they made special note of it, and stated that though not normally mentioned, published times are assumed to be UTC.  Interestingly, even though they were dealing with uncertainties of +/- 5 sec, they still used HJD.

Re BJD vs HJD:  well, the max difference is +/- 4 sec = 0.00005 days.  It would only be in exceptional circumstances that I would believe data uncertainties that small, so although it is being a little sloppy,  BJD and HJD are essentially equivalent.  (I am personally skeptical of the uncertainties reported by the KvW algorithm:  I think they are too small, but it is generally accepted practice to publish them as produced by the algorithm. Mikulasek et al have looked at this... arXiv 1311.0207)  Since e.g. Gerry's publications of ToMs state them as being HJD, that's what you should use.

It is left up to researchers who need exceptional precision to deal with UT - TT, and HJD - BJD conversions.

I hope this helps…  and I welcome further discussion on this topic!

Regards,  Gary BIllings


American Association of Variable Star Observers (AAVSO)
Time Systems for Eclipsing Binary Observations and Analysis

Jim – Your question is very timely (no pun intended) since recent data acquired electronically are far more accurate than the old visual observations.

Gary – I agree with what that you’ve written. I also have my own two cents to contribute. One cent is with regard to published times of minimum and the other is concerning the analysis of orbital periods.

Publishing times of minimum in JD heliocentric referenced to UTC is alright and most people will interpret those results correctly. However, the USNO Time System web page http://tycho.usno.navy.mil/systime.html  suggests explicitly specifying the time system being used, for example “JD heliocentric UTC”.

Orbital period studies are a different matter though. They should be based on dynamical time rather than Earth rotation. Otherwise, O-Cs diagrams will reflect ‘delta T’ which is the difference between the two systems. Delta T changes rather erratically but it has increased by nearly one minute during the past century. The divergence is expected to grow larger as the Earth’s rate of rotation continues to decrease.