How to correctly calculate HJD?

Dear all,
I am confused as to the correct calculation for HJD form JD. Both are of value in the discussion of EB-observations, but HJD is particularly important in O-C diagrams. I have used BAA Computing Section and tried to brew a Python notebook based on Dogett&Kaplan in Henden &Kaitchuck Astronomical Photometry p113. They are not in agreement. Would it not be advisable to have an option in the AAVSO site to calculate HJD from JD, RA, Dec ?
Yours truly,
Serge

1 Like

Hi Serge,

Unfortunately, doing math on date-time objects is deceptively difficult, especially when implementing transformations between reference frames. In Python, I exclusively use astropy.time for light travel time corrections for these types of calculations. In C/C++, I often used the Standards of Fundamental Astronomy libraries.

The great thing about these tools is that they are already very well tested.

Brian

Serge,
You are absolutely correct about using HJD (or BJD) for working with O-C diagrams for short period variables. I’ve used the BAA website and I’ve trusted it for research work. I’ve compared to the results from BAA to those from the IRAF setjd command and from corrections in Persanso and find them to match. As Brian says it is a hard thing to calculate since there are a lot of moving parts in the calculation.

Eric

Greetings,

I’m curious as to the amount of the difference between BAA and Doggett you got for HJD?

Lew’s JD Sun Time spreadsheet is ok. In one of the cells he states that he estimates that values are probably good to “3 seconds”.

I have used for a long time Dr. Dan Bruton’s excel spreadsheet to convert JD to HJD. Below is the link to his online calculator and the link at the bottom will get your his excel. Inside his macro he states that his method is good to 4-seconds of time. Both note that there is no consideration in their equations for the difference between the heliocenter and the barycenter.

https://www.physics.sfasu.edu/astro/javascript/hjd.html

Running the example that came with Lew’s JD Sun Time excel and Bruton’s excel the difference between the two spreadsheets is 1.7-seconds, 0.00002-days.

Strictly speaking observers are not observing from the geocenter but from somewhere on the surface of the Earth this position is called a topocentric place. Also for really high precision and accurate timing the solar system barycenter (BJD) is the correct reference to use. However, for most objects we amateurs observe HJD should be good enough.

p.s. What ever method you use you should specify it. HJD, BJD and what software you used to compute it.

Jim (DEY)

The other consideration for variable star astronomers is what type of event is being timed, and how accurate the timing can be, as opposed to how closely a computer clock matches a time standard.

For example, I try to have my computer set to within 1 second of real time, but don’t agonize over it, because the only times I measure are the peaks of light curves of short period pulsators (delta Scuti stars), and the times of mid eclipse for EBs. If I think my times for such events have an error of no more than (say) 20 seconds (although for some light curves it may be a larger figure), the precision on O-C diagrams is good and I am happy. The accuracy of the computer clock is clearly not the limiting factor, but of course one wants an accurate HJD conversion.

Roy

Following up on Jim’s (@DEY_VAR) post, AstroImageJ (AIJ) has a very nice calculator for the various JDs, including BJD. AstroCC (Coordinate Converter) resides under the Data Processor tool. AstroCC will write the various coordinates directly back to the FITS header of the image file.

AIJ is freely available. There’s a learning curve, but not too steep.

George (IGE)

And, following up on @Roy_Axelsen_ARX post. I believe the time stamp in a FITS header ultimately derives from the computer controlling the imaging camera. I use Dimension4 to keep my imaging computer on-time. Dimension4 uses the Simple Network Time Protocol (SNTP) to adjust the computer’s clock, but after reading Dale Ghent’s How Mounts Keep Time, I think the better way is to enable Network Time Protocol.

I am chasing exoplanet transits; perhaps this is overkill for variable star timing.

George (IGE)