Gliese 372 : A Nearby Binary Star System

The story of how I discovered Gliese 372's binary nature is an illustration of just how little we know about the nearest 25 parsecs. It was known since 1958 (Vyssotsky 1958, Astronomical Journal, v.63, p.211) as a typical 11th magnitude M dwarf, with a colour-inferred distance of 10 parsecs. As an undergraduate studying Physics and Astronomy, I was doing my B.Sc. Thesis project at the David Dunlap Observatory at the University of Toronto. I was using the 2-metre telescope there to measure radial velocities of a sample of M and K dwarfs from the Vyssotsky catalogue, with the aim of combining the measurements with proper motions to get space velocities of the sample.. Normally, to check for consistancy, I obtained 2 or 3 spectra of each star, averaging the radial velocities. By Feb 1994 I had three spectra of Gliese 372, each taken about 2 weeks apart, and each with completely different radial velocities: +5 km/s, +17 km/s, and +54 km/s. The instrumental uncertainty was about 1 km/s, so I was a little shocked, and at first I thought I might have been observing the wrong star two of those three times.. But the spectra looked identical, except for the RVs.

It turned out Gliese 372 is a binary system of two M dwarfs. I was seeing the doppler shift of the brighter component at different times in its 48 day period. I took 39 more spectra over two seasons, and published my analysis in Harlow (1996, Astronomical Journal, v.112, p.2222). Careful spectral analysis revealed the spectra of two M dwarfs, with apparent magnitudes of 11 and 12, in a system at a distance of 17 parsecs. So the system is still within the 25 parsec limit of the Gliese catalogue, and my discovery adds one more known star to that catalogue. Also, Gliese 372 is one of only nine known M dwarf double-lined spectroscopic binary systems and it is the only one which does not have unusually high amounts of magnetic activity. This makes it an interesting test-bed for theories of normal M dwarfs, the lowest mass main sequence stars.

Why was this not discovered until 1996? The average separation of the components is about 0.26 AU, and at a distance of 17 parsecs, the maxium angular separation on the sky would be about 0.02 arc-seconds. So the only way to know about the binary nature of the system is with medium or high resolution spectroscopy. It turns out that detailed spectral analysis of many of the nearest and brightest stars just hasn't been done! We spend most of the big telescope time looking at faint galaxies and distant quasars, while we still know very little about our own small neighbourhood of this Galaxy!

Future possibilities? It is possible that Gliese 372 is an eclipsing binary system. From spectral data, I compute a 25% chance that this system eclipses. Eclipses would last 2 to 4 hours, once per 48-day period. This is difficult to observe, and it is even more difficult to do a photometric analysis which completely rules out the chance of eclipses, since the ephemeris is getting a little old. But if it were eclipsing, it would be the only system of its kind, offering information on temperature, mass and radius, and it would be an excellent testing ground for theories of M dwarfs.


Here is an optical spectrum of Gliese 372 taken at the David Dunlap Observatory.


Here is the cross correlation function of Gliese 372 against the single star template of HD 36395, plotted for three different representative orbital phases. The cross correlation function is displaced vertically by X, where X=1.0, 0.5 and 0 for phase=0.1, 0.5 and 0.94, respectively. The secondary peak is marked by arrows and phases 0.10 and 0.94, and is seen to disappear behind the promary peak at phase 0.50. This is direct evidence for the presence of two sets of absorption lines in the spectra.


Here are Heliocentric radial velocities of each component of the system plotted versus orbital phase. The solid circles are velocity measurements of the primary, or Star 1, and the open squares are velocity measurements of the secondary, or Star 2. The solid lines correspond to the orbital solution with elements as given in the Table below.

P = 47.709 $\pm$ 0.053 days  
T0 = 2449345.44 $\pm$ 0.52 HJD  
$\omega_1$ = 80.8 $\pm$ 2.0 degrees  
e = 0.530 $\pm$ 0.016    
K1 = 30.24 $\pm$ 0.59 ${\rm km\ s^{-1}}$  
K2 = 40.0 $\pm$ 1.5 ${\rm km\ s^{-1}}$  
$\gamma$ = 31.34 $\pm$ 0.37 ${\rm km\ s^{-1}}$  
             
$M_1 \sin^3 i$ = 0.594 $\pm$ 0.045 $M_\odot$  
$M_2 \sin^3 i$ = 0.449 $\pm$ 0.035 $M_\odot$  
M1/M2 = 1.323 $\pm$ 0.042    

$a_1 \sin$ i

 = 16.82 $\pm$ 0.38 $\times 10^6$ km  
$a_2 \sin$ i = 22.25 $\pm$ 0.87 $\times 10^6$ km  
$a\sin i$ = 39.06 $\pm$ 0.95 $\times 10^6$ km  

           




Contact Information:

Email: jharlow@pacific.edu

Prof. Jason Harlow
Department of Physics
University of the Pacific
3601 Pacific Ave.
Stockton, CA 95211

Tel: 209-946-3130
FAX: 209-946-3131

This page is maintained by Jason Harlow. Any comments or criticisms are welcome.