The black hole in the Bright Star Catalog

Sometimes, slow and steady wins the race, even in modern research, and sometimes progress is not so steady and interrupted by extended periods without new insights. How it took twenty years to realize that we had a black hole in our data archive.

Like 1 Comment
Read the paper

For us, Dietrich Baade and myself, the story began back in 1999 when the FEROS spectrograph was installed and commissioned at the 1.52m Telescope on European Southern Observatory’s (ESO’s) La Silla Observatory. The instrument-building team received "guaranteed observing time", and we could use part of this time for a little survey of bright stars of our interest, the so-called Be stars.

Sunset panorama at La Silla
Telescopes on La Silla at sunset (Credit: ESO/B. Tafreshi, twanight.org)

PhD student Monika Maintz had a first look at the spectra of HR 6819.  She recognized it as the signature of stars in close orbit, because the Doppler shift of some spectral lines  varied for one of the two stars by a large amount. However, the available spectra were not enough in number and well enough distributed in time to find the orbital period.  So more observations were done from La Silla in 2004 over the course of three months, where FEROS had meanwhile been moved to another telescope, the 2.2m of the Max Planck Society.

Of course, we could not leave our duties and spend months in the desert just to make a 20-minute observation every other day. At the time, ESO had introduced service observing at its observatories. Ground-based observers did not have to do their observations themselves anymore, but could rely on observatory staff to carry them out. ESO’s Director General, later Nobel laureate Riccardo Giacconi, had developed and realized the vision that, for much improved observing efficiency, ESO provide its community with similar services as the Hubble Space Telescope does.

Our friends and team members Stanislav (Stan) Štefl and Petr Hadrava had a closer look at these new observations obtained in 2004, and he found an orbital motion of a binary with a period of 40 days.  It is not that uncommon, more than one half of such massive B-type stars are in binaries.  But the second star (the Be star) did not participate in this motion. One star had to have a dance companion, but the other star could not be it.

Trailed spectra showing no trace of a companion star

The smoking gun of a third, invisible object: One signature is moving, the other is not

Petr’s  expertise lies in the disentangling of spectra of multiple stars that are not resolved in images, exactly the problem we had.  The results were clear:  The Be star had not changed velocity in 5 years, yet the other star moved back and forth every 40 days. There was a moving body of at least five solar masses, and Kepler's and Newton's laws are strict: To let that object swing with a velocity of 60 km/s and a period of 40 days takes another object of similar mass.  But only two such stars showed up in the results, there was no trace of a third. The implication was tantalizing: the invisible body should be a black hole. 

This was perhaps a too spectacular result, especially for an extremely modest person like Stan, and so he decided to put in on the back burner, while concentrating on the challenge of moving to radio astronomy and working for the ALMA observatory. He wanted to finish the project after his return to the Czech Republic. But just weeks before that, the project came to a full stop with Stan’s death in a car accident in Santiago de Chile in 2014.

 Stan Štefl in the mountains he loved.
Stan Štefl in the mountains he loved.

At the end of 2019, Nature published an article on a system named LB-1.  The study interpreted the observations in the frame of a binary model with a black hole of nearly 70 solar masses, in significant tension with conventional stellar-evolution models.

Discussing it in the Journal Club, ESO research fellow Marianne Heida drew our attention to this study, and it was instantly clear that the spectra of LB-1 included the tell-tale signature of a Be star. And it finally dawned on us: The similarity of HR 6819 and LB-1 was just stunning. We quickly put back together the team, including also Robert Klement, a former student of Stan, and worked frantically on making sure we did not overlook something. But the minimum companion mass of at least four times that of the Sun still required a black hole as the explanation of the data, there was no trace of anything else that could provide that pull of gravity.  And the system was the closest to the Earth of well-demonstrated black holes of all types.

The scientific account of the story is dedicated to Stan, in deep gratitude, respect, and friendship.  We shall continue to miss him.  



Go to the profile of Thomas Rivinius

Thomas Rivinius

Science Operations Astronomer, European Southern Observatory

No comments yet.