HVGC-1: The outcast cluster

We’ve already discovered stars moving at thousands of kilometers per second. Is this globular cluster doing the same?

To the naked eye, stars and galaxies often look just like tiny pinpricks of light. While for many sources, telescopes can provide images in almost startling detail, there are plenty of cases where the target of interest still looks just like on a dot on a screen. In those cases, astronomers may be reliant on spectroscopy to give them even basic insights into just what it is they’re looking at.

Take, for example, the object HVGC-1, located about 16 million parsecs from Earth in the Virgo Cluster of galaxies. When it was first discovered in 2014, its position in the sky meant it could either be a star in the Milky Way’s halo or a rogue globular cluster being ejected from the elliptical galaxy M87. Radial velocity measurements seemed to indicate that the latter hypothesis was more likely, but it took optical and infrared photometry, accompanied by measurements of hydrogen lines in its spectra, to finally close the case.

A Hubble Space Telescope image of M87, along with a jet emitted from its center. Image credit: NASA/Hubble.

HVGC-1 — short for High-Velocity Globular Cluster 1 — is interesting as more than just a demonstration of why spectral line measurements are incredibly useful. The object is moving away from M87 at a speed of over 2300 km/s, greater than the galaxy’s escape velocity. This is unparalleled by any globular cluster discovered to date, and seems to indicate a turbulent past for both the cluster itself and the galaxy it once called home.

The culprit? Quite possibly a supermassive black hole.

Spectroscopy strikes again

The discovery came in the midst of a search (Caldwell et al. 2014) for globular clusters towards the Virgo Cluster. Old, dense, and full of low-metallicity stars, these objects and the stars that make them up are likely as old as the galaxies they orbit. Their early formation and evolution is still an active topic of research, and while our best and most detailed samples of globular clusters are within the Local Group, studying them in other galaxy groups — like M87’s — can help us better understand some aspects of galaxy formation.

Fig. 1, Caldwell et al. 2014. The peaks of the star and globular cluster distributions are clear, and HVGC-1 is far from both of them.

The astronomers studied approximately 2500 globular cluster candidates and were able to determine redshifts — and therefore radial velocities — for about 1800 of them. The velocity distributions showed that the candidates actually fell into three distinct categories: galaxies, globular clusters in the Virgo Cluster, and foreground stars in the Milky Way’s halo. When the galaxies were removed from the sample, one data point remained that was far from the peaks of the star and globular cluster distributions, with a radial velocity of about -1000 km/s — an enormous outlier. This raised a question: Was HVGC-1 a star, a galaxy, a globular cluster, or something else entirely?

Spectroscopy allows us to determine an object’s motion along our line of sight, but spectral lines themselves provide us with even more information. For instance, the presence and shape of various lines can be a key indicator of a star’s composition and temperature. The strengths of lines is also important, but line strengths can be decreased by things like extinction and, of course, the inverse square law. Therefore, if you want to try to identify an object based on how strong its lines are, it’s better to look at ratios of different lines, rather than absolute intensities.

Fig. 3, Caldwell et al. 2014. The HVGC-1 is pretty clearly a globular cluster.

In this case, the team looked at two different line ratios: Hγ divided by the G-band luminosity, and a ratio of two types of Ca II emission, utilizing higher-energy Balmer lines. In general, M87’s globular clusters have substantially lower Ca II ratios than either globular clusters in Andromeda or field stars. HVGC-1’s Ca II ratio was about 0.5 — roughly half of what you might expect to see in a halo star, but perfect for a globular cluster.

Filtering out the stars

The other major diagnostic used to separate the object from possible foreground stars was simple photometry. In a survey of the Virgo Cluster released earlier that year, called The Next Generation Virgo Cluster Survey-Infrared (NGVS-IR), Muñoz et al. 2014 had figured out a clever way to distinguish globular clusters from stars by comparing photometry from different filters. Taking luminosity measurements in the K, u, and i bands (centered at 2190, 365, and 806 nm, respectively), they plotted i-k against u-i, and identified a number of regions in what’s called a uiK diagram, including lines of stars and globular clusters.

Fig. 14, Muñoz et al. 2014. The two neighboring lines of main sequence stars and globular clusters are very densely populated.

The NGVS-IR team had been specifically studying M87, making their discovery a really useful tool for Caldwell et al., who were looking for some more evidence that HVGC-1 was a globular cluster. Their uiK diagram did not disappoint, and the target object fell beautifully into the globular cluster sequence, as expected.

Fig. 2, Caldwell et al. 2014.

Between the photometry and the line ratio analysis, it seemed that the group had definitively classified HVGC-1 as a globular cluster, moving at the incredible speed of of 2300 km/s relative to M87. A quick calculation showed that this speed is much greater than the galaxy’s escape velocity, indicating that the globular cluster will soon leave it, as well as the Virgo Cluster as a whole. We do see a number of stars moving like this in the Milky Way called hypervelocity stars (I wrote about an interesting case, HE 0437–5439, earlier this year!). These stars are thought to have reached their immense speeds after encounters with Sagittarius A*, the supermassive black hole at the center of the galaxy. Could something similar have happened with HVGC-1?

Black holes and baby clusters

The idea of a violent encounter with a supermassive black hole is a tempting one that would imply a violent history for M87. For an object to be ejected from a system, there must be at least three bodies involved. Hypervelocity stars are often stripped of binary companions — the third bodies involved — when they meet a supermassive black hole, but that binarity is highly unlikely for globular clusters! The astronomers proposed a modification to the setup, where two supermassive black holes spaced no more than a couple parsecs apart could have provided the necessary push to send HVGC-1 into intergalactic space. Such an encounter would strip away most of the globular cluster’s stars, leaving only its dense core.

There’s support for this idea. While M87 only holds one supermassive black hole, it’s offset from the center, which could indicate that the black hole is the result of a binary merger. The gravitational waves from the event might be emitted anisotropically, propelling the remnant away from the galactic center. We have evidence that the same thing happened in the quasar 3C 186.

Hypervelocity stars like HE 0437–5439 can be remnants of binary or triple-star systems. Image credit: NASA.

That said, as is almost always the case, there are other explanations for the peculiar velocity. One possibility the group considered is interactions between galaxies in the Virgo Cluster could have been stripped from M87 — HVGC-1 among them. However, the observed distribution from other candidate objects in M87 doesn’t extend as far as the measured relative velocity of 2300 km/s, meaning HVGC-1 would still be an outlier. Another possibility was that HVGC-1 is the companion of a “subhalo” dwarf galaxy orbiting M87, displaced through a three-body interaction. Again, though, this would be unlikely to generate the required velocity.

Finally, the team considered another situation related to supermassive black hole binaries. Given how dense HGVC-1 is, it seemed possible that it was actually what’s called a hypercompact stellar system — a supermassive black hole surrounded by a group of stars, all ejected from a black hole binary system at the center of the galaxy through gravitational wave emission. However, the measured metallicity of the cluster disagreed strongly with theory — and, again, so does its velocity.

The astronomers still favor the idea of a supermassive black hole binary merger — an exciting prospect, as those mergers might be detectable by systems like pulsar timing arrays. If we are indeed observing M87 long after such an event, there might be other signs in the stellar populations close to the galaxy’s core. Perhaps the astrophysical jet emerging from its center could give us some clues. M87 is an absolutely enormous galaxy, likely many times the mass of the Milky Way, with rich groups of stars, globular clusters, and gas and dust. It’s a constant target of study at a wide variety of wavelengths. We might not be too far from the next survey to provide evidence as to the curious past of HVGC-1.