Accessibility Tools

Skip to main content

The Hubble telescope discovered a surprising accretion disk around a black hole

22 July 2019

Ari Laor took part in an international research team

that discovered a surprising accretion disk around the black hole in the galaxy NGC 3417, using the Hubble Space Telescope. The research was published in the journal Monthly Notices of the Royal Society.

 

Link to paper published in the journal Monthly Notices of the Royal Astronomical Society

NASA/STScI and ESA press release

Link to image (from NASA/STScI and ESA press release)

 


 Ari Laor

 

Introduction-What are active galaxies?

The stars in the universe (like our sun) are not scattered evenly in space, and as people gather in settlements, the stars in the universe gather in communities called galaxies. As settlements range in size from dozens of individuals to tens of millions, galaxies range in size from several million stars to hundreds of billions of stars (in our galaxy, The Milky Way galaxy, there are about ten billion stars). And just as in cities, which always have a center with a denser population, also in galaxies the density of stars generally rises towards the center.

But beyond having more stars at the centers of galaxies, there is also something completely different there, as discovered about 20 years ago, a huge black hole that has a mass of about 0.1% of the mass of all the stars in the galaxy. Its typical mass is a million to a billion times the mass of a typical star. The stars in a galaxy spread over several tens of thousands of light years (the distance light travels in one year), while the size of the central massive black hole (the area from which matter cannot get out) is tiny. It ranges from just several light seconds to at most a few light hours. The existence of the central massive black hole is revealed by its gravitational pull, which affects stars already from a few dozen of light years from the center. But other than the acceleration of the stars that stray at its direction, and then escape, the black hole usually just sits there in the dark. Rarely (once every ten thousand years) a star strays too close to the black hole, and the resulting tremendous acceleration rips the star, which is mostly swallowed by the black hole. This process is accompanied by a flash of radiation that can be discovered from far away.

In a small portion of galaxies, that are called Active Galaxies, gas is accreted onto the black hole. The gas spirals-in, forms an accretion disk, which emits radiation that can be comparable to the radiation of all the stars in the galaxy, and even up to a thousand times more. The strong radiation illuminates the galaxy and ionizes clouds of gas throughout the galaxy, which produce radiation at certain wavelengths, so called emission lines. These lines have a certain width (they cover a range of wavelengths) that indicates the velocity of the gas. The wider the line – the higher the velocity of the emitting gas. Very close to the center, only a few dozen of light days away from the black hole, which is located at the edge of an accretion disk, there is a wide gas torus that is also illuminated by the disk and emits radiation in lines. Due to the high velocities so close to the black hole, the broad-lines correspond to velocities of thousands of kilometers per second, and the area is called the broad-line region.

 

The problem - is the unified model always correct?

The active galaxies appear in two types; type 1 have the following characteristics (by distance from the black hole), continuum emission emitted by the accretion disk, in the optical, UV, and X-ray, broad emission lines from the broad-line region, and narrow emission lines from the gas in the galaxy (these lines are narrow due to relatively low velocities in the galaxy, a few hundred kilometers per second). In type 2 active galaxies only the narrow lines are observed. The broad-lines and continuum emission from the accretion disk are missing. Over the years it became clear that type 2 active galaxies are actually the same as type 1, and the difference is just a matter of viewing angle. Close to the center there are large amounts of gas and dust, distributed in the form of a torus. In a galaxy which is observed close to the plane of the torus (that is, observed from the side, rather than from above), the torus obscures the inner regions, including the accretion disk and the broad-line region. The same galaxy will appear as a type 2 active galaxy. A galaxy with face on (top) view will appear as type 1 active galaxy, since all areas can be seen.

In recent years, however, there have been cracks in the unified model. Active galaxies have been detected without broad lines, but with X-ray radiation, which is emitted very close to the black hole, indicating that there is no obscuration of the internal regions. In other words, there are probably two different types of active galaxies, those with a broad line region, and those where the broad line region is truly missing. Can the gas torus in the broad-line region disappear? If yes, Why? If we believe that we understand the production mechanism of the broad-line region, we must inevitably understand when does this region ceases to exist.

 

Research at the Technion

In Dr. Jonathan Stern's Ph.D. thesis under my supervision, that focused in part on the properties of active galaxies in low luminosities, we found that there is a high correlation between the intensity of the X-ray emission and the broad-line emission. In an article from 2012 we studied the expected emission of several galaxies identified as type 2 without a broad line region, and noted that the lack of broad-lines may stem from the low luminosity of the emission lines and the resulting difficulty in their detection.

In an article from 2017 of Prof. Stefano Bianchi of the University of Rome, Italy, and his associates, on the Galaxy NGC 3417, the researchers noted that this is the most likely case of a type 2 galaxy without obscuration, based on new X-ray observations that showed no sign of absorption. Prof. Robert (Ski ) Antonucci from the University of California in Santa Barbara, who proved unequivocally (about 30 years ago) that type 1 and type 2 galaxies were indeed the same objects, questioned their claim based on our 2012 work, and sent Stefano's team and us an email asking what is their opinion on our claim that there might be a broad-line hidden in the spectrum but is difficult to detect.

In the discussion that emerged, I raised the argument that the distance of the broad-line region from the black hole depends on the intensity of the continuum emission from the accretion disk, and since this galaxy is of low luminosity, the broad-line region is likely very close to the black hole (a distance of about one light day, compared to a typical distance of dozens of light days). Consequently, the rotation velocity is higher, tens of thousands of kilometers per second, instead of the few thousand kilometers per second estimated by Stefano. Hence the emission lines are expected to spread over a large range of wavelengths and will be more difficult to detect. On the other hand, Stefano may be right and the broad-line region truly does not exist. The origin of the gas torus in the broad-line region is most likely the outer accretion disk, and based on theoretical work it is indeed expected to cease to exist in very low luminosities, and with it the broad-lines will disappear. In my work with Stern in 2012 we did not claim the broad-line region exists, but that the existing observations are not accurate enough to prove it is not there.

In my experience, a scientific confrontation ends with each party remaining in its side, and therefore email exchanges generally do not lead anywhere. This time Stefano surprisingly said, how can I test that your proposed explanation is right? For me and for Ski (Antonucci) the answer was clear, by using the Hubble Space Telescope. Hubble has a high angular resolution (one tenth of an arcsecond) compared to ground based telescopes that observe through the atmosphere (limited to 1-2 arcsecond), and therefore can reduce the contribution of the galaxy by a factor close to 100, focus on the weak emission from the active nucleus, and then identify clearly whether or not the broad-lines exist.

Stefano asked us to collaborate with him and submit a proposal asking for Hubble Space Telescope observing time, along with his other collaborators, who have experience in data reduction from the telescope. In my personal experience, proposals for observation on the Space Telescope tend to be a waste of time, because of the very low percentage of approval, so I stopped submitting proposals many years ago. Stefano specializes in X-ray telescopes and was happy to try to and submit also to Hubble. To our surprise, the proposal was accepted for observations.

 

The results - three surprises

The observation focused on the spectrum of the galaxy. To do this, we used the narrowest slit of Hubble that completely eliminated the contribution of the galaxy and showed a clean spectrum of the active nucleus in the galaxy. A huge and broad line immediately popped up. The broad-line region indeed exists. The unification theory still applies. When you see the continuum source you see the broad-line region as well. As the lines become very broad they just become harder to detect.

Another surprising result was the line profile (the velocity profile of the gas). The base of the line indicated speeds of about 30,000 km/second (one tenth of the Speed of light), fitting preliminary estimates of 20-40,000 km/second. But the line profile was very asymmetric, with a steep shoulder at shorter wavelengths (blue shift), and a very moderate-shoulder at longer wavelengths (red shift). I derived a similar profile in calculations done about 30 years ago (1991) of the line profile from gas in a thin disk moving very close to a black hole, as a result of a combination of the Doppler effect (special relativity) and gravitational redshift (general relativity). The overall effect gives a clear signature of an asymmetric profile with a specific structure.

In principle an asymmetric profile can also be produced when the gas distribution is not uniform, or when the radiation is scattered on its way to us by free electrons. Surprisingly, we got a perfect fit to the observed profile using the thin disk models. The broad-lines clearly originate from a thin relativistic disk close to the black hole.

The third surprise came from the measured distance of the emission line region from the black hole. From the known relation between the luminosity of the active nucleus and the size of the broad-line region, we got that the size in this weak active nucleus is one light day. Given that the black hole mass is approximately 300 million solar masses, this distance corresponds to 77 ± 15 times the radius of the black hole. In addition, fitting the line profile allows direct measurement of the size of the region, in units of the black hole radius, without a need to know the black hole mass and the luminosity of the active nucleus. The distance received by this independent method is 62 ± 16 times. In other words, we received complete agreement between the two independent measurements. This result is particularly surprising because the relation between the luminosity of the active nucleus and the size of the broad-line region is measured for galaxies that are much brighter, in which the size of the region is about 10,000 times the radius of the black hole. Although the broad-line region in the galaxy NGC 3147 Is less than 100 black hole radii, and the area is located deep within the accretion disk, the relation between the brightness and the size of the area still exists.

The luminosity from the active nucleus in this galaxy is only a mere one thousandth of that in a typical active galaxy, with a black hole with similar mass (about 300 million of the sun's mass). At such low luminosities, the conventional theoretical assumption was that the accreted gas is dilute and very warm, with a low cooling efficiency, and as a result creates a somewhat spherical cloud that is accreted onto the black hole almost without emitting any radiation, instead of creating and accretion disk. Since the source of the broad-line region is the accretion disk, if the disk disappears, the broad-lines will also disappear. The observations showed that the broad-lines do exist, and moreover, that they originate from gas in a thin disk close to the black hole, which is the accretion disk itself. The accreted gas is therefore dense and relatively cold, and creates a thin accretion disk.

Lines from relativistic gas in disks very close to the black hole (a few times the black hole radius) have been observed in X-ray telescopes. The huge advantage of measuring relativistic emission lines using the Hubble Space Telescope in the visible part of the spectrum is the large number of photons, at least 100,000 times more than in the X-ray. The successful attempt with NGC 3147 opens the door to try and observe other similar galaxies, in particular galaxies in which a smaller radius is expected. The line profile will then allow us to get the disk structure in the innermost region, and perhaps even allow us to measure the rotation ("spin") of the black hole. Our proposal for follow-up observation using the Hubble Space Telescope on this galaxy has recently been approved. Stay tuned!