An analysis of a sample comprising some 20,000 mass-accreting supermassive black holes, known as quasars, shows that most of the diverse properties of these cosmic beacons are explained by only two quantities. See Letter p.210
If a picture is worth a thousand words, then a spectrum can be worth a thousand pictures. That is perhaps an underestimate when dealing with star-like blobs of light that look fuzzy even through the world's largest telescopes, as is the case with quasars. First recognized more than five decades ago as counterparts to radio sources1, these extremely energetic entities are supermassive black holes in the nuclei of distant galaxies2. The black holes themselves do not emit light, but their gravity accelerates gas into swirling accretion disks that can outshine the galaxies they dwell in. Determining the physical properties of these systems from spectroscopic observations is challenging. But a study by Shen and Ho3 reported on page 210 of this issue accomplishes that elusive feat in the clearest way so far, using not one quasar spectrum, but the spectra of more than 20,000 quasars.
Visible light from a quasar has two main sources: a continuous spectrum emitted by the hot accretion disk, and discrete line emission from gas clouds that orbit the black hole and disk and that are ionized by the disk's intense radiation. The emission lines reveal information about the local environment. In particular, intensity ratios of emission lines due to different levels of ionization depend on the characteristics of the disk's radiation field. Many of these spectral properties are correlated in systematic ways4 (a set of correlations referred to as 'Eigenvector 1', or EV1), which suggests that they are driven by one fundamental physical parameter: the Eddington ratio. This is the ratio of the luminosity of the quasar to that of a black hole that has maximal gas accretion — a limit reached when the quasar's radiation pressure balances its gravity.
Another major property of the emission lines is their width, measured as the full-width at their half-maximum (FWHM) intensity. The lines are broadened by the Doppler effect: gas in the line-emitting clouds moving away from Earth emits light shifted to longer wavelengths (redshifted), and gas moving towards us emits light shifted to shorter wavelengths (blueshifted). The FWHM of broad lines, in particular that of the Hβ hydrogen line (FWHMHβ), provides the component along Earth's line of sight of the gas's orbital velocity. The gravitational field associated with supermassive black holes, which are millions to billions of times more massive than the Sun, sets those velocities, and so measurements of FWHMHβ help to determine a black hole's mass.
However, measurements of FWHMHβ depend on how the quasar is tilted relative to our perspective. The large-scale jets that quasars emit, and which are seen in radio observations (Fig. 1), permit a determination of the jet's orientation, together with the quasar's symmetry axis, around which the accretion disk and the line-emitting gas clouds rotate. The value of FWHMHβ, and the inferred velocity, tend to be small when the jets are pointing in our direction, and large when they are pointing away5. But we do not clearly see strong, distinct jets in the general quasar population. Therefore, correcting for the geometric effect of the quasar tilt on FWHMHβ-based estimations of black-hole masses has not generally been possible.
Shen and Ho have tackled this problem. Making new and convincing arguments, they have plotted what they describe as “a main sequence of quasars” (see Fig. 1 of the paper3). The horizontal axis is the emission-line intensity ratio of iron (Fe II) to Hβ, denoted RFe II, which characterizes EV1 and also tracks the Eddington ratio; the vertical axis is FWHMHβ, which segregates quasars for a given EV1 by the orientations of the systems. Shen and Ho's Figure 1 allows astronomers to go from spectral properties that are easily measured — FWHMHβ and RFe II — to two fundamental quantities that account for the observed diversity of quasars: the Eddington ratio and orientation.
Furthermore, the authors have, for the first time, reported a statistically significant difference in the large-scale environments of quasars with differing Eddington ratios. They find that the more-massive black holes, which have lower accretion rates and hence lower Eddington ratios, exist in large-scale environments in which quasars and their host galaxies are more strongly clustered — in accordance with theoretical expectations. Tying the properties of quasars, which operate on tiny scales compared with the galaxies that harbour them, to even grander large-scale structure is a most intriguing development. The behaviour of galactic nuclei is thus linked to the largest scales of galaxy clusters, indicating evolutionary relationships between these two entities on cosmic scales.
A century ago, stellar astronomy underwent a similar breakthrough in linking observable parameters to more-fundamental physical quantities. By plotting the colours of stars against their luminosities, astronomers noticed a band, called the main sequence, along which most stars fall. The position of a star on this band is determined by its mass, which in turn governs many stellar properties, from temperature to size to lifetime. Quasars are very different from stars, as is their newly identified main sequence — which is perhaps more accurately described as a main wedge. But in an analogous way, we may now hope to develop a deeper understanding of quasars, their physical properties and perhaps even more. Clearly, the main sequence of quasars needs further testing, and only time will tell whether its utility is equal to that of the stellar main sequence.
In any event, we are now better placed to use our telescopes for collecting spectra and other data from quasars, and in turn to associate a set of observed data with a set of physical parameters. The quasar research field is maturing, just as stellar astronomy once did. Even though a quasar's Eddington ratio, black-hole mass, inclination angle, luminosity and other properties all simultaneously affect its spectrum, establishing a main sequence promises to provide an invaluable tool for separating out competing effects. Perhaps some day, advances in technology will allow us to obtain an image of a quasar that is clear enough to verify the picture that Shen and Ho advance today.
Schmidt, M. Nature 197, 1040 (1963).
Lynden-Bell, D. Nature 223, 690–694 (1969).
Shen, Y. & Ho, L. C. Nature 513, 210–213 (2014).
Boroson, T. A. & Green, R. F. Astrophys. J. Suppl. Ser. 80, 109–135 (1992).
Wills, B. J. & Browne, I. W. A. Astrophys. J. 302, 56–63 (1986).