Chapter 16: The Expanding Universe

 

An image of one of the first quasars discovered in the radio galaxy 3C273.

The first quasars were discovered using radio surveys, and by the end of the 1960s, a few hundred were known. Radio searches for quasars are very efficient because strong radio emission is a good indication of non-thermal activity. However, it turns out that relatively few quasars are strong radio sources; most must be found by their optical emission. Optical searches are more difficult because the sky is crowded with faint stars and galaxies. Luckily, the extremely blue colors and broad, and highly redshifted emission lines of quasars distinguish them from stars. Optical techniques are now the most efficient way of finding quasars. The number of quasars increased from a few hundred in 1970 to a few thousand in 1985 to 20,000 by the year 2000. Since then, large surveys using the Anglo-Australian Telescope at the Sloan Digital Sky Survey telescope have increased the numbers dramatically. Over a million quasars are now known!

Quasars are extraordinarily distant. Their redshifts range from z = 0.2 up to z = 7.1. Using the approximation d = z c / H₀, the nearest quasar is therefore 0.2 x 3 x 10⁵ / 65 = 1000 Mpc, or over 3 billion light years away. The distance to the highest redshift quasars depends on the cosmological model assumed, but it is in the range 7000 to 8000 Mpc or 20 to 25 billion light years. The range in look-back time is 25% to 95% of the age of the universe. Quasars span a large range in luminosity, from 10¹² to over 10¹⁵ times the luminosity of the Sun — equivalent to up to 1000 times the luminosity of the Milky Way Galaxy.

Astronomers have a clever technique for figuring out the size of a quasar's emission region. This technique relies on the fact that non-thermal radiation usually varies in intensity. Most quasars vary in brightness over a span of several weeks or a month. Because of the finite velocity of light, this data translates into an upper limit on an object's size. Imagine that the object is one light year in diameter. If the light output from the entire object varies it will take one year longer for the signal from the far side of the object to vary than the signal from the near side. An object's brightness cannot vary any faster than the time it takes light to travel from one side of the object to the other. If a quasar varies on a time scale as short as one week, its diameter must be less than one light week or 3 × 10⁵ × 3600 × 24 × 7 = 1.8 × 10¹¹ km. This is roughly 1000 A.U., which is only about 10 times the diameter of the Solar System. The estimated size based on this light travel time argument is orders of magnitude better than any size estimate that could come from the best space-based imaging.

The fantastic energy output of quasars comes from a tiny volume at the center of the host galaxy. As an analogy for this remarkable concentration of energy, imagine flying above a large city at night with the lights of the city spread out below you. The city probably contains about 100 million house lights, streetlights, and car lights spread over a region of 50 kilometers across. Think of each light as a star (actually, each light represents more like a thousand stars). A quasar is then a light source only a centimeter across, with an intensity equal to 1000 times the sum of all the lights in the city! Packing so much energy into such a small volume is a real challenge for theories of the quasar power source. A supermassive black hole operating as a gravity engine is the only plausible way to get so much energy out of such a small volume.

Spectra of QSO's (quasi-stellar objects) reveal a plethora of absorption lines, most of which actually have nothing to do with the QSO itself. Instead they are due to cosmologically distributed gas which intersects the line of sight from Earth to the QSO and imprints a distinct absorption signature onto the spectrum of the QSO.

The spectra of quasars show strong, broad emission lines. Astronomers have created a composite spectrum that illustrates the typical spectral features of a quasar. If the width of the lines is due to Doppler shifts, then the hot gas emitting the lines must be moving between 10,000 and 20,000 km/s, or roughly 5% of the velocity of light. The strongest emission lines are due to ionized hydrogen, carbon, magnesium, neon, oxygen, and nitrogen. Heavy elements have been observed in the spectra of even the most distant quasars, with look-back times of about 12-13 billion years. We conclude that the laws of physics seem to be unaltered over large distances — the elements we see in the far reaches of the cosmos are the same elements we find in the neighborhood of the Sun. Also, since we see these heavy elements at a look back time of 95% of the age of the universe, generations of stars must have lived and died in the first 5% of the lifetime of the universe. These elements were then dispersed into the interstellar medium to become ionized atoms in the gas swirling close to the quasar core.

 

The anatomy of a quasar.

To move beyond a simple description of quasar properties, we need to know more about their demographics. How many quasars are there in a volume of space, how old are they, and how do they relate to normal galaxies? The most important finding is that quasars evolve. There is a steep increase in the number of quasars at fainter brightness levels, which indicates that quasars were both more luminous and more numerous in the past. The study of large and complete samples allows us to recount the history of the quasar population. Quasars were first born about 10 billion years ago. They increased in number for 2 or 3 billion years, reaching a maximum space density at z = 2. Since then they have been gradually fading, like brilliant embers. The space density of luminous quasars now is hundreds of times less than it was 6-8 billion years ago, at the glorious peak of quasar activity. The most luminous quasars are extremely rare now — astronomers would need to search a volume of space about 500 Mpc across just to find one!

Quasar host galaxies

If we consider the relative numbers of quasars and galaxies in a fixed volume of space, about one galaxy in 1000 has quasar activity. This information can be interpreted in two ways. Perhaps one galaxy in 1000 is special in some way that makes it develop a quasar nucleus. Or perhaps every galaxy goes through a phase of quasar activity lasting 1/1000 of the age of the universe or about 10 million years. We are limited to a snapshot of the universe, an inevitable consequence of our own short lives and the universe's great age. At any given time, only 1 in 1000 galaxies will be switched on to show quasar activity. The two results look the same, and we cannot tell the difference with a simple census. However, if we add the information that every nearby galaxy has a supermassive black hole but almost all of them are inactive, we learn the answer: every galaxy has the potential for quasars activity, but the activity is episodic and only occurs 0.1% of the time for any particular galaxy.