Galaxies and structured universe evolve.
So, active galactic nuclei evolve as well,
and in ways that are actually very
interesting.
At first, it was hard to figure out how
does that happen because the luminosity
function of quasars, at least then is
brightened, it's, is well-represented as a
parallel.
So, if you plot log of the number of
quasars per unit volume, unit luminosity
versus log of the luminosity, a parallel
looks like a straight line.
Then, you can compare luminosity functions
to different spreadsheets and you still
see a straight line.
Now, the question is, how did it get to be
that way?
If that quasar numbers evolve, that means
the, that luminosity function goes up and
down.
If the luminosities evolve, it goes left
to right, but you can't tell without any
features.
And so, for a long time, it wasn't clear
whether it was a density evolution or a
luminosity evolution.
All we knew that there were many more
quasars at large redshifts than there here
now.
The, the,[UNKNOWN] was broken once the
observations got to be deep enough so we
can see that there is actually a feature
and luminosity function is not a pure
parallel.
And here it is, how it looks like.
The truth is the combination of both.
There were more quasars back then, but
they also tended to be more luminous, and
you can think of it as, in a sense of,
well, there are things back then that are
not here now, but quasars are not so much
objects as events, as phases of evolution
in general behavior of supermassive black
holes.
We think that any galaxy with a massive
black hole in the middle can go through
multiple episodes of quasar-like activity.
And so, those were more frequent in the
past, and they're also more luminous.
And so, this is now reasonably
well-understood and we can integrate the
number of quasars attending even redshift
and see how that whole thing changes as a
function of time.
And the result is something like this.
As we look from here now and go to higher
redshifts, the numbers increased.
The comoving number density of quasars
over all luminosities goes through a broad
maximum round redshifts 2 to 3, and then
past about redshift of 4 declines again.
Now, if you flip this, and think in terms
of cosmic time, this makes a perfectly
good sense.
At first, there is nothing.
You have to start building them up and
fueling through, say, merging of galaxies,
that activity reaches a peak some,
somewhere between[UNKNOWN] one and few
gigayears and that's the peak of the
quasar era.
And then, it declines because the galaxy
encounters are not as frequent as the
universe expands, and there is less gas to
be used to fuel them.
So, we think, we have a reasonably good
understanding of the picture.
You may recall that history of cosmic star
formation, is something like that, not
exactly the same, but qualitative,
qualitatively very similar.
Another way to look at this, is through
radio sources.
The radio sources we believe are largely
powered by supermassive black holes, but
not always.
And if you look at luminosity function of
radio sources and then get the redshifts
and understands where they are, you'll
find there are really two components.
On the bright end, most of the luminous
radio sources are powered by the creation
to supermassive black holes.
On the faint end, actually, they're
dominated by star formation.
Star formation can produce radio emission
through supernova remnants and such and so
then those two will evolve differently
depending on the history of star formation
versus history of active nuclei.
They're correlated but not perfectly.
And indeed, the, as far as the active
nuclei are concerned in radio, they do
evolve very much in the same way as
quasars.
Since the beginnings of radio astronomy,
astronomers counted radio sources as a
function of flux.
And this was essentially a form of
cosmological number density test but since
they're evolved rapidly it is, it was
difficult to derive any cosmological
conclusions aside from disproving steady
state cosmology.
And now, this actually is done in detail
but we're measuring those redshifts as
well.
Another question is, are quasars
clustered, just like galaxies are
clustered?
And the answer is yes.
Obviously, it has to be.
Not only are they clustering in a similar
way as galaxies, or rather groups or faint
of small clusters of galaxies, but that's
very small separations, their clustering
is even stronger and there are many cases
of binary quasars, now actually two
physically quasars orbiting each other
just like binary stars and their numbers
seem to be increasing as the separation
gets closer then say about 100
kiloparsecs, which is where galaxian halos
begin to merge, in other words, when
galaxy merging really begins.
So, the quasars are fueled by galaxy
interactions, this is perfectly
understandable.
There is also a couple of cases of triple
quasars known.
And if the quasars were following
clustering galaxies exactly, this will be
essentially impossible to achieve because
quasars are rare, we have two quasars,
together, will be even more rare, we have
three, an even more rare.
But if the proximity is correlated with
the onset of QSO activity, say because of
[unknown] merging, then you will see cases
like this and that's exactly what's
observed.
What about properties of individual
quasars?
Well, we have their spectra and as far as
we can tell, they do not evolve very much
at all.
It seems the highest redshifts, we find
them beyond redshift of 6 to the present
day.
There is changing spectrum due to the
intergalactic absorption by hydrogen as
you will recall.
But aside from that, as far as we can
tell, there is very little evolution if
any.
So, quasar phenomenon itself is always the
same.
The occurrence of the phenomenon changes
in time but what happens in the central
engine remains more or less the same.
We can infer masses of supermassive black
holes, the power quasars from their
spectra and plot that as a function of
redshift.
Since the masses are correlated with
luminosities of high redshifts, we only
see the most massive ones.
And here is an interesting plot that shows
log of the mass of central engine versus
redshift.
At lower redshifts, we probe to lower
luminosities and there is a big spread and
as we go to higher redshifts, we can only
see high mass end.
But what's really interesting is that even
to the largest redshifts we probe, there
are quasars powered by massive black holes
that go out to 10 billion solar masses,
which is remarkable.
Assembling 10 billion solar masses in a
volume smaller than a solar system is not
an easy thing, and at redshift, say, of 6
or there abouts, we only had one billion
years to do it.
So, that posed little bit of a problem.
How do you effectively make such massive
black holes so quickly?
We will talk a little more about this in
the next module.
Another interesting thing is that we can
use the spectra of quasars to infer
metallicities of the gas around.
And that, of course, is a product of a
chemical evolution.
The metals are being cooked up in stars,
their exploded supernovae, and so on.
And it turns out that even to most distant
quasars known, have very high abundances
of heavy elements like carbon and oxygen
and so on.
This is also confirmed through detections
of molecular gas around these high
redshift quasars.
And, in fact, not only do they have these
really massive black holes, but they also
have metallicities exceeding solar
metallicity.
We can plot metallicity of quasars or
other broad line regions around central
engines as a function of redshift and you
find that they can be even ten times more
metal-rich then sun.
Normally, this would require a lot of
chemical evolution to happen.
So, what this means is in the cores of
host galaxies of quasars, there was
already a great deal of star formation and
chemical enrichment history.
And not only that, but those metals have
been retained, they were not expelled by
supernova-driven weights.
So, what this immediately implies is that
quasars at very high redshift occur in
deep potential wells and the places where
we know such things can happen are cores
of the most massive elliptical galaxies
today.
So, that probably signifies that that's
where they're originating.
They're coforming with the most massive
galaxies.
A, a somewhat more subtle effect is that
the abundance patterns are slightly
different.
The abundances of heavy elements found
depend on the type of supernovae.
You may remember the type 1, supernovae
are detonating light burst.
Type 2 are really massive stars.
And the abundance pattern for quasars is
more characteristic of type 1 supernovae
and so that means systems that give them
rise, detonating, accreting light burst
must have formed even earlier.
So, the star formation in these objects
must have started at the largest redshift
we can, higher than our redshift of 10.