Transverse Proximity Effect with a foreground quasar

In the thread "Plasma redshift, coronal heating, QSOs, CMB, DM halos
etc." I wrote about the Transverse Proximity Effect (TPE) with a
foreground quasar. Steve Willner (May 14) suggested that dust
obscuration and beaming may explain the lack of the expected effect.
Here is my response. For a fuller account of the failure to find the
TPE with a foreground quasar, please refer to:

http://astroneu.com/plasma-redshift-1/#TPE

In the intial thread, I wrote:

The TPE effect is expected according to Big Bang cosmology - the
foreground quasar is believed to lie close to the sightline to a
background quasar and the foreground quasar is predicted to
ionize all neutral H in its vicinity, which should result in an
absence of Lyman alpha absorption in the spectrum of the
background quasar at a wavelength corresponding to the redshift
of the foreground quasar.

1 - The foreground quasar turns on and off - and was off
at the time it would have had to be on to ionize the
neutral H in the sightline to the background quasar.

2 - The foreground quasar's light (UV at least) is beamed
towards us and does not affect the sightline to the
background quasar.

3 - The foreground quasar is surrounded by a cloud which
prevents its light from ionizing the neutral H in the
sightline to the background quasar.

However, a simpler explanation is that the redshift of light
from these quasars happens primarily near them (due to plasma
redshift or some other such process) so firstly the quasars are
closer than usually assumed and secondly the redshift along the
sightline doesn't happen in a linear or easily predictable
fashion. In this explanation, we have no clear idea of the
distances to the quasars. Maybe the so-called "background"
quasar, the one with the higher redshift, is closer than the
lower redshift quasar, but has more of its total redshift
occurring in the region close to it.

Steve Willner wrote:

I haven't studied this in detail, although I did look at the thesis
cited above. I'll just offer a couple of comments.

a) It doesn't take very much dust to absorb all the ionizing photons.

b) In the usual QSO model, the active nucleus is surrounded by a dusty
torus.

c) Optical searches will preferentially select QSO's where the torus
is pole-on to our line of sight and thus edge-on where the line of
sight to a background QSO passes nearest the foreground QSO.

I suspect the issue could be settled by a combination of infrared
observations to detect the dusty toruses and using hard X-ray
selection to pick unbiased samples. Maybe that has already been done
-- as I say, I didn't look very hard -- but I didn't see any treatment
of these issues at first glance.

Of the three papers I cited - Rupert Croft's, Michael Schirber's thesis
and his paper with two other researchers:

http://arxiv.org/abs/astro-ph/0310890 (Rupert Croft)
http://arxiv.org/abs/astro-ph/0307563
http://www.physics.ohio-state.edu/~astro/thesis.pdf

I recall that the strongest rejection of beaming is in Croft's paper.
Here are some quotes:

(p17)

Comparing to the SDSS results in Figure 13 it does not seem that
the lack of a proximity effect in that data can be due to the
beaming from an angle close to that we have used here (90 degree
cone angle). Reproducing Figure 13 with beaming alone would
require a much smaller angle.

(p18, discussing his modelling attempts to reproduce the lack of TPE.)

The anisotropy of quasar emission was investigated in our
simulations using a half-opening angle of 45 degrees. This did
have a noticeable effect on the absorption plotted in the sigma
- pi plane, with shadowing evident of regions at greater angles
from the sightline. The extra absorption in these regions did not
lead to much difference in the angle-averaged mean absorption
around quasars though, and in order to reproduce the observed
results, a very small opening angle would appear to be required.
For example, in the observational sample, there are 5 sightlines
with an impact parameters between 2.5 to 5 h^-1MPc, and these show
no evidence of a proximity effect. The excess absorption over the
mean seen close to quasars is evident out to 10 h^-1MPc, which means
that a maximum half opening angle of ~ 15 degrees is required. As
also calculated by Schirber and Miralda-Escude (2003), this seems
too small to be consistent with expectations of quasar emission.

So they estimate that in order to explain the observations with beaming,
beams of UV ionizing radiation only 30 degrees wide would be needed.

The abstract and the paper itself ends with a discussion of quasars
turning on and off.

Variability of quasars in bursts with timescales 10^4years and
10^6 years could reconcile these two facts.

From Micheal Schirber et al's paper, page 19:

A beam radius as small as 20 degrees for the QSO radiation seems
implausible. In unified models of AGN, the continuum ionizing
radiation is supposed to come from the accretion disk, which may be
absorbed by an obscuring torus near the equator, but typical
half-opening angles are ~ 30 - 45 degrees (Antonucci 1993;
Schmitt etal 2001), and they are thought to increase with luminosity
(Rudge & Raine 2000). A separate possibility is that the QSO has not
ionized the gas in its host halo, and that the ionizing radiation is
able to escape only along a narrow tunnel among clouds. However, the
fact that most QSOs of luminosity similar to the foreground one in
pair 1 do not exhibit intrinsic Lyman limit absorption in their
spectrum implies that this explanation could not account for a narrow
beam of emission in most QSOs.

To summarize, beaming of the ionizing radiation might be one of the
reasons for the absence of the transverse proximity effect in our
three pairs, but if this absence is generally confirmed on a larger
sample of pairs, then beaming alone cannot be the sole explanation.

I just used Google and AdsAbs to search for any references to the
"transverse proximity effect" since I last looked. It seems that the
recent papers are generally based on the notion that this lack of TPE
indicates that quasar lifetimes and/or variablitity. Their focus on
lifetimes is clear in their title, or in the quoted text:

Multiepoch Sky Surveys and the Lifetime of Quasars
Paul Martini, Donald P. Schneider 2003 Sep 23 (ApJ 597 L109-L112)
http://arxiv.org/abs/astro-ph/0309650

Calibrating the Galaxy Halo - Black Hole Relation Based on the
Clustering of Quasars
Stuart Wyithe, Abraham Loeb 2004 Mar 31 (Submitted to ApJ.)
http://arxiv.org/abs/astro-ph/0403714

Preliminary results suggested quasar lifetimes of tq ∼ 10^6 −
10^7 years, consistent with the values determined by other
methods (see Martini 2003 for a review), including the
transverse proximity effect (Jakobsen et al. 2003) . . .

QSO Lifetimes
Paul Martini 2003 Apr 1
http://arxiv.org/abs/astro-ph/0304009

Measuring the Radiative Histories of QSOs with the Transverse
Proximity Effect
Kurt L. Adelberger 2004 May 25 (Accepted ApJ.)
http://arxiv.org/abs/astro-ph/0405505

I found only one new paper mentioning the TPE which doesn't seem to be
following the path of quasar variability or limited lifetimes:

The Mysterious Absence of Neutral Hydrogen within One Mpc of a
Luminous Quasar at Redshift 2.168
Paul J. Francis, Joss Bland-Hawthorn 2004 May 25 (Accepted MNRAS.)
http://arxiv.org/abs/astro-ph/0405506

We showed in 4.1 that if the region surrounding PKS 0424-131 were
typical, we should have expected to have seen the fluorescent Ly
alpha emission from a considerable number of clouds. We should also
have seen internally ionised clouds. Instead, we saw nothing.

Since the quasar-quasar TPE researchers do not seem to contemplate the
possibility that the quasar distances are other than that conventionally
predicted by a Doppler interpretation of their redshift, they have to
choose between the quasar turning on and off, beaming, and dust
obscuration. It seems that in general they have ruled out the latter
two, and so are going with the first - though there are many objections
to that as well.


- Robin http://astroneu.com

Four weeks have passed and no-one has written about how they account for
these observations within the Big Bang Theory.

The researchers consider that their choices are between beaming,
absorption clouds or limited quasar lifetimes. It seems they have ruled
out the first two and concluded that quasars must have limited lifetimes
or that they turn on and off.

Michael Schirber et al. in http://www.arxiv.org/abs/astro-ph/0307563 :

Variability demands that the luminosity of the QSO with the largest
predicted effect was much lower 10^6 years ago . . .

Though he cites one study which apparently found potential TPE in HE II
in one pair of quasars, at a redshift 0.006 from where it was expected.

Rupert Croft, in http://www.arxiv.org/abs/astro-ph/0310890 :

For both quasar lifetimes which we simulate (10^7 yr and 10^8 yr),
we expect to see a strong decrease in the Lya absorption close to
other quasars (the "foreground" proximity effect). We then use data
from the Sloan Digital Sky Survey First Data Release to make an
observational determination of this statistic. We find no sign of our
predicted lack of absorption, but instead increased absorption close
to quasars. If the bursts of radiation from quasars last on average
10^6 yr, then we would not expect to be able to see the foreground
[implicitly: proximity] effect.

As far as I am aware, there is little or no observational or theoretical
support for such limited lifetimes - or for most quasars turning on and
off with such short time periods. The only theory I am aware of to
explain a massive black hole in a galaxy reducing its accretion and
radiation significantly is along the lines of the high-emission stage
being caused by a temporary disturbance throwing lots of matter into the
accretion disk, with the rest of the galaxy orbiting at a safe distance
and avoiding being drawn in once the disturbed material has been
devoured. Its hard to imagine sharp turnoffs with this mechanism, since
there isn't any obvious safe distance from a black hole. Also, I don't
think that the centres of spiral galaxies are characterised by neat
circular orbits.

Furthermore, quasars are conventionally thought to generally reside in
elliptical galaxies, where all the stars are in scattered elliptical
orbits. This would mean that firstly there are a great many stars
occasionally getting "close to" the black-hole as they return to the
centre of the galaxy and that secondly this cloud of stars would
inevitably and presumably continually be perturbed by the black-hole's
gravity into orbits which would, in some fraction of cases, cause them
to be devoured.

Another such objection to quasar cores turning on and off - based on how
a core could possibly be starved of stars at all, and especially how this
starvation could change drastically in less than a million years - is
that the orbital motions of stars are generally a lot slower than this
million year timeframe. I haven't looked at current understandings of
orbital motions in spiral or elliptical galaxies, but if the Sun takes
around 240 million years to orbit our Galaxy, it is hard for me to
imagine that the orbital motions of stars in the middle of any galaxy
could all attain a black-hole avoidant state under any circumstances,
much less gain or lose that state in such a short time.

What is needed to explain the lack of TPE is a fast turn on, as well as
an overall low "duty cycle" (on to off ratio) which requires that
quasars be even more numerous than the visibly active subset we observe.

According to the BBT, the larger radio galaxies have lobes well over
10^6 light years from the core - with these lobes generally regarded as
resulting from presumably continuous jet activity over "several tens of
light years":

http://www.astron.nl/wsrt/press/text_english.html

though this page shows some objects in which the jet activity evidently
does turn on and off faster than this.

The 19 giant radio galaxies in Table 1 of:

A new sample of giant radio galaxies from the WENSS survey: I -
Sample definition, selection effects and first results
A.P. Schoenmakers et al. http://www.arxiv.org/abs/astro-ph/0107309

average 1 Mpc from core to lobe. The newly discovered radio galaxies in
Table A.4. average 1.18 Mpc from core to lobe - which is 3.84 million
light years. Since the velocity the jet components (whatever they are)
is below the speed of light and the lobes are clearly the result of a
period of activity which took longer (in total, even if broken up into
periods of activity and inactivity) than the time it takes material to
travel to the lobe, it seems reasonable to suggest that the jets (and
therefore presumably the cores) of these objects are active for at least
tens of millions of years and perhaps hundreds or thousands.

While some objects do exhibit uneven jets - indicating changes in
jet-making activity and so perhaps the core's radiation at other
wavelengths (such as the UV which would ionize hydrogen and give rise to
the Transverse Proximity Effect) - it seems that if quasar activity (at
least as far as UV emissions are concerned) really does normally turn on
for periods as brief as a million years or less, then that these objects
must generally undergo many such cycles of turning on and off.

The BBT involves the Universe in the past being populated with many
quasars of absolute luminosities far greater than that of whole
galaxies. A June 1993 article in Scientific American by George K. Miley
and Kenneth C. Chambers illustrates a "quasar era" at redshifts 1.5 to
2.7, spanning just over a billion years - about 1.4 to 2.5 billion years
after the BB. While currently popular cosmologies may give somewhat
differing figures, it seems the BBT has an "era" of quasars at least a
billion years long - which doesn't seem to fit with the idea that each
quasars is "On" for only a single stint of a million years.

I think that a better explanation for the lack of TPE, and the
observational evidence against time dilation with higher redshift
quasars indicates that the quasars are not as far away as their
redshifts conventionally indicate, that they (and likewise galaxies) are
not moving apart from each other or from Earth, and that therefore a
large part of the redshift of quasars is "intrinsic" to them - most
likely occurring in the space relatively close to the core.

I will write more on this in another message - this one is long enough.
I just wanted to point out that I think these researcher's failure to
find the TPE, even with their greatest efforts and whilst they
apparently have no desire to disprove or even question the Big Bang
Theory, consitutes strong evidence that something is very wrong with
the Theory.

I would like to see BBT supporters reconcile these observations with
their understanding of the Theory.


- Robin http://astroneu.com

Robin Whittle wrote:


Another such objection to quasar cores turning on and off - based on how
a core could possibly be starved of stars at all, and especially how this
starvation could change drastically in less than a million years - is
that the orbital motions of stars are generally a lot slower than this
million year timeframe. I haven't looked at current understandings of
orbital motions in spiral or elliptical galaxies, but if the Sun takes
around 240 million years to orbit our Galaxy, it is hard for me to
imagine that the orbital motions of stars in the middle of any galaxy
could all attain a black-hole avoidant state under any circumstances,
much less gain or lose that state in such a short time.


Some stars within the nuclei of galaxies have orbits only a few parsecs
or tens of parsecs in size, which means that their orbital timescales
can be as short as a few decades. Look, for example, at

http://spiff.rit.edu/classes/phys240.../mwcenter.html

and the references therein (Ghez et al., Nature, vol 407, p. 349 (2000)).

Michael Richmond

In article ,
Robin Whittle wrote:
The researchers consider that their choices are between beaming,
absorption clouds or limited quasar lifetimes. It seems they have ruled
out the first two and concluded that quasars must have limited lifetimes
or that they turn on and off.

A lot of your posting is about whether quasars do have limited
lifetimes, so let me make some comments about that.

[...] with the rest of the galaxy orbiting at a safe distance
and avoiding being drawn in once the disturbed material has been
devoured. Its hard to imagine sharp turnoffs with this mechanism, since
there isn't any obvious safe distance from a black hole.

This looks like a `cosmic vacuum cleaner' model of the black hole.
It's not the distance but the angular momentum that counts; if
material near the black hole can't shed its angular momentum, it's not
going to be accreted. It's entirely possible that, once the mass that
was dense enough to form a conventional dissipative accretion disc has
been swallowed up, lots of other matter that is capable of being
dynamically affected by the black hole will sit around not being
accreted at any significant. Indeed, this is thought to be exactly
what's going on in the centres of the many nearby galaxies, our own
included, where there is strong (in our own galaxy's case extremely
strong) evidence for a central black hole with dynamical effects, but
little or no evidence for ongoing AGN activity, still less a quasar.

Furthermore, quasars are conventionally thought to generally reside in
elliptical galaxies,

The most luminous quasars are in ellipticals; many are in spirals.

This would mean that firstly there are a great many stars
occasionally getting "close to" the black-hole as they return to the
centre of the galaxy and that secondly this cloud of stars would
inevitably and presumably continually be perturbed by the black-hole's
gravity into orbits which would, in some fraction of cases, cause them
to be devoured.

Sure. People have been very interested in the results of tidal
disruption of a star by a black hole. But the timescales for this
process, for an individual star, are very short compared to even 10^6
years. It would produce an interesting observational effect, but it
wouldn't look like a quasar. (See
http://www.astro.rug.nl/~spijkman/aoz/stars.html and refs therein).
The stellar cores of galaxies aren't dense enough to provide a
continuous flow of stars onto a black hole. To feed a quasar you need
gas.

According to the BBT, the larger radio galaxies have lobes well over
10^6 light years from the core - with these lobes generally regarded as
resulting from presumably continuous jet activity over "several tens of
light years":

you mean `several tens of millions of years' I think.

The estimated lifetimes of powerful radio galaxies are certainly of
the order of 10^7 years. You should note a couple of things, though:
one, there's a potential selection effect in that powerful radio
galaxies are likely to be luminous systems well supplied with fuel, or
we wouldn't see them: two, there certainly are radio galaxies (or
possibly former radio-loud quasars) which can be detected by their
radio emission but which have *no* apparent ongoing nuclear activity.
These systems have certainly been through a recent AGN phase and then
stopped; which proves, if we need proof, that that's possible.

The 19 giant radio galaxies in Table 1 of:

A new sample of giant radio galaxies from the WENSS survey: I -
Sample definition, selection effects and first results
A.P. Schoenmakers et al. http://www.arxiv.org/abs/astro-ph/0107309

average 1 Mpc from core to lobe.

So we know that at least some radio galaxies -- the largest known --
must have lifetimes longer than a couple of million years.

it seems that if quasar activity (at
least as far as UV emissions are concerned) really does normally turn on
for periods as brief as a million years or less, then that these objects
must generally undergo many such cycles of turning on and off.

Yes, that is the standard picture.

The BBT involves the Universe in the past being populated with many
quasars of absolute luminosities far greater than that of whole
galaxies. A June 1993 article in Scientific American by George K. Miley
and Kenneth C. Chambers illustrates a "quasar era" at redshifts 1.5 to
2.7, spanning just over a billion years - about 1.4 to 2.5 billion years
after the BB. While currently popular cosmologies may give somewhat
differing figures, it seems the BBT has an "era" of quasars at least a
billion years long - which doesn't seem to fit with the idea that each
quasars is "On" for only a single stint of a million years.

A couple of points:
1) I don't think anyone suggested that each quasar was only on for a
million years and then never went through any subsequent phase of
activity. All that's necessary is that the quasars go off for a while
(probably of the order of another 10^6 years or so) before starting
again.
2) There's no problem in producing a billion-year `quasar era' (though
in passing I should say that the idea of an `era' of quasars, as
opposed to a time when they were most common, is slightly misleading,
not to say outdated) with quasars that individually last only of the
order of millions of years, provided that there are plenty of them
and/or that they repeat. We wouldn't have a problem calling the last
5,000 years the `human era' of Earth, even though human lifetimes are
not of that order.
3) The `quasar area' is nothing to do with the BB: it's an
observational requirement in any model in which distance and redshift
are related in a conventional way.

Martin
--
Martin Hardcastle Department of Physics, University of Bristol
A little learning is a dangerous thing; / Drink deep, or taste not the
Pierian spring; / There shallow draughts intoxicate the brain ...