1
|
|
2
|
- OJ287 was first detected as a radio source in the Ohio surveys in 1968.
- It was soon identified with a 15th magnitude compact blue
object like a typical quasar.
- Unlike a normal quasar, the optical spectrum was found to be
featureless, but very weak, variable emission lines have since been
detected, with a red shift of z=0.306.
- OJ287 was classified first as a BL Lac Object and, later, as a blazar –
a quasar with extreme variability.
|
3
|
- OJ287 seems to be an exceptional object in that its relativistic jet is
closer to our line of sight than any other AGN – we look down inside the
throat of the jet.
- There is a very high degree of relativistic beaming.
- The host galaxy of OJ287 is barely detectable, even in the infrared,
where it should show up more clearly.
|
4
|
- OJ287 is highly variable at all wavelengths from the radio to x-rays
(and possibly gamma rays).
- The variability can be extremely violent – half a magnitude, or more, in
less than an hour.
- There is even a report of a 0.8 mag variation in the near-IR in 50
seconds!
|
5
|
- The optical light curve is characterised by occasional large outbursts
and lower amplitude variability on all time scales.
- There is a strong suggestion of long-term trends in the light curve.
|
6
|
- Light curve data for OJ287 exists since 20/01/1892 from archival
photographic plates, although the level of sampling and data quality are
much greater since identification.
- Large outbursts are seen back to at least 1913.
|
7
|
- In 1988 Mauri Valtonen, Aimo Sillänpää at Tuorla Observatory in Finland
noticed that the outbursts appeared to repeat at regular intervals.
- They suggested that there was a period of 11.6±0.5 years and that the outbursts were due to the
interaction of a binary supermassive black hole with an eccentric orbit.
|
8
|
- Every 12 years the small black hole passes through the accretion disk of
the large one.
- The encounter causes a huge dump of material onto the large black hole.
- The result is a big increase in brightness.
|
9
|
- The 11.6 year period predicted correctly that a new outburst would occur
in November 1994.
- The maximum was also correctly predicted to be smaller than the 1983
outburst, according to the apparent 60-year modulation of amplitude.
- This was a success for the model and widely regarded as confirming the
periodicity.
- But… maybe it did not! (watch this space)
|
10
|
- The optical-infrared outburst showed a double maximum with separation »13 months.
- Only the second maximum is seen in the radio data.
- Even at 860GHz (0.35-mm) there is no evidence of the first maximum.
- OJ287 (apparently) was not in outburst in 1994 in ISO data, but the
outburst is seen clearly at 2.2 microns.
- Why the differences?
|
11
|
- Traditional methods of periodic analysis of light curves used by
scientists do not work.
- The light curve sampling has increased from <1 point per year in
1900 to 5600 points per year from 1993-95.
- The error on each point has decreased by a factor of »10.
- The analysis is totally dominated by data since 1972 (the 1993-95 data
has 104-105 times more weight in the archive than
the oldest data).
- Taking any kind of light curve average changes the results.
- Although the outbursts have been “periodic” since 1972, only about three
complete periods have been observed
- The less accurate historical data is critical in the light curve
analysis.
|
12
|
|
13
|
- Finnish astronomers Mauri Valtonen & Harry Lehto propose that OJ287
has:
- A binary black hole with masses 17x109 and 108MQ.
- A rest frame orbital period of 8.9 years.
- Eccentric orbit, rather like a comet such as Encke.
- Pericentre advance »35º
per orbit.
- Superflares due to massive infall from the accretion disk caused by
gravitational perturbations at pericentre.
- Lifetime against decay by emission of gravitational radiation of ~106 years.
- Implies progressive decrease of orbital period (~1 week/orbit).
|
14
|
- How good is the predictive power of the model?
- Is this a true periodicity, or a pseudo-periodicity stable over a few
periods? No successful prediction ð No Good!!!!
- There were many predictions for the epoch of the next maximum (see
Kidger, 2000, AJ, 119, 2053 for a summary). Most suggested that it
would be between 2006.25 and 2006.75.
- Much of this range is a period of poor visibility near solar
conjunction.
- The secondary maximum in 2007 was also predicted to occur near
conjunction.
- Exact timing of the maxima would allow the model to be verified and
refined.
- Slightly different assumptions of initial parameters in the binary black
hole model lead to very different predictions.
- A sad consequence of the poor quality of the historical light curve.
|
15
|
- If the light curve is truly periodic, why is it that after >100 years
there is such uncertainty in the period?
- It is very worrying that we cannot do better, but…
- We have had only two well-observed outbursts (small number
statistics!)
- A lot of people have jumped on the periodicity bandwagon with
ridiculously short light curves (£20 years!!).
- What is the relationship between the optical and radio outbursts?
- Should alternatives to the binary black hole such as a precessing jet
(lighthouse model) perhaps be considered?
|
16
|
- The best-known prediction for the 2006 outburst (Pietilä, H. 1998, ApJ, 508,
669 ) and the periodic solution predict an outburst in summer 2006.
- No proper coordinated monitoring programme has been organised to observe
the outburst, unlike in 1993-1996.
- At the 2005 AGM of the amateur “The Astronomer” Group in Basingstoke
(England) a chance conversation with former British Astronomical
Association Variable Star Section (BAA-VSS) Director, Gary Poyner, led
us to set up an amateur monitoring campaign.
|
17
|
- The BAA-VSS has an archive of 343 observations of OJ287 from 1993 to
mid-2005.
- The majority are visual estimates made by the observer by eye at the
telescope.
- The first measure of the 2005-06 campaign was made on October 1st
2005.
- 391 additional measures were made up to June 11th 2006 with
telescopes from 20-50 cm.
- 88 visual estimates
- 186 unfiltered CCD measures
- 55 CCD + V (8 in B, 25 in R, 29 in I)
|
18
|
|
19
|
- 1993 - request from Dr. Harry Lehto for observations of OJ287 during
1993-1994.
|
20
|
|
21
|
- BAA-VSS now takes up the challenge
- Campaign announced on BAA-VSS-Alert, VSS Circular & BAA Circular
- Web page is born with regularly updated light curves http://www.garypoyner.pwp.blueyonder.co.uk/oj_camp.html
- New Sequence is made available
|
22
|
|
23
|
- Visual observations made in the same way as a Variable Star estimate
|
24
|
|
25
|
- How good are visual magnitude estimates at such faint magnitudes?
- Is there a difference between CCD measures with and without a filter?
- Just how good is amateur data anyway?
|
26
|
- An eye and a CCD are sensitive to different ranges of colour
- The human eye is not very sensitive to red light, but quite sensitive
to blue.
- A CCD is not very sensitive to blue light, but is very sensitive to red.
- An unfiltered CCD takes in a range of wavelengths of light much greater
than the human eye.
- OJ287 has a colour quite similar to that of the Sun, with more red than
blue light.
- So it is normal to expect different results with each kind of measure.
|
27
|
- Search for all observations in the archive made within an hour of each
other.
- Compare to see if one kind of data is systematically brighter or fainter
than another.
- If all the types of data are equivalent we would see all the
observations lie along the diagonal line.
- Obviously they do not!
|
28
|
- Both unfiltered CCD measures and visual estimates turn out to give
slightly BRIGHTER magnitudes than filtered CCD data.
- V = Visual + 0.14
- V = Unfiltered V + 0.09
- However, a typical difference between CCD data and a visual estimate of
0.1-0.2 magnitudes at magnitude 15 is not too bad…
|
29
|
- The corrected light curve shows an excellent agreement between filtered,
unfiltered and visual data.
- The quality of the amateur data is excellent.
|
30
|
- Initially we were disappointed to see a continuous fade rather than a
slow rise to maximum.
- OJ287 faded from V=13.6 on November 5th 2005 to V=16.1 on
Feb. 16th 2006.
- Several large flares are visible…
- …but there was obviously no rise to a summer 2006 outburst.
|
31
|
- By April Gary was getting a bit depressed…
- Then we started to look back a bit further into the archive.
- We were so convinced that there would be an outburst in summer 2006 that
we had been blinded to the obvious…
|
32
|
|
33
|
- The outburst had happened nearly a year earlier than expected.
- In fact, we had almost missed it.
- Fortunately, a few observers had followed OJ287 low in the morning sky
at dawn in October 2005.
- In early November OJ287 was magnitude 13.6, having been 14.8 in October
and fainter than 15 in May.
|
34
|
- The peak of the 2005 outburst may have been missed, but even so the peak
brightness was higher than in 1994-95.
- It is still a small outburst, at the minimum of the 60 year cycle.
- It is advanced at least 10 months with respect to the 11.84 year
period.
|
35
|
- On a phase diagram the 1949, 1961, 1972, 1983 and 1994 maxima all line
up for a stable 11.84 year period (red arrow).
- Unless the 2005 maximum (blue arrow) was simply a precursor, there has
been a large phase jump since 1994.
- The 2005 maximum clearly does not fit a stable period.
|
36
|
- The Lehto & Valtonen model predicted that the 2005 encounter would
be at low velocity and distant from the centre of the accretion disk.
- A long delay from plane-crossing to optical flare.
- A rather poorly defined maximum due to the encounter geometry.
- A large advance in the date of the optical flare.
|
37
|
- Disk impact at 2005.09
- Primary accretion disk lifted 280AU out of plane.
- Delay in optical flare 0.7yr.
- Impact duration 0.23yr (same as peak of outburst).
- Secondary blast wave from impact reflection »0.2yr after impact.
|
38
|
- To fit the parameters of the black hole orbit we need 5 well-timed
maxima.
- The outbursts in 1948, 1960,1972, 1983 & 1994 seemed to fit the
bill.
- The 1994 outburst was the best observed one ever and tied the model down
very tightly.
- But, suppose that the 1994 outburst was a mistake?
|
39
|
- Just before OJ287 disappeared at conjunction in 1994 it was brightening
rapidly.
- It was much fainter again in September.
- Maybe the real outburst happened while OJ287 was invisible?
- If the wrong date was used in the calculations for the 1994 outburst,
the model and its predictions will be wrong.
|
40
|
- Either the date of the 1994 outburst is wrong, or the 2005 outburst
proves every model to be wrong!
- If we use the date of the 2005 outburst to fix the calculations, instead
of that of 1994, we find that the 1994 outburst would have happened in
early July and have been finished by the time OJ287 reappeared in the
morning sky.
|
41
|
- Superimposing the 2005 data on the 1994 light curve shows strong
similarities, but the true 1994 maximum would have been hidden if it
occurred when the revised model predicts ð Perhaps we only saw the initial rise towards
maximum and an aftershock in 1994.
|
42
|
- The 2005 maximum has allowed the model parameters to be revised.
- The mass of the primary increases from 17 to 19 billion solar masses.
- The precession rate increases from 33º to 39º per orbit.
- The fit to the early data improves greatly.
- A second, sharp maximum from a high-velocity ascending node encounter
should occur in 2007.
|
43
|
- Mauri Valtonen’s model predicts that there will be a new outburst
peaking around September 15th 2007.
- A rival model from Esko Valtaoja says that there will be an outburst
peaking in early January 2007.
- Which, if either, will be right?
- Gary and his merry men will be the referee in this duel of titans!
|
44
|
- Mauri Valtonen’s model gives a long range light curve forecast:
- A series of outbursts starting in September 2007 will keep the quasar
bright until 2010.
- As these fade the jet will realign more towards us, beaming more
emission and making the base level of the light curve rise.
- New outbursts will start in Spring 2016.
- OJ287 will remain relatively bright until about 2030.
|
45
|
- The 2005 outburst and the pre-1948 data show that the light curve is not
really periodic (sorry guys, but you’re wrong!).
- OJ287 will show periodicity for »50 years at a time, but no longer.
- The periodicity has just broken down (again).
- But the binary singularity model can explain this behaviour.
- The lifetime of the system against gravitational radiation is ~106 years ð There will be very
few close AGN binaries because they have a short lifetime. The black
holes coalesce rapidly.
- The next 12 months will make or break the rival theories.
- To a large degree it will be amateur astronomers who show which theory
(if any) is correct.
|