An Auger Observatory View of Centaurus A Roger Clay, University of Adelaide based on work particularly done with: Bruce Dawson, Adelaide Jose Bellido, Adelaide Ben Whelan, Adelaide and the Auger Collaboration
An Auger Observatory View of Centaurus A This talk will introduce the need to incorporate high energy astrophysics into our understanding of mainstream astrophysics. Some basic technique ideas will be discussed and then used to understand the Pierre Auger Observatory. Some astrophysical ideas central to high energy astrophysics will be described. Recent results from Auger will be presented.
Galactic Centres and AGN are of particular interest in high energy astrophysics, which concentrates on studies of the Universe through high energy radiations in many forms. Particles, including photons, at energies of GeV and above (TeV, PeV, EeV) are adding a new dimension to our understanding of familiar, and new, objects.
The Pierre Auger Observatory was conceived to attempt directional astronomy with charged particles. It is close to doing that and Centaurus A looks to be (should be) a likely source. The Observatory has already measured interactions with the CMB in the local Universe. It is now producing data which will teach us about magnetic fields in the same region. There are many remaining serious puzzles to be resolved before we will feel that we have a real understanding of the Universe out to 100 Mpc. First of all some context from lower energies.
As noted with H.E.S.S., our galaxy, the galactic centre, supernova remnants, pulsar wind nebulae, etc. are the types of object which are characteristically observed. These are generally recognised as energetic/violent objects which are often observed through synchrotron and inverse Compton processes clearly complementing radio astronomical studies.
H.E.S.S. operates by observing specific sources for extended periods of integration time. The alternative process involves imaging the whole of the accessible sky. Such a monitoring process is more traditional for cosmic ray studies and is currently employed by MILAGRO, the Tibet Array, Pierre Auger etc.
Ap.J. 595, 803, 2003 Milagro all sky
Synchrotron and inverse Compton processes producing observable photons require high energy particles (cosmic rays, energetic electrons, TeV etc.). Our goal is to directly observe such high energy particles and to determine their origins through directional measurements.
The Cosmic Ray Spectrum Things to note are: It covers a huge range of energies so the physics is likely to change from one part to another. It is steep, with a power law slope about 3. It is rather featureless but steepens a little in the middle and flattens near the highest energies. Radius of gyration 1 kpc in 1micro Gauss, 0.1 nt Radius of gyration 100 kpc in 1micro Gauss, 0.1 nt
Primary cosmic rays interact and a cascade develops which converts primary particle kinetic energy into secondary particles.
At energies of interest to Auger (above 1 EeV 1018 ev), the flux is low and a large telescope is required. In our case, with a 3000 square kilometre collecting area in Argentina. A northern observatory is planned with many times that area.
The Pierre Auger Observatory records the cascades using ground based detectors and atmospheric fluorescence detectors.
A large collecting area is required because of the low flux 3000 square kilometres (1600 tanks).
Observing the cascade with nitrogen fluorescence and particle detectors. We deliberately avoid the forward directed Cerenkov light.
We can occasionally see cascades from four fluorescence telescopes and the ground array. This gives us confidence in our calibrations and our understanding of the cascade physics (above accelerator energies).
Comments on measurement uncertainties. Uncertainties in the primary energy are due to unknowns associated with the shower physics (above, but not greatly, accelerator energies), the atmospheric nitrogen fluorescence yield under various conditions, and many small effects at the few % level. The energy error budget adds to a couple of tens of percent. It is continually being improved and better understood.
Comments on measurement uncertainties.continued. The primary cosmic ray composition is unknown which means that the effect of any astrophysical magnetic fields is unknown. We tend to (erroneously) assume a simple proton composition. We have upper limits to photon and neutrino fluxes which are close to those expected from robust modelling (e.g. of GZK processes). It is possible that Auger events correlate best with AGN (or other) directions for the proton component and that other events re more highly charged (e.g. Hillas 2009).
The cascade development gives us both energy and composition information.
Studies of the cascade development give us composition information.
Auger directional data: The first data were published in 2007 These were NOT ISOTROPIC. They are STILL NOT ISOTROPIC. The lack of isotropy seems to be correlated with directions of AGN within reasonable distances given the GZK process. But AGN are correlated with other things..
At energies above a few times 1019 ev, interactions with the CMB result in a small interaction mean free path for protons. THIS IS THE GZK EFFECT (but notice that the energy loss per interaction is not huge the energy attenuation length is typically ~90 Mpc) Mean free path below 10 Mpc For energies above 60 EeV
We expect to begin to see GZK photons in the non distant future.
Simple idea of a spectrum depending on the source distance. These are TIMES. Source Differential spectral index. This assumes a power law source spectrum (Surprisingly non critical.)
Suppose that sources are randomly distributed in space how will the spectrum depend on intergalactic diffusion properties i.e. magnetic field strength and turbulence scales. 56 Log(E *flux) Milky Way 3 Extragalactic MW 0.2 at2 55 10kpc 100nG 10kpc 10nG 54 10kpc 1 ng 1Mpc 10 ng 1Mpc 1nG 53 1 Mpc 100nG 52 17 17.3 17.5 17.8 18 18.3 18.5 18.8 19 19.3 19.5 19.8 20 20.3 20.5 Log(E ev) 1 ng too low. 10 100 ng possible depending on the turbulence parameters.
The spectrum is now also sensitive to the LOCAL (real) distribution of sources. Another way of looking at this is to recognise that, above the GZK cut off, the flux will have useful directional properties because we can only see locally. Fractional Anisotropy 0 log(fractional anisotropy) 0.5 17 18 19 20 1 1.5 100 ng 2 10 ng 1 ng 2.5 3 3.5 4 log(e)
So We expect to see an anisotropy for sources at distances less than about 100 Mpc because more distant sources have strongly attenuated fluxes. This is seen. This means that the dominant source distribution is anisotropic but it does not tell us how to interpret that. But we still need to identify some sources. Are there a few very strong ones or many rather similar ones?
A uniform exposure plot of (2007 data) Auger events (circles) is NOT isotropic. Hillas, arxiv 2009
The Auger events show evidence of small scale clustering. Hillas, arxiv 2009
There is some evidence that cosmic rays (circles) are correlated with AGN directions. This broad brush correlation may not seem great, there is clustering near Cen A. Hillas, arxiv 2009
On the other hand, AGN and the highest energy cosmic rays do seem to have a correlated component. How can this be? Hillas, arxiv 2009
It also works for a subset (extended radio sources) of sources. Hillas, arxiv 2009
Physics/reality check: Radius of gyration of a 1 EeV (1018 ev) proton in a 1 microgauss regular field is 1 kpc. Of course, this scales linearly so as we saw, 100 EeV has 100 kpc. More highly charged nuclei (iron 26 times) have smaller radii of gyration. Cen A is many times this distance away so retention of direction requires sub microgauss fields or small turbulence scales.
Cartoon of possible scattering field sites. Possible intergalactic or halo field. May cause loss of flux or systematic deviations magnetic spectrometer. Milky Way and Halo B Source cluster. This fills with particles which are emitted from its surface. Subtends a small angle. May be large propagation delays. Possible local group field. Scattering here results in large deviations. Could be large delays.
Galactic Deflections of protons 20 EeV and above (regular field). (Stanev 1996)
Time for the official sky map (circa 2007).
Recent Auger data AGN and Auger directions correlate. Crosses AGN Circles Auger events
Auger point spread Auger Directions on this image Galactic Field Deviation
2009 ICRC result
50 These events make up about one third of all PAO events at the highest energies even though this is not in the area of greatest sky exposure. They arguably contribute strongly to the 2007 published excess of correlations with AGN because they are within the supergalactic plane. 148 80 30 82 70 69 10 57 63 79 65-30 -40 66-50 -60-70 -10 Galactic Longitude Cen A
To incorporate the other surrounding events, we must invoke intergalactic scattering in turbulent fields, a rather natural requirement. e.g. a 10 kpc turbulence scale plus a 0.05 nt field out to Cen A would give a spread of around 300. or 10 kpc turbulence for 0.1 nt in a galactic halo to 100 kpc => around 200. Neither too unnatural and not too bad a fit. And all this takes extra time say an extra 108 years to get from Virgo an interesting number.
Conclusions. Auger has measured a GZK cut off. The cosmos out to 100 Mpc is not isotropic so we expect Auger data to be anisotropic. They are. There is evidence for small angle correlation with AGN directions. This is still somewhat perplexing to me. However, there is a clear correlation with the supergalactic plane and an apparent large scale correlation centred on Cen A. The data are still sparse, but it is not clear that the more energetic particles are more concentrated on Cen A. There are basic theory issues. How does Cen A, or even our Milky Way accelerate particle to EeV energies.
Non conclusions. So, there are loose ends and, of course, more data will be required to resolve them. Why is there no clear change in mean event energy with angular distance from the Cen A direction? Do regular intergalactic/halo fields play a role? Is it a coincidence that the higher energy events are to the north of Cen A? Why is there clustering around Cen A but nothing towards Virgo? What role is composition playing?
With charged particles, there is an extra degree of freedom the choice of magnetic field structure. What if it had a regular component, an extended galactic halo perhaps Maybe we get Virgo?
Thanks