Roland Crocker Monash University The  -ray and radio glow of the Central Molecular Zone and the Galactic centre magnetic field

Preliminaries Tera-electronvolts: TeV = eV ~ erg Work in progress With: David Jones, Fulvio Melia, Juergen Ott, Elizabeth Wommer, Ray Protheroe

-turbulent! Credit: T. Oka ~10% Galaxy’s molecular gas

Pieces of Puzzle

LaRosa et al. (2000) Galactic Center Region at 90 cm ~ 300 pc Nonthermal (radio-emitting) filaments NTFs Large scale magnetic fields and relativistic electrons SNRs, HII regions Poloidal magnetic field within ~100 pc of nucleus Sgr A: compact radio sources at nucleus of Milky Way (3 x10 6 solar mass BH) Credit: Dermer and Atoyan Gal. plane

Magnetic Field? Large-scale magnetic field amplitude is uncertain by two orders of magnitude “Equipartition” between CR and magnetic energy densities favours ~10  G field (e.g. LaRosa et al 2005). On the other hand… 330 MHz GBT

Magnetic Field? Non-thermal filament observations favour ~mG field (Morris and co-workers). Credit: Yusef-Zadeh/VLA

Credit: HESS Collab HESS GC Observations Sgr A

What is the mechanism for  -ray production? 1.Leptonic models - high-energy electrons (10 TeV+) inverse-Compton scatter ambient light to TeV energies 2.Hadronic models - protons (and heavier ions) collide with ambient gas (H 2 ) and produce pions 3.Dark matter annihilation?

What is the mechanism for  -ray production in Sgr A? 1.Leptonic models - high-energy electrons (10 TeV+) inverse-Compton scatter ambient light to TeV energies 2.Hadronic models - protons (and heavier ions) collide with ambient gas (H 2 ) and produce pions 3.Dark matter annihilation? ?

Credit: HESS Collab

Interpreting the data

Credit: HESS Collab CS contours from Tsuboi et al. (1999)

Protons or electrons for CMZ emission? If electrons –Why the correlation with the molecular material? –Strong magnetic fields  short cooling times  compact sources with X-ray counterparts…not seen If protons –Angular correlation between gas and  -rays naturally explained

Protons or electrons for CMZ emission? If electrons –Why the correlation with the molecular material? –Strong magnetic fields  short cooling times  compact sources with X-ray counterparts…not seen If protons –Angular correlation between gas and  -rays naturally explained

Where are the requisite CRs coming from? Spectrum much harder than local CR population Spectral index of the diffuse emission and the GC point source are so similar  common accelerator? (at Sgr A*?)

Single central source? No: propagation modelling by Wommer, Melia & Fatuzzo (2008) over size scales of diffuse emission ~200 pc diameter region shows that CR p’s have extremely restricted diffusion spheres: e.g. 5 point sources and 10  G field Need about 50 point sources to get sufficiently smooth distribution

A truly diffuse source mechanism is required A natural candidate is “2nd order Fermi” or stochastic acceleration on the turbulent and time-varying magnetic fields expected to be threading the region e.g. Distributed injection and 10  G field

Consequences of hadronic picture The CMZ should be a source of 1.Radio synchrotron 2.Neutrinos π0π0 p p p π±π± π±π±   e±e± ν radio H2H2

Radio Observations of ‘DNS’ GBT 330 MHz 2.7 GHz Parkes

Radio Observations of ‘DNS’ ATNF 2.4 GHz mosaic by David Jones

log[ /Hz] log[F /Jy]

log[ /Hz] log[F /Jy] 330 MHz LaRosa et al. (2005) GBT 74 MHz LaRosa et al. (2005) VLA 330 MHz LaRosa et al. (2005) VLA

log[ /Hz] log[F /Jy] 330 MHz LaRosa et al. (2005) GBT 74 MHz LaRosa et al. (2005) VLA 330 MHz LaRosa et al. (2005) VLA D. Jones ATCA 2.4 GHz 1.4 GHz

log[ /Hz] log[F /Jy] 330 MHz LaRosa et al. (2005) GBT 74 MHz LaRosa et al. (2005) VLA 330 MHz LaRosa et al. (2005) VLA 1.4, 2.4 GHz D. Jones ATCA 2.7 GHz Duncan et al. (1995) Parkes 10 GHz Handa et al. (1987) Nobeyama 2.4 GHz 1.4 GHz D. Jones ATCA

log[F /Jy] log[ /Hz] bk SPIN = 2.9

too much mass

too much 300 MeV brems

U p too high (assuming 2ndaries) too much mass too much 300 MeV brems

too much mass U p too high (assuming 2ndaries) too much 300 MeV brems

z  x s -1

 eqprtn with 8 keV plasma

Connecting radio and  -rays If B = 0.1 mG then proton primaries responsible for ~GHz emission (of secondary electrons) have energy ~30 GeV proton primaries responsible for  -rays > 0.3 TeV have energy ~5 TeV The spectral indices are different: ~2.9 for radio ~2.3 for  -rays Conclusion: the proton distribution steepens towards low energies

Log[E  /eV] Log[F  /(eV -1 s -1 cm -2 sr -1 )]

extras

HESS Observation of Sgr A Aharonian et al Consistent with point source, radius < 3 arcmin ~ 7 pc Located within 1 arcmin of Sgr A* -- Hard spectrum: ± 0.04(stat.) ± 0.10(syst.) Data consistent with steady source

Sgr A TeV source detected by HESS at > 40  Systematic pointing error only 6”  centroid of emission inconsistent with Sgr A East HESS Observation of Sgr A