Active Galactic Nuclei at TeV energies

Hélène Sol LUTH, Observatoire de Paris

MG12, UNESCO, Paris, July 2009

Outline

• Introduction • The sample of AGN detected at TeV • Similarities and discrepancies : some specific cases of blazars • An emerging family of TeV AGN : the radiogalaxies M87 and Cen A • Remarks on SSC scenarios

Outline

• Introduction • The sample of AGN detected at TeV • Similarities and discrepancies : some specific cases of blazars • An emerging family of TeV AGN : the radiogalaxies M87 and Cen A • Remarks on SSC scenarios

VERITAS MAGIC

Ground-based TeV gamma-ray astronomy : Atmospheric Cherenkov Telescopes

HESS CANGAROO

Typical example of VHE spectrum and SED

VHE power-law, two-peaked SED, variability

Basic scenarios for SED modeling

leptonic acceleration

e (TeV) Synchrotron  (eV-keV)

hadronic acceleration

p + (>>TeV)  0 matter

(or photons )

 +   (TeV) B  (eV)  (TeV) Inverse Compton SSC or EC IC  0 decay

energy E

(adapted from De Lotto, 2009)

TeV emitting zone(s)

relativistic bulk motion : in a jet or outflow with Radio galaxies FR I ,

FR II Radio quasars BL Lac (HBL, LBL) and FSRQ Relativistic jet Strong relativistic boosting

(~ factor δ 4 ) favours detection of blazars/BL Lac

emitting plasma V ~ c

Blazars : HBL, IBL, LBL, FSRQ …

from J.P. Lenain

The blazar sequence

FSRQ HBL

Average SED for a sample of 126 blazars binned according to L radio (

Fossati et al, 1998

) : A continuous sequence from the most powerful FSRQ, through LBL and IBL, to the weaker HBL A proposed unifying scheme based on leptonic models with SSC and external Compton (EC) emission, the importance of EC decreasing from FSRQ to HBL (

Ghisellini et al, 1998

) However : growing evidences that not all objects fit the trend

(2003-2009; Padovani et al)



need of a revisited sequence …

Outline

• Introduction • The sample of AGN detected at TeV • Similarities and discrepancies : some specific cases of blazars • An emerging family of TeV AGN : the radiogalaxies M87 and Cen A • Remarks on SSC scenarios

Now 28 AGN

from J. Hinton

source redshift type flux in Crab unit photon index Γ shortest variability discovery paper Cen A M 87 3C 66B ?

Mkn 421 Mkn 501 1ES 2344+514 Mkn 180 1ES 1959+650 BL Lac PKS 0548-322 PKS 2005-489 RGB J0152+017 W Comae PKS 2155-304 0.00183

0.004233

0.0215

0.0300

0.03364

0.044

0.046

0.048

0.0688

0.068998

0.071022

0.080

0.1020

0.117

FR I HBL HBL HBL HBL HBL LBL HBL HBL HBL IBL HBL FR I FR I 0.008

0.014 0.033

0.027 ?

0.3-10 0.06-11 0.08-0.5

0.11

0.05-2.2

0.03

0.014

0.025

0.02

0.09

0.1-15 2.7

2.2

3.1 ?

2.1 2.7-1.9

2.8-2.5

3.3

3.2-2.5

3.6

2.8

4.0

2.95

3.8

3.2-3.9

- day

Aharonian et al, 2009 Aharonian et al, 2004

year ?

mins mins day - week - - year month day mins

Aliu et al, 2009 Punch et al, 1992 Quinn et al, 1996 Catanese et al, 1998 Albert et al, 2006 Nishiyama, 1999 Albert et al, 2007 Aharonian et al, 2007 Aharonian et al, 2005 Aharonian et al, 2008 Acciari et al, 2008 Chadwick et al, 1999

source redshift type flux in Crab unit photon index Γ shortest variability discovery paper RGB J0710+591 H 1426+428 1ES 0806+524 1ES 0229+200 PKS 1424+240 H 2356-309 1ES 1218+304 1ES 1101-232 1ES 0347-121 1ES 1011+496 S5 0716+714 3C 66A 3C 279 PG 1553+113 0.125

0.129

0.138

0.1396

0.16 ?

0.1671

0.182

0.186

0.1880

0.212

0.31

0.444 ?

0.5362

>0.25

HBL HBL HBL HBL IBL HBL HBL HBL HBL HBL L / IBL IBL FSRQ HBL 0.02

0.19

0.018

0.018

0.023

0.08

0.023

0.02

0.07

0.06

0.034

2.8

3.5

3.6

2.5

4.6

3.09

3 2.94

3.1

4 3.5

4.1

4.11

4.5

- days - - month - year year - months day day year

Ong, 2009 Horan et al, 2002 Acciari et al, 2009 Aharonian et al, 2007 Ong, 2009 Aharonian et al, 2006 Albert et al, 2006 Aharonian et al, 2006 Aharonian et al, 2007 Albert et al, 2007 Teshima et al, 2008 Acciari et al, 2009 Albert et al, 2008 Aharonian et al, 2006

The VHE sample

(July 2009)

• 25 blazars : - 19 HBL

(High-frequency peaked BL Lac)

- 4 IBL and 1 LBL

(Intermediate and Low-frequency peaked BL Lac)

- 1 FSRQ

(Flat Spectrum Radio Quasar)

• 2 (or 3) radio galaxies (+ 1 G.C. ?) Number of TeV sources per type : highly peculiar !

Redshifts : from 0.00183 to 0.536 (+ 3 uncertain) TeV variability : already seen in 18 sources (despite poor temporal coverage) « Shortest observed time scales » minutes : 3 sources (flares) day : 6 sources week : 1 source month : 3 sources year : 5 sources

flux-z

12000 10000 6000 2000 0 0 10 0.2 0.4

z

50 Fluxes in Crab units versus redshift (high and low states)

flux-z

Distribution versus redshift at lower fluxes 600 500 300  No clear trend with distance 100 0 0 10 0.2 0.4

z

A highly biaised sample

• Current dynamical range of ACT ~ 5000 • Active states of TeV AGN : x 10, x 200 • Doppler boosting : x 10 4 , x 10 6  Entirely biaised towards strong boosting : 100% of boosted sources (except possibly FR I ?)  Biaised towards active states : due to sensitivity limits and to strategy of observation (VHE and multi-lambda alerts)  Possibly biaised towards low redshifts : due to strategy of observation, to optimize detection probability despite EBL absorption effect.

Absorption of VHE gamma-rays by the EBL, Extragalactic Background Light

γ VHE γ EBL  e + e , dominant absorption, where σ is maximal for ε ~ (500 GeV / E) eV i.e. opt/IR photons for VHE gamma-rays Photon spectrum : Φ

obs

(E,z) = e

τ γ (E,z)

Φ

em

τ

γ

(E,z) = optical depth Г

obs

(z) ~ Г

em

+ τ

γ

(E,z)

~ E -2

Observed spectral index Γ versus redshift

Γ obs

Domain expected for Г em ~ 2.4 and EBL model from Stecker et al, 92 Domain expected for a photon-axion oscillation scenario

z

from De Angelis et al, 2009

Observed spectral index Γ versus redshift

(July 2009)

Γ obs z

No apparent increase of Г with z above z > 0.2

Does not follow the expected standard trend.

Even below « Axion-Like-Particle scenario » …?

Observed spectral index Γ versus redshift

(July 2009)

Hardening of emitted spectral with redshift : possible if VHE spectrum hardens with increasing fluxes (as observed in Mkn 421, Mkn 501, and PKS2155-304), and if high states are more often detected at high z but IC bump enters more into the Klein-Nishina regime and Г em remains basically unchanged Stronger relativistic shocks at high z would imply smaller index Г em but why stronger shocks at high z ?

Evolution of moderate photon absorption inside the source  change of Г em but why absorption evolves with z ?

Entirely different emission scenarios for high z « bright » sources compare to low z « fainter » sources ?

but it is Doppler boosting and activity level which dominate the apparent fluxes, not the z ! 

Open question, requires further analyses

Curvature in VHE spectra

power-law E cut log(E) In several cases, spectra are known to deviate from a power law in their high energy part. This is better seen during active states : Mkn 421, Mkn 501, PKS 2155-304 Origin not yet identifed ( γ

max , KN, EBL …)

Such effects come to limit the validity of any pure power-law studies. Biais also due to observed spectral ranges ?

Curvature also seen in the Galactic Center VHE source, possibly the nearest weak « AGN ».

PKS2155 Big Flare HESS G 1 power-law E break G 2 log(E)

The VHE AGN « non-sample »

Targeted sources

: Several tens of upper limits already obtained by present ACT. Sometimes constraining the SED (ex: 3C454.3).

Untargeted sources

galactic plane survey : in the field of view of observed sources and 694 nearby AGN < 100Mpc observed area Example with HESS (2004-2008) : Upper limits of 1-10% of the Crab flux for 61 untargeted AGN

Study by Herr, Hofmann, HESS paper in preparation

Outline

• Introduction • The sample of AGN detected at TeV • Similarities and discrepancies : some specific cases of blazars • An emerging family of TeV AGN : the radiogalaxies M87 and Cen A • Remarks on SSC scenarios

The HBL PKS 2155-304

Quiescent state

Various hadronic and leptonic models can often fit present available spectra of HBL

in stationary state

The HBL PKS 2155-304

Extremely active state

2 nd flare 1 st big flare Monitoring an extraordinary active state of PKS 2155-304 in 2006, detected by HESS + multi-lambda campaign. 

Variability down to minute time scale Emitting zone smaller than R g or very high bulk Lorentz factor

The HBL PKS 2155-304

Extremely active state : 1 st flare

Example of modeling light curves and SED by time dependent SSC scenario, with 5 compact components in jet with slightly different parameters C 2-6 + a more extended slowly evolving component C 1

(from Katarzynski et al, 2008)

The HBL PKS 2155-304

Extremely active state : 2 nd flare

from A. Zech, ICRC, 2009

The HBL PKS 2155-304

Extremely active state : 2 nd flare

Fit of the 2 nd flare by SSC time dependent modeling : Reproduces light curves and spectra in X and gamma rays

(from J.P. Lenain)

Variation of VHE spectral index over years highly correlated with VHE flux variation but : - Harder spectra for higher flux during active state - Steeper spectra for higher flux during quiescent state.

Variation of VHE spectral index during the 2 nd flare highly correlated with VHE flux variation : - Harder spectra for higher flux

The blazars W Comae (IBL) and BL Lac (LBL)

BL Lac High variability and broad band spectra  Stringent necessity of coordinated HE and multi-lambda monitoring to constrain SED and evolution.

The FSRQ 3C279

Flat spectrum radio-quasar at z=0.536

Brightest EGRET source. Highly variable, fast variability (~6 hours) MAGIC observed it in 2006 for 9.7 hours in 10 nights : clear detection

First FSRQ in TeV gamma-rays and highest z

Strategy of multi-lambda triggers KVA optical telescope at la Palma Mkn 180 z = 0.045

Optical ToO trigger

1ES 1011+496 z = 0.212

ToO trigger

S5 0716+714 z = 0.31

MAGIC

S5 0716+714 MAGIC PRELIMINARY Significance 6.8 σ

The region of 3C66 A and B A : an IBL B : a radiogalaxy

MAGIC VERITAS 2 sources separated by only 0.12 ° : - MAGIC favours a detection of 3C66B in 2007, and excludes 3C66A - VERITAS favours a detection of 3C66A in 2008, and excludes 3C66B

3C66A

: Detection of a flare by VERITAS and Fermi Simultaneous observations of the VHE flare with Fermi/LAT

(Reyes et al, ICRC, July 2009)

 Firmly establishes 3C 66A as VHE source SED is better fitted by a SSC+EC scenario  Either it was 3C66A all along, or both sources A and B varied in opposite way between 2007 and 2008

Outline

• Introduction • The sample of AGN detected at TeV • Similarities and discrepancies : some specific cases of blazars • An emerging family of TeV AGN : the radiogalaxies M87 and Cen A • Remarks on SSC scenarios

The radiogalaxy M87 TeV M87

 : TeV day variability 3 possible emitting zones : - The peculiar knot HST-1 at ~ 65 pc from the nucleus - The inner VLBI jet - The central core and the black hole environment

HST X radio

Monitoring of the core of M87 by VLBA at 43 GHz every 5 days Explore sub-mas scale (0.21mas x 0.43mas) to probe the jet collimation region Significant rise at in core at the time of VHE activity and enhanced emission along inner jet

(Wagner et al, ICRC, 2009; Science, 2009)

0.5 mas = 0.04 pc = 140 R s

X-ray light curve of HST-1 obtained by Chandra in 2008 does not follow the TeV one Correlated core emission (radio, X and VHE)  favours scenarios with TeV emission from inner jet or central core.

Inner jet

: Spine-layer SSC+EC scenario

(Tavecchio, Ghisellini, 2008)

Multi-blob SSC scenario at 100R s

(Lenain et al, 2008)

Core

: Acceleration in Black Hole magnetosphere

(Neronov, Aharonian, 2007; Rieger, Aharonian, 2008; Istomin, Sol, 2009)

The radiogalaxy Cen A

• Recent discovery of VHE emission from

Cen A

(Aharonian et al, 2009)

with HESS • Together with M87,

establishes radio galaxies as a new class of VHE emitters

• Three different types of AGN now detected at VHE (blazars, radiogalaxies, and weak AGN as Galactic Center) 

is VHE emission a general feature of AGN (and SMBH) ?

Richness of the extragalactic space at VHE, to further explore with MAGIC II, HESS II and later on with CTA / AGIS

Cen A 5 σ

Possible VHE zones ?

- BH magnetosphere - base of jets - jets and inner lobes - pair halo in host galaxy Link to UHECR ?

Origin of the VHE emission ? Compatible with radio core and inner kpc jets of Cen A

Ex : SSC emission from jet formation zone (Lenain et al, 2008)

Starburst galaxies and ULIRG :

TeV detection still difficult but flux upper limits approaching theoretical predictions 

Recent detection of M82 by VERITAS and of NGC 253 by HESS

(ICRC, July 11th, 2009) AGN-starburst connection ?

Chandra HST

Collision, star formation, superwind …

Arp 220 (ULIRG) NGC 253

Outline

• Introduction • The sample of AGN detected at TeV • Similarities and discrepancies : some specific cases of blazars • An emerging family of TeV AGN : the radiogalaxies M87 and Cen A • Remarks on SSC scenarios

M87

Size constraints from variability

PKS2155-304 Mrk 501 • TeV variability of M87, PKS2155-304, Mrk501 … requires very small emitting zone, of the order of a few r g or even less (even for high causality argument. δ) under •

Challenge

to efficiently accelerate particles in such small zones (core around BH, or very inner jet). • Fermi processes in shocks and turbulence  widely invoked • Alternatives : magnetic reconnection, direct electric forces, centrifugal force

Applying simple SSC scenarios to SED and light curves

Assuming given a particle distribution with broken power-law : N e ( γ) = K 1 γ -n 1 for γ min < γ < γ break and N e (γ) = K 2 γ n 2 for γ break < γ < γ max 1) High energy SED of stationary states of HBL can be well reproduced by one-zone scenario with typical parameters : B ~ 0.1 – 0.2 G Size of emitting zone R ~ 10 14 Doppler factor ~ 20 – 50 Density factor K 1 Index n 1 ~ 2 (1.5 – 2.5) Index n 2 ~ 2.7 ~ 10 – 4.7 3 – 10 4 – 10 cm -3 16 cm 2) Active variable states of HBL can be reproduced by SSC time dependent models with at least two types of SSC emitting zones (compact components + an extended slowly varying one) 3) IBL, LBL and FSRQ : SSC can work, but EC contribution helps. 4) Radiogalaxies : need SSC multi zone models. Seems to possibly work but still several options. Open issue. Different situation for M87 and Cen A.

Domain of parameters ?

Index of particle distribution can significantly differ from canonical values in relativistic shocks

(Stecker et al, 2007) (Baring, 2004)

Spectral index sensitive to compression ratio r, upstream flow, obliquity of the shock, nature of scattering, strength of turbulence and anisotropy of the diffusion.

Parallel ultrarelativistic shocks, index  OK for n 1 . For n 2 2.25 : corresponds to r ~ 3 and small scattering : oblique relativistic shocks ? more inefficient. (needs t acc < t cool )

Further observing AGN at VHE • TeV observations

disentangle non-thermal effects from thermal ones

possibly present at others wavelengths  provide a simplified view of the physics at the highest energies.

• Explore

variability at the shortest time scales

 jet physics, particle acceleration and radiation processes  physics of supermassive Black Hole environnement; accretion physics  build a sample of sources at different redshifts (evolution, validity of Lorentz invariance, analysis of the EBL).

• Gather samples of different AGN types to allow unification schemes (are HBL highly beamed LBL ?). Check the ‘blazar sequence’, probe the quiescent states

statistical studies

for • Look for VHE emission from « dormant » BH or « dead » quasars. Studies of

AGN and SMBH evolution

, AGN feedback and co-evolution with host galaxies.

•

Importance of multi-messenger and multi-lambda analyses.

----- Complements -----

A limited number of VHE γ -rays emission mechanisms : 2 (+1)

• Leptonic scenarios relativistic electrons : synchrotron and Inverse-Compton (positrons) e + B  (IC) radiation of e + B +

γ

, in magnetic field B (also X-rays) e +

γ 0

 e +

γ

, with h ν ~ min [

γ

e 2 h ν 0 ,

γ

e m e c 2 ] , IC on synchrotron emission (SSC) or on external photon field (EC) • Hadronic scenarios : Interaction of energetic radiation backgrounds protons (CR) with local gas and p + p p +

γ

  N + N + n 1 ( π + + π ) + n 2 π 0 ( N = p or n) p + π 0 , n + π + , others (for

γ

p h ν > m π c 2 ) ; or p + e + Then decay

π 0

+ Decay pions  

2 γ produce VHE photons with E γ

muons 

~ E

π + e (for

γ

p h ν > 2m e c 2 )

/2 ~ 10% E

secondary electrons and neutrinos

p,i

(also X rays) Alternatives : curvature and synchrotron radiation of VHE protons. • (Annihilation of Dark Matter particles theories, Kaluza-Klein scenarios  : predictions of supersymmetric open questions to explore. No detection yet. A great challenge, but not yet granted !)

EBL : optical depth

3C279 and the gamma-ray horizon

Recent inner jet scenarios

(adapted from standard TeV models for HBL)

• Spine-layer scenario

(Tavecchio, Ghisellini, 2008)

Assumes a structured jet with two components - a fast core emitting radio to X-ray/GeV photons , surrounded by - a slower sheet emitting TeV photons , which can dominate at large viewing angles SSC + EC (spine/layer) emission Turbulent acceleration in layer ?

2 sets of free parameters, but it reproduces SED of LBL at small angles GLAST limit 1yr B ~1G (0.3G) in spine (layer) Г ~ 12 (4) in spine (layer)

TeV high state in 2005 ‘Low state’ B ~ 10 mG at 100 R g Г ~10 θ ~ 15° M87 : Differential Doppler boosting in emitting zone SSC emission. Multi-blob scenario (Lenain et al) Electron distribution described by a broken power law Fitting low and high TeV states with inner jet parameters reasonably well consistent with values from GRMHD simulations of McKinney, 2006 (used as initial constraints).

Galactic Center Point like source: intrisic size < 1.2’ (≈ 2.9 pc) at the 99% C.L.

 position: l=359 ° 56’41.1’’ ± b=-0 ° 2’39.2’’ ± 6.4’’ 5.9’’ ± ± 6’’ 6’’  centroid emission located at 7’’ ± 12’’ from Sgr A*  Sgr A East excluded at the 7 s C.L.

 G359.95-0.04 still inside error bars (8.7’’ from Sgr A*) SNR SgrA East (90 cm) van Eldik et al., ICRC (2007) from M. Vivier

 MWL source in the central parsecs of our Galaxy: emitting from radio to TeV  rays  From radio to X-rays: highly variable, originates from the SMBH Sgr A*  Origin of the hard X-rays/Tev  -rays?

X-rays Radio IR Sgr A* ?

TeV