Two exceptional γ -ray flares of PKS and their implications on blazar and fundamental physics Rolf Bühler MPIK Heidelberg KIPAC Tea Talk SLAC 9th December 2008
H.E.S.S. Telescopes ● Located in Namibia (1800 a.s.l.) ● Energy threshold of ~150 GeV ● Sensitivity of 5 σ detection of 5% crab in one hour ● Point spread function of ~0.1 o ● Energy resolution of ~15%
Blazars at VHE 25 detected at VHE (E>100 GeV) to date: ● All (but four) High Frequency Peaked BL Lacs ● Typical flux of ~2% crab ● Mkn 421, Mkn 501, PKS brighter (~10% crab) From R. Wagner
Synchrotron Self Compton Preferred model to explain EM emission: ● X-rays and gamma-rays correlate ● Synchrotron dominated Radio to X-ray emission Inverse Compto n E E 2 dN/dE Synchrotron ~eV ~Ge V
PKS in ● Observed for 74.4 hours with excess events Typical flux of I 0 = 4 × cm -2 s -1 (compatible with data) Runwise (28 min) flux Excess events
Lightcurve in nightly flux monthly flux Close to I 0 in Highly active in 2006 with extreme outburts in July JuneJulyAugustSeptember
Exceptional flares on 28 and 30 July with flux values more the >100 × I 0 and fast flux variability Lightcurve in nightly flux JuneJulyAugustSeptember Big Flare Chandra Flare runwise flux
The Big Flare Five sub-bursts with rise/fall times of ~200s → Causality connects Doppler factor and size of emitting region: 1 min bins F. Aharonian et al. Arp. J. L. 664 (2007) 10×I 0
The Chandra Flare 4 min bins Unprecedented MWL coverage: ● High correlation between X-rays and rays ● Optical flare starts at the same time and evolves slower
The Chandra Flare ● Large Compton dominance during the flare (L γ / L x ~5) ● Increasing variability with energy: 20× at γ- rays, 2× at X-rays, 1.15× at optical Max. sim. Min. sim. Max. VHE 2003
The Chandra Flare 7-14 min bins High flux and spectral correlation, cubic flux correlation during decreasing phase VHE X- ray FγFxFγFx a → No one zone SSC [1] → Likely several X-ray emitting regions [1] K. Katarzynski et al., A&A (2005)
Summary PKS underwent two exceptional VHE flares in July 2006: ● Peak flux values >100× more then typical ● Fast flux variability (~200 s) Multi-wavelength observations during the second flare reveals: ● A cubic decrease of the VHE with X-ray flux ● The inverse Compton component dominates the energy output → No one zone SSC → Likely several X-ray emitting regions No time delays seen between X-rays and γ - rays Energy spectrum hardens at VHE with increasing flux Inverse Compton SED peak shifts to higher energies during flare Publication in preparation..
Is the speed of light constant? “Light is always propagated in empty space with a definite velocity c which is independent of the state of motion of the emitting body” Is the speed of light constant?
Are you sure? “Light is always propagated in empty space with a definite velocity c which is independent of the state of motion of the emitting body” Is the speed of light constant?
Several Quantum Gravity models have predicted energy dependence of the speed of light. General parametrization:, with Leads to dispersion of poly-energetic signals: Is c energy dependent ? Linear:Quadratic:
Measurable effect ? Time Requirements: Long distance, fast variability → GRB's (X-ray) and AGN (VHE) Caveat: Cancellation due to source intrinsic effect (unlikely if no dispersion) → Population studies High Energ y Low Energ y Measurable effect?
Current limits ( population studies: ) Time of flight measurements GRB's [1,2] : AGN's [3] : Vacuum Birefringence Radio Galaxies [4] : Modified thresholds UHECR [5] : Crab Nebula [6] : ( for electrons ) ( if sign negative ) ( if sign helicity dependent ) [3]Albert et al., Phys. Lett. B (2008) [4]Fan et al., MNRAS (2007) [1]Boggs et al., Ap. J. L. (2004) [2]Ellis et al., Astr. Part. Phys (2006) [5]Galaverni et al., PRL (2008) [6]Maccione et al., JCAP (2007) Current limits
GeV >800 GeV Oversampled, 2-min bins ΔE mean = 1 TeV Probing c(E) with the Big Flare
Dispersion measurement RMS = 28 s 21% → < 73 s TeV -1 ( 95% confidence ) τ peak = 20 s Dispersion measurement
Systematic tests Recovery of injected dispersion in data Confirmed with “bootstrap” simulations & independent analysis Systematic tests
Limits For more details see Aharonian et al., Phys. Rev. Lett. 101 (2008) Dispersion limit translates into: Limit s ● Most stringent linear limit from time of flight measurements ● Finally rule out possibility of time delay cancellation requires population studies Linear: Quadratic: Reminder:
Summary & Conclusions The absence of dispersion during the first flare results in stringent limits on speed of light modifications: Summary PKS underwent two exceptional VHE flares in July 2006: ● Peak flux values >100× more then typical ● Fast flux variability (~200 s) Multi-wavelength observations during the second flare reveals: ● A cubic decrease of the VHE with X-ray flux ● The inverse Compton component dominates the energy output → No one zone SSC → Likely several X-ray emitting regions
Backup slides
Correlations in SSC IC E E 2 dN/dE Synch. E E 2 dN/dE electron s photon s Thomson regime (electrons upscatter their own synchrotron): F x n e (E x ) F γ n e (E γ ) n ph n e (E γ ) n x (E x ) F γ F x 3 SS C F γ F x Klein-Nishina regime in decaying phase (electrons upscatter synchrotron from lower energies) : n ph = const. → F γ F x [1] K. Katarzynski et al., A&A (2005) [1] 2 Generally:
Simplest solution X-ray emission from (at least) two regions:
Blazar population [1] Biller et al., PRL (1999) [2] Albert et al., Phys. Lett. B (2008) No redshift dependent trend (yet) [1] [2] Blazar population
Derived from s lag between and TeV [1] Biller et al., PRL (1999) [2] Albert et al., Phys. Lett. B (2008) [3] Albert et al., Ap. J. (2007) [1] [2] [3] Blazar population
Simulation of bin correlations ● Throw random numbers on a fine grid Add and multiply with measurement error on coarse grids Simulation of bin correlations