Longterm Xray Variability in Blazars its Multiwaveband Context
Long-term X-ray Variability in Blazars & its Multiwaveband Context Alan Marscher Boston University Research Web Page: www. bu. edu/blazars Free Downloads! - Monthly 43 GHz VLBA images of 29 blazars - Original songs (MP 3 format) on science topics
Main Collaborators: Major International Effort Svetlana Jorstad, Ritaban Chatterjee, Frannie D’Arcangelo (Boston U. ) Vladimir A. Hagen-Thorn & Valeri Larionov (St. Petersburg State U. , Russia) Margo & Hugh Aller (University of Michigan) Ian Mc. Hardy (U. Southampton, UK) José Luis Gómez & Iván Agudo (IAA, Granada, Spain) Anne Lähteenmäki & Merja Tornikoski (Metsähovi Radio Observatory) Walter Gear (Cardiff U. , UK) Esko Valtaoja (Turku U. , Finland) Thomas Balonek (Colgate U. ) Gino Tosti (U. of Perugia Observatory, Italy) Omar Kurtanidze (Abastumani Observatory, Rep. of Georgia) Martin Gaskell (U. of Texas) H. R. Miller, Kevin Marshall, Wesley Ryle (Georgia St. U. ) Matthew Lister (Purdue U. ) Thomas Krichbaum, Lars Fuhrmann, Tuomas Savolainen, Yuri Kovalev (MPIf. R) Funded by NASA & NSF
Main Instruments - A Truly Multiwaveband Effort Lowell Obs. Perkins Telescope: Prism (optical) & Mimir (H-band) (polarimetry) Calar Alto 2. 2 m telescope (optical polarimetry) Liverpool Telescope (optical & near-IR light curves & polarimetry) Very Long Baseline Array (7 mm) & Global mm VLBI Array (3 mm) Crimean Astrophysical Observatory 70 cm telescope (polarimetry) St. Petersburg State University 0. 4 m telescope (polarimetry) Rossi X-ray Timing Explorer, XMM-Newton, & Swift (X-ray light curves) Chandra, HST, Spitzer (imaging of kpc-scale jets) GLAST (gamma-ray light curves) IRAM 30 m (3 mm polarimetry) University of Michigan Radio Observatory (radio light curves & polarimetry) Metsähovi Radio Observatory (radio light curves) Numerous other optical telescopes & collaborators
Relativistic Jets on Parsec & Kiloparsec Scales 3 C 111 Scale: 1 mas = 0. 92 pc = 3. 0 lt-yr (Ho=70) Approximate location of black hole mas
Goal: Probe Blazar Jets as far Upstream as Possible We want to probe the section of the jet from the mm-wave core down to the central engine - where we think the jet is accelerated & collimated We also want to study shocks & other disturbances, particle acceleration, changes in flow Lorentz factor, magnetic field geometry, turbulence, interaction with surrounding medium, bending/precession Unbeamed Beamed
The Quasar PKS 1510 -089 (z=0. 361) X-ray/Radio/Optical Light Curves Arrows: Times of superluminal ejections at 43 GHz Major outbursts usually seen at all wavebands, coincident with ejection of superluminal radio knot Which wavebands lead and lag is time dependent Need robust method for determining correlations & their uncertainties & connections with events seen in VLBI images of jet
Method for Probing Jet Upstream of Core: Identify Jet Features Responsible for High-frequency Variations Look for correlated polarization variability in optical, IR, submm, & mm-wave Locate sites of variable high-frequency emission on mmwave VLBA images relate optical/near-IR light curves to images Connect radio-optical emission to X-ray & -ray radiation by correlating light curves (RXTE, Swift, GLAST) with optical/near-IR, & mm-wave light curves Dynamic multiwaveband emission maps
The Quasar 3 C 279 (Chatterjee et al. 2008) Extreme variations in flux at optical & X-ray (& γ-ray) Variations in X-ray flux appear well-correlated with changes in optical brightness Cross-correlation analysis of 11 years of X-ray, optical, & radio (15 GHz) monitoring (Uttley et al. ) 1. Compute raw PSD 2. Create many simulated light curves with random variations following raw PSD with sampling same as observed 3. Determine “true” PSD + error on slope red noise spectrum 4. Use true PSD to create many simulated light curves with sampling same as observed 5. Use discrete cross-correlation method to determine correlation of light curves at different 6. Use simulations to determine probability that correlations could be obtained by chance
The Quasar 3 C 279 (cont. ) Variations in X-ray flux well-correlated with changes in optical brightness → time lag of 15± 6 days, with optical leading Discrete Cross-Correlation Coefficient (optical vs. X-ray) Optical/X-ray correlation varies on timescale of few yr - X-ray leads by up to 20 days (1995 -1997) Positive delay: optical leads - Optical leads by up to 20 days (1999 -2002) - ~ zero lag (1997 -1999, 20022004) - Weak, broad correlation (20032007)
The Quasar 3 C 279 (cont. ) Variations in X-ray flux well-correlated with changes in jet direction → time lag of 80± 150 days, with X-ray flux lagging Direction of jet determines mean strength of next flare Discrete cross-correlation coefficient (X-ray vs. PA of inner jet, 0. 2± 0. 1 mas from core)
The Quasar 3 C 279 (Chatterjee et al. 2008) Cross-correlation Analysis highly significant correlations between light curves: • X-ray - optical (time lag varies between +20 and -20 days) • X-ray - radio core and X-ray peaks superluminal ejections (X-ray leads by 114± 21 days) Profiles of optical flares narrower than X-ray (higher E electrons make optical) Optical-leading flares: shocks or SSC light-travel delays X-ray-leading flares: gradual acceleration of electrons X/opt radiative E output ratio: ~1 when delay ~ 0 (upstream) Low when delay longer (downstream) » » X-ray & optical emission occurs in ≥ 2 sites upstream of 43 GHz core
BL Lac Reveals its Inner Jet (Marscher et al. 2008, Nature, 24/04/08) Late 2005: Double optical/X-ray flare, detection at Te. V energies, rotation of optical polarization vector during first flare, radio outburst starts during 2 nd flare Strong Te. V detection Superluminal knot coincident with core Scale: 1 mas = 1. 2 pc Te. V data: Albert et al. 2007 Ap. JL X-ray Optical EVPA matches that of knot during 2 nd flare P decreases during rotation B is nearly circularly symmetric
Physical Picture of BL Lac: Exactly as Expected BL Lac: Physical Theoretically* (It’s Picture a Miracle!) Moving shock follows spiral streamline (does not cover entire jet cross-section) Passes through helical field pattern in acceleration + collimation zone of jet Becomes bright shortly before it emerges into zone of turbulent plasma *Vlahakis (2006, in Variability of Blazars: Entering the GLAST Era) - Similar to model for OJ 287 by Kikuchi et al. 1988 A&A, 190, L 8 & Sillanpää et al. 1993, Ap. SS, 206, 55 - Polarization rotation was smooth not turbulence as in 0420 -014 (D’Arcangelo et al. 2007, Ap. JL)
BLLac in Late 2005: What Caused the Flares Interpretation: First optical/X-ray/Te. V flare: knot Lorentz factor increases up to peak Flare from increase in Doppler beaming - X-rays: synchrotron - No radio flare 2 nd flare (includes radio): knot interacts with standing shock in 7 mm core - X-rays: SSC - Major radio flare
Interpretation of BL Lac Observations Shock follows spiral streamlines until it exits region of coiled magnetic field 1 st flare occurs just before shock exits acceleration & collimation zone 2 nd flare caused by moving shock interacting with standing shock(s) in core site of 1 st flare site of 2 nd flare
Sketch of Physical Structure of Jet, AGN • Magnetic acceleration zone with helical field gives way to turbulence (in spine of jet) Transverse velocity gradients shear magnetic field, partially align it parallel to jet • Outbursts of radiation - including X- and -rays - occur in disturbances (probably shocks that become superluminal knots) as they pass through: (1) end of acceleration zone, (2) standing shock(s) farther downstream (e. g. , the “core”), or (3) locations (time dependent) where jet bends closer to line of sight
Radio galaxies show a connection between X-rays from central engine region & activity in jet, ~ as in microquasars Sequence of VLBA images (Marscher et al. 2002) Scale: 1 mas = 0. 64 pc = 2. 1 lt-yr The FR I Radio Galaxy 3 C 120 (z=0. 033) (Ho=70) HST image (Harris & Cheung) • Superluminal apparent motion, ~5 c (~2 milliarcsec/yr) • X-ray spectrum similar to Seyferts • Mass of central black hole ~ 5 x 107 solar masses (Marshall, Miller, & Marscher 2004; Peterson et al. 2004)
X-Ray Dip/Superluminal Ejection Connection in 3 C 120: Older data Superluminal ejections follow Xray dips by ~ 60 days Somewhat similar to microquasar GRS 1915+105 3 mm core must lie at least 0. 4 pc from black hole to produce the observed X-ray dip/superluminal ejection delay of ~ 60 days Marscher et al. 2002, Nature, 417, 625
X-Ray Dips/Ejections in 3 C 120: Data since 2002 Superluminal ejections still follow X-ray dips by ~ 60 days Anti-correlation between X-ray flux or spectral index & 37 GHz flux delay of 62 days
Comparison of GRS 1915+105 with 3 C 120 Light Curves + BH mass of 3 C 120 ~4 x 106 times that of GRS 1915+105, so timescales of hours to months in 3 C 120 are similar to the scaled-up quasi-periods (0. 15 to 10 s) of 1915+105 +Typical fractional amplitude of dips is also similar - Random fluctuations higher amplitude in 3 C 120 - Long, deep dips not yet seen in 3 C 120, jet not seen during this stage of 1915+105 blowup 150 s of blow-up should scale up to ~17 yr in 3 C 120 if timescales Mbh GRS 1915+105 over 3000 s on 9/9/97 Light curve (top) & PSD (bottom) (Taken from Markwardt et al. 1999 Ap. JL) Above: X-ray light curves of GRS 1915+105 over 150 s & 3 C 120 over 5 yr
FR II Radio Galaxy 3 C 111 (z=0. 0485) Seems to Do the Same May 2004 1 milliarcsec August 2004 New knot Ejection of bright superluminal knot follows start of X-ray dip by 0. 35± yr 3 mm core must lie at least 0. 6 pc from black hole to produce the observed X-ray dip/superluminal ejection delay
FR II Radio Galaxy 3 C 111 X-ray/Optical Correlation X-ray variations lead optical by 20 days - as expected if X-rays come from corona near black hole & optical from farther out in accretion disk
Connection between Disk + Jet Events Co ro na Possible scenarios of X-ray dip & hardening + faster jet flow → shock down jet: 1. Wind → jet; faster wind off disk → less dense corona, weaker X-ray, faster jet 2. B from ergosphere makes jet; fewer particles injected at base → weaker X-ray (if corona = base of jet), faster jet downstream (borrowed idea from G. Ghisellini) Flatter X-ray spectrum because downstream jet contributes higher fraction of X-rays when X-rays from base are weaker
• Summary X-ray emission in the jet can occur in 2 or more sites: outer part of jet’s acceleration/collimation zone & standing shocks (or bends) • At least some outbursts, flares, & pairs of flares come from superluminal knots • Time delays between optical & X-ray sometimes consistent with shock model, sometimes with gradual e acceleration • 3 C 120 and 3 C 111 behave in a manner not ridiculously different from GRS 1915+105 except that (1) we see no QPOs (2) we have yet to observe a prolonged low-hard state (3) the stage of 1915+105 with X-ray light curve most closely resembling 3 C 120's has no observable jet The physical connection between disk/corona/ & jet doesn't simply scale with black hole mass
BL Lac in Late to Optical EVPA Rotation BL 2005: Lac: Fit to EVPA Rotation Curvature of rotation: uniform rotation in source frame, axis at an angle of 45 o Can explain if streamline direction is aberrated (jet axis 7. 7 o from l. o. s. ; Jorstad et al. 2005 J 05) Requires Lorentz factor ~ 7 (same as derived from knots by J 05) Magnetic helical pattern cannot be aberrated (Poynting flux advection speed not highly relativistic) without causing dominant polarization parallel to axis - consistent with stationary helical field pattern expected in MHD launched jets
- Slides: 25