Multiwaveband Variability of Seyferts and LINERS Relationship to X-rays Ian M c Hardy University of 1
Outline X-ray / UV / Optical variability X-ray / Radio variability 2
X-ray / UV/ Optical Variability What causes UV/optical variability in Seyferts? Reprocessing of X-rays? - X-rays lead uv/optical by short (hour-days) light travel time. Expect Lag Wavelength 1.33 Intrinsic disc variability X-rays lag: two possibilities Long lag (months), viscous propagation timescale for perturbations to reach X-ray region from optical in disc Short lag (hour-day), light travel time of UV seed photons to corona 3
NGC 4051: optical X-ray 4
NGC 4051 5
NGC4051 Optical lags by 1.5+/- 0.5 d (above 99% confidence) Breedt et al 2010 6
MKN 79 (Breedt et al, 2009, MNRAS) Long timescales (years) uncorrelated behaviour. Intrinsic disc variations in optical? Short timescales (days-weeks) - well correlated. Hint that optical lags, but lag not well defined 7
Swift UV / optical interband lags Line is Lag Wavelength 1.33 NGC2617 Shappee et al 2014 Between 20 50 data points per band Fit to UV-optical with fixed power law slope of 4/3 does not go through X-ray point 8
Swift Monitoring of NGC5548: (> 500 observations) Good correlation, but not perfect, eg large W2 rise after day 6480 McHardy et al, 2014, MNRAS, 444, 1469 9
Lag of X-rays by UVW2 Mean-subtracted lightcurves Intensively sampled period Lag distribution (Javelin Zu et al 2011) 10
NGC 5548 All Swift Bands 11
Lags as function of wavelength Lag Wavelength 1.23 Expect 4/3 power for Shakura-Sunyaev disc. So good agreement. Fit goes through X-ray point Model lags (McHardy et al, 2014) BUT observed lags are longer than expected for the Mass and m Red line is time for HALF of reprocessed light to arrive. Microlensing obs (eg Morgan et al 2010) also require larger disc than SS model Hotter than expected disc (eg higher Inhomogeneous disc (Dexter and Agol 2011)? m, higher Lx)? Time associated with the reprocessing mechanism? Same result in extensive follow up observations (Edelson et al 2015, Fausnaugh et al 2015) 12
XMM and ground based lags for NGC4395 Mass 3.6x10 5, x100 lower than NGC5548, and x10 lower Does it also have longer lags than predicted? m UVW1 with OM Ground based g-band monitoring around globe (Peterson, Elvis, Uemura Chand, Netzer, Kaspi and others) (Connolly et al. in prep) 13
NGC4395: Swift X-ray / UVW2 Looking within individual Swift visits (TOO 12 ks) uvw2 lags X-rays by ~400s 14
XMM and ground based monitoring of NGC4395 X-ray XMM UVW1- XMM OM g-band ground 30-31 December 2014 15
NGC4395 - DCFs X-rays vs UVW1 X-rays vs g-band (Using Emmanouloupolos et al 2013 improved lightcurve simulation method for simulations) 16
NGC4395 Javelin lags UVW1 lags X-rays by 473 (+47, -98) s g-band lags X-rays by 788 (+44, -54) s 17
NGC4395 Lags vs Wavelength Lags very short. Intrinsic reprocessing time must be very short. Simple linear fit (red) is best fit (forced through zero). However powerlaw of index 4/3 (blue) is also acceptable So reprocessing dominates here also, but agreement with Shakura-Sunyaev models is closer than for NGC5548 does disc structure depend on M, mdot, Temp? 18
NGC4395 - Models observations model emission profiles Solid lines total disc energy release in band, including X-ray contribution Dashed lines gravitational energy release Observed lags correspond to peak emission radii (models from P. Lira) 19
Dependence of reprocessing on disc temperature: X-ray / UV Variability of NLS1s DCF Swift UVW2 X-rays Cameron et al, in prep Poor correlation 20
X-ray / optical correlation strength depends inversely on disc temperature From Cameron PhD thesis; also Breedt Phd thesis 21
LINERS M81
M81 NuSTAR NuSTAR 1000s bins 4d observation With Andy Young, Bristol 23
LINER M81 NuSTAR Spectrum Disc reflection very weak or absent So optical is NOT reprocessed X-rays Cut-off PL is good fit Γ=1.85 E_cut=230 kev Hot iron line 6.62 kev -ADAF? (Young, McHardy et al, in prep) 24
X-ray / UV Variability of M81 IF there is a UV/X-ray correlation, then - X-rays lagging UV: UV could be seed photons for X-rays, eg cyclo/synchrotron from corona - UV lagging X-rays: UV from synchrotron jet, downstream from X-ray corona 25
X-ray / UV Variability of M81 Swift UV X-rays Cameron 2014 Phd Thesis 26
X-ray / W1 Variability of M81 Some short timescale correspondence 27
X-ray / UV Variability of M81 W1 W2-4 -2 0 2 4 Lag of X-ray by UVW2 (days) Weak correlations, close to zero lag, possibly small UV lag Suggests UV are not seed photons for X-rays but are from downstream jet Javelin result from Harvey-Taylor UG Could just be part of same emission component 28
X-ray / UV Variability of LINERS UVW1 UVW2 X-RAYS 29
X-Ray / mm / Radio Relationship in LINERS NGC7213, M81 (More luminous radio sources than Seyferts) Do the perturbations which drive the X-rays carry on into the jet? Are liners the equivalent of `hard state X-ray binaries?
NGC7213 X-RAY/RADIO 3800 4400 5000 MJD-50000 S ~ ν + α (Bell et al, 2011) 31
NGC7213 X-RAY/RADIO 8 GHz lags X-rays by ~24 days 5 GHz lags X-rays by ~40 days 32
M81 sub-mas structure Sub-mas bending jet similar to blazars Here from Ros, McH et al; See also Marti-Vidal et al 2011 33
M81 radio-mm variability: strong correlation Radio lags by approx 3d - consistent with standard synchrotron jet 34
SMA 1.3mm vs AMI 15GHz radio Radio lags mm by a few days ie definite synchrotron jet 35
M81 Swift X-ray and AMI 15 GHz Radio Guy Pooley 36
M81 X-ray vs Radio AMI 15GHz from Pooley; OVRO from Richards, Readhead and Pearson 37
M81 X-ray / Radio ICCF / DCF ICCF DCF Good overall correlation. Perturbations (from disc?) which drive X-rays may propagate down the jet to drive radio variations. Lag of X-ray by Radio (days) Centroid of lag, using Peterson lagerror8 prog simulations 21 +/- 3d Peak of lag 44 +/- 3d (Thanks to Brad Peterson) Scaling to binaries is unclear: X-ray / optical binary lag of 0.1s (Gandhi et al 10) is consistent (assuming also linear lag scaling with wavelength) But X-ray/radio lag of 7 minutes (Wilms et al 2007) is too long. 38
M81 X-ray and Radio scaled When scaled for mass, M81 data fits on Fundamental Plane for jet dominated sources very well observed Merloni et al 2003, Falcke et al 2004, Koerding et al 2006 39
LINER M81 Lags Radio lags X-ray by approx Radio lags mm by UV lags X-rays by ~ 21d - a messy correlation, OK on long timescales ~ 3d - well defined correlation ~ 0.5d - weak correlation but lag, well defined. Long X-ray / radio lag with shorter mm / radio lag could mean synchrotron emitting part of jet starting some way from BH and X-ray emitting corona. Then UV has to be part of the X-ray emission component. Or maybe X-ray / radio lag is actually shorter, then jet could start near BH 40
Another Coronal / Jet Connection : QSO 3C273 RXTE X-ray Gamma-ray Only occasional simultaneous flares, but perhaps the Gamma-ray emission, overall, is greatest when the average X-ray level is highest?
X-ray / Gamma-ray correlation : 3C273 So there may be a small amount of co-spatial X- and Gamma-ray emission, but most comes from separate emission regions.
X-ray / Radio Variability of Radio Quiet Seyferts 43
X-ray bright Seyfert: NGC4151 Mundell et al 2003 Radio jets commonly seen in nearby Seyferts but rarely (never?) luminous or compact enough to expect significant Synchrotron Self-Compton X-rays. So is there a radio / X-ray relationship? If so, why?
Radio variability from Seyferts, ie high accretion rate, Soft State, AGN NGC5548 Wrobel 2000 - radio variability over months but no parallel X-ray observations See also Mundel et al 2009 Seyferts are thought to be the equivalent of soft state X-ray binaries. No detectable radio emission from soft state binaries Russel et al 2010 45
NGC4051 - Seyfert Looks just like a classical radio galaxy except much smaller and of much lower luminosity. (Jones et al, in prep) Component separation is ~50 light years (Girolleti and Panessa 2009)
NGC4051 VLA A array observations Radio Default beam Fixed beam Smoothed X-ray X-ray At best, only a very weak hint of X-ray/radio correlation Jones et al 2011 47
NGC4051 Radio vs. X-ray - all arrays No strong evidence for large amplitude radio variability (Jones et al, 2011) 48
NGC4051 on radio `fundamental plane for jet-dominated sources (Merloni et al 2003, Falcke et al 2004, Koerding et al 2006 Here plotting just hard state objects from Koerding etal NGC4051 is ~1 decade radio quiet Jet orientation? 49
NGC4051 as a coronal radio source? L R /L x ~10-5 If so, X-ray is integral of radio emission (Neupert effect seen in stellar coronae) Too faint to search for that effect here. From Laor and Behar 2008 50
NGC5548 AMI 15GHz Compact core (Guy Pooley) Other source is unrelated 51
NGC5548 X-ray / Radio Variability AMI 15GHz - Guy Pooley Swift RXTE 52
NGC5548: X-ray / Radio Lag Radio lags X-ray by 42 +/- 17d Too long for radio to be seed photons in corona Accretion rate fluctuations from disc, perturbing X-ray corona, could propagate down jet to radio emission 53
CONCLUSIONS SEYFERTS: Short timescale UV/optical variability is X-ray reprocessing but implied disc size in NGC5548 is ~3x too big for Shakura Sunyaev model. Are all AGN discs bigger than theoretically expected? If so, why? LINERS: Short timescale UV variability in M81 correlates weakly with the X-rays with very short lag. No disc, so j et emission? M81 radio lags X-rays by ~20d. Radio lags mm by ~3d. Possibly perturbations which drive X-rays also go down the jet. Seyferts: weak correlation of X-ray and radio with radio lagging by about 40d. Some similarities to liner M81.
Material I won t show is below
Infrared Optical Lags: NGC3783 Green is V-band, for comparison. J-band similar lightcurve to V-band, but K-band lags. V and J mostly reprocesing from disc, K includes some torus contribution Lira et al, submitted also Saganuma et al 2006 57
XMM and ground based monitoring of NGC4395 100s bins 28-29 December 2014 58
X-ray / UV Variability of NLS1s DCF Swift UVW2 X-rays Cameron et al, in prep 59
NGC4051 X-ray Spectra and radio 60
X-ray / Gamma-ray variability : 3C273 Gamma-ray X-ray Gamma-ray Much more high-frequency variability in Gamma-rays
X-ray / Gamma-ray correlation : 3C273