Variability Of Active Galactic Nuclei Research Articles

Properties of long-term optical variability of active galactic nuclei with double-peaked broad low-ionization emission lines

In this manuscript, we study properties of long-term optical variability of a large sample of 106 SDSS spectroscopically confirmed AGN with double-peaked broad low-ionization emission lines (double-peaked emitters). The long-term optical light curves over 8 years are collected from the Catalina Sky Surveys Data Release 2. And, the Damped Random Walk (DRW) process is applied to describe the long-term variability of the double-peaked emitters. Meanwhile, the same DRW process is applied to long-term optical light curves of more than 7000 spectroscopically confirmed normal quasars in the SDSS Stripe82 Database. Then, we can find that the DRW process determined rest-frame intrinsic variability timescales $\ln(\tau/{\rm days})$ are about 5.8 and about 4.8 for the double-peaked emitters and for the normal quasars, respectively. The statistically longer intrinsic variability timescales can be confirmed in the double-peaked emitters, after considerations of necessary effects, such as the effects from different distributions of redshift, BH mass and accretion rate between the double-peaked emitters and the normal quasars. Moreover, a radial dependence of accretion rate $\dot{m}_{\rm R}~\propto~R^\beta$ with larger values of $\beta$ could be an acceptable interpretation of the longer intrinsic variability timescales in the double-peaked emitters. Therefore, there are different intrinsic properties of emission regions between the double-peaked emitters and the normal quasars. The double-peaked emitters can be well treated as an unique subclass of AGN.

Theoretical modelling of the AGN iron line vs. continuum time-lags in the lamp-post geometry

Context: Theoretical modelling of time-lags between variations in the Fe K$\alpha$ emission and the X-ray continuum might shed light on the physics and geometry of the X-ray emitting region in active galaxies (AGN) and X-ray binaries. We here present the results from a systematic analysis of time-lags between variations in two energy bands ($5-7$ vs $2-4\,\mathrm{keV}$) for seven X-ray bright and variable AGN. Aims: We estimate time-lags as accurately as possible and fit them with theoretical models in the context of the lamp-post geometry. We also constrain the geometry of the X-ray emitting region in AGN. Methods: We used all available archival \textit{XMM-Newton} data for the sources in our sample and extracted light curves in the $5-7$ and $2-4\,\mathrm{keV}$ energy bands. We used these light curves and applied a thoroughly tested (through extensive numerical simulations) recipe to estimate time-lags that have minimal bias, approximately follow a Gaussian distribution, and have known errors. Using traditional $\chi^2$ minimisation techniques, we then fitted the observed time-lags with two different models: a phenomenological model where the time-lags have a power-law dependence on frequency, and a physical model, using the reverberation time-lags expected in the lamp-post geometry. The latter were computed assuming a point-like primary X-ray source above a black hole surrounded by a neutral and prograde accretion disc with solar iron abundance. We took all relativistic effects into account for various X-ray source heights, inclination angles, and black hole spin values.

Ultraviolet and X-ray variability of active galactic nuclei withSwift

We analyse a sample of 21 active galactic nuclei (AGN) using data from the Swift satellite to study the variability properties of the population in the X-ray, UV and optical band. We find that the variable part of the UV-optical emission has a spectrum consistent with a powerlaw, with an average index of $-2.21\pm0.13$, as would be expected from central illumination of a thin disc (index of -7/3). We also calculate the slope of a powerlaw from UV to X-ray variable emission, $\alpha_{\rm OX,Var}$; the average for this sample is $\alpha_{\rm OX,Var} = -1.06 \pm 0.04$. The anticorrelation of $\alpha_{\rm OX}$ with the UV luminosity, $L_{\rm UV}$, previously found in the average emission is also present in the variable part: $\alpha_{\rm OX,Var} = (-0.177 \pm 0.083) {\rm log} (L_{\rm \nu,Var} (2500\,\AA)) + (3.88 \pm 2.33)$. Correlated variability between the emission in X-rays and UV is detected significantly for 9 of the 21 sources. All these cases are consistent with the UV lagging the X-rays, as would be seen if the correlated UV variations were produced by the reprocessing of X-ray emission. The observed UV lags are tentatively longer than expected for a standard thin disc.

Ensemble X-ray variability of active galactic nuclei

Most investigations of the X-ray variability of active galactic nuclei (AGN) have been concentrated on the detailed analyses of individual, nearby sources. A relatively small number of studies have treated the ensemble behaviour of the more general AGN population in wider regions of the luminosity-redshift plane. We want to determine the ensemble variability properties of a rich AGN sample, called Multi-Epoch XMM Serendipitous AGN Sample (MEXSAS), extracted from the fifth release of the XMM-Newton Serendipitous Source Catalogue (XMMSSC-DR5), with redshift between 0.1 and 5, and X-ray luminosities in the 0.5-4.5 keV band between 10^42 and 10^47 erg/s. We urge caution on the use of the normalised excess variance (NXS), noting that it may lead to underestimate variability if used improperly. We use the structure function (SF), updating our previous analysis for a smaller sample. We propose a correction to the NXS variability estimator, accounting for the light curve duration in the rest frame on the basis of the knowledge of the variability behaviour gained by SF studies. We find an ensemble increase of the X-ray variability with the rest-frame time lag tau, given by tau^0.12. We confirm an inverse dependence on the X-ray luminosity, approximately as L_X^-0.19. We analyse the SF in different X-ray bands, finding a dependence of the variability on the frequency as nu^-0.15, corresponding to a softer when brighter trend. In turn, this dependence allows us to parametrically correct the variability estimated in observer-frame bands to that in the rest frame, resulting in a moderate shift upwards (V-correction). Ensemble X-ray variability of AGNs is best described by the structure function. An improper use of the normalised excess variance may lead to an underestimate of the intrinsic variability, so that appropriate corrections to the data or the models must be applied to prevent these effects.

The Extremes of AGN Variability

AbstractWe present results from our ongoing monitoring programs aimed at identifying and understanding Active Galactic Nuclei (AGN) in extreme flux and spectral states. Observations of AGN in extreme states can reveal the nature of the inner accretion flow, the physics of matter under strong gravity, and they provide insight on the properties of ionized absorbers and outflows launched near supermassive black holes (SMBHs). We present new results from our long-term monitoring of IC 3599, WPVS007, and Mrk 335, multi-wavelength follow-ups of the newly identified changing-look AGN HE 1136–2304, and UV–X-ray follow-ups of the binary SMBH candidate OJ 287 after its 2015 optical maximum, now in a new optical-X-ray–high-state.

Pan-STARRS and PESSTO search for an optical counterpart to the LIGO gravitational-wave source GW150914

We searched for an optical counterpart to the first gravitational wave source discovered by LIGO (GW150914), using a combination of the Pan-STARRS1 wide-field telescope and the PESSTO spectroscopic follow-up programme. As the final LIGO sky maps changed during analysis, the total probability of the source being spatially coincident with our fields was finally only 4.2 per cent. Therefore we discuss our results primarily as a demonstration of the survey capability of Pan-STARRS and spectroscopic capability of PESSTO. We mapped out 442 square degrees of the northern sky region of the initial map. We discovered 56 astrophysical transients over a period of 41 days from the discovery of the source. Of these, 19 were spectroscopically classified and a further 13 have host galaxy redshifts. All transients appear to be fairly normal supernovae and AGN variability and none is obviously linked with GW150914. We illustrate the sensitivity of our survey by defining parameterised lightcurves with timescales of 4, 20 and 40 days and use the sensitivity of the Pan-STARRS1 images to set limits on the luminosities of possible sources. The Pan-STARRS1 images reach limiting magnitudes of i = 19.2, 20.0 and 20.8 respectively for the three timescales. For long timescale parameterised lightcurves (with FWHM=~40d) we set upper limits of M_i <= -17.2 -0.9/+1.4 if the distance to GW150914 is D = 400 +/- 200Mpc. The number of type Ia SN we find in the survey is similar to that expected from the cosmic SN rate, indicating a reasonably complete efficiency in recovering supernova like transients out to D = 400 +/- 200 Mpc.

The MIXR sample: AGN activity versus star formation across the cross-correlation ofWISE, 3XMM, and FIRST/NVSS

We cross-correlate the largest available Mid-Infrared (WISE), X-ray (3XMM) and Radio (FIRST+NVSS) catalogues to define the MIXR sample of AGN and star-forming galaxies. We pre-classify the sources based on their positions on the WISE colour/colour plot, showing that the MIXR triple selection is extremely effective to diagnose the star formation and AGN activity of individual populations, even on a flux/magnitude basis, extending the diagnostics to objects with luminosities and redshifts from SDSS DR12. We recover the radio/mid-IR star formation correlation with great accuracy, and use it to classify our sources, based on their activity, as radio-loud and radio-quiet AGN, LERGs/LINERs, and non-AGN galaxies. These diagnostics can prove extremely useful for large AGN and galaxy samples, and help develop ways to efficiently triage sources when data from the next generation of instruments becomes available. We study bias in detail, and show that while the widely-used WISE colour selections for AGN are very successful at cleanly selecting samples of luminous AGN, they miss or misclassify a substantial fraction of AGN at lower luminosities and/or higher redshifts. MIXR also allows us to test the relation between radiative and kinetic (jet) power in radio-loud AGN, for which a tight correlation is expected due to a mutual dependence on accretion. Our results highlight that long-term AGN variability, jet regulation, and other factors affecting the $Q/L$$_{bol}$ relation, are introducing a vast amount of scatter in this relation, with dramatic potential consequences on our current understanding of AGN feedback and its effect on star formation.

REVISITING STOCHASTIC VARIABILITY OF AGNs WITH STRUCTURE FUNCTIONS

ABSTRACT Discrepancies between reported structure function (SF) slopes and their overall flatness as compared to the expectations from the damped random walk (DRW) model, which generally well describes the variability of active galactic nuclei (AGNs), have triggered us to study this problem in detail. We review common AGN variability observables and identify their most common problems. Equipped with this knowledge, we study ∼9000 r-band AGN light curves from Stripe 82 of the Sloan Digital Sky Survey, using SFs described by stochastic processes with the power exponential covariance matrix of the signal. We model the “subensemble” SFs in the redshift–absolute magnitude bins with the full SF equation (including the turnover and the noise part) and a single power law (SPL; in the “red noise regime” after subtracting the noise term). The distribution of full-equation SF (SPL) slopes peaks at (0.52 ± 0.06) and is consistent with the DRW model. There is a hint of a weak correlation of γ with the luminosity and a lack of correlation with the black hole mass. The typical decorrelation timescale in the optical is year. The SF amplitude at one year obtained from the SPL fitting is mag and is overestimated because the SF is already at the turnover part, so the true value is mag. The asymptotic variability is mag. It is strongly anticorrelated with both the luminosity and the Eddington ratio and is correlated with the black hole mass. The reliability of these results is fortified with Monte Carlo simulations.

SIMULATING THE TIMESCALE-DEPENDENT COLOR VARIATION IN QUASARS WITH A REVISED INHOMOGENEOUS DISK MODEL

ABSTRACT The UV–optical variability of active galactic nuclei and quasars is useful for understanding the physics of the accretion disk and is gradually being attributed to stochastic fluctuations over the accretion disk. Quasars generally appear bluer when they brighten in the UV–optical bands; the nature of this phenomenon remains controversial. Recently, Sun et al. discovered that the color variation of quasars is timescale-dependent, in the way that faster variations are even bluer than longer term ones. While this discovery can directly rule out models that simply attribute the color variation to contamination from the host galaxies, or to changes in the global accretion rates, it favors the stochastic disk fluctuation model as fluctuations in the inner-most hotter disk could dominate the short-term variations. In this work, we show that a revised inhomogeneous disk model, where the characteristic timescales of thermal fluctuations in the disk are radius-dependent (i.e., τ ∼ r; based on that originally proposed by Dexter &amp; Agol), can reproduce well a timescale-dependent color variation pattern, similar to the observed one and unaffected by the uneven sampling and photometric error. This demonstrates that one may statistically use variation emission at different timescales to spatially resolve the accretion disk in quasars, thus opening a new window with which to probe and test the accretion disk physics in the era of time domain astronomy. Caveats of the current model, which ought to be addressed in future simulations, are discussed.

USING LEAKED POWER TO MEASURE INTRINSIC AGN POWER SPECTRA OF RED-NOISE TIME SERIES

ABSTRACT Fluxes emitted at different wavebands from active galactic nuclei (AGNs) fluctuate at both long and short timescales. The variation can typically be characterized by a broadband power spectrum, which exhibits a red-noise process at high frequencies. The standard method of estimating the power spectral density (PSD) of AGN variability is easily affected by systematic biases such as red-noise leakage and aliasing, in particular when the observation spans a relatively short period and is gapped. Focusing on the high-frequency PSD that is strongly distorted due to red-noise leakage and usually not significantly affected by aliasing, we develop a novel and observable normalized leakage spectrum (NLS), which sensitively describes the effects of leaked red-noise power on the PSD at different temporal frequencies. Using Monte Carlo simulations, we demonstrate how an AGN underlying PSD sensitively determines the NLS when there is severe red-noise leakage, and thereby how the NLS can be used to effectively constrain the underlying PSD.

Using gravitationally lensed images to investigate the intrinsic AGN variability

We discuss about how the relative flux densities among the images of gravitationally-lensed active galactic nuclei, AGN, can be used to study the intrinsic AGN variability with high accuracy. Multi-frequency monitoring observations of resolved gravitational lenses can allow us to detect signals of very weak variability and also provide information about the jet opacity and structure. As an example, we investigate the variability of the flux-density ratio between the two lensed images of the blazar B0218+357, using dual-frequency cm-wave observations. Similar to our previously reported submm-wave observations of the lensed blazar PKS1830-211, we observe a clear chromatic variability, starting short before an increase in the flux-density of the blazar. The evolution of the flux-density ratios between the blazar images shows a more clear and rich structure than that of the mere lightcurves of each individual image. The accuracy in the ratio measurements is allowing us to see variability episodes in the blazar that are weaker than the natural scatter in the absolute flux-density measurements. A simple opacity model in the jet is used to consistently explain the difference between the flux-density-ratio evolution at the two frequencies.

The origin of UV‐optical variability in AGN and test of disc models: XMM‐Newton and ground‐based observations of NGC 4395

AbstractThe origin of short timescale (weeks/months) variability of AGN, whether due to intrinsic disc variations or reprocessing of X‐ray emission by a surrounding accretion disc, has been a puzzle for many years. However recently a number of observational programmes, particularly of NGC 5548 with Swift, have shown that the UV/optical variations lag behind the X‐ray variations in a manner strongly supportive of X‐ray reprocessing. Somewhat surprisingly, the implied size of the accretion disc is ∼3 times greater than expected from a standard, smooth, Shakura‐Sunyaev thin disc model. Although the difference may be explained by a clumpy accretion disc, it is not clear whether the difference will occur in all AGN or whether it may change as, eg, a function of black hole mass, accretion rate, or disc temperature. Measurements of interband lags for most AGN require long timescale monitoring, which is hard to arrange. However for low mass (&lt;106 M⊙) AGN, the combination of XMM‐Newton EPIC (X‐rays) with the optical monitor in fast readout mode allows an X‐ray/UV‐optical lag to be measured within a single long observation. Here we summarise previous related observations and report on XMM‐Newton observations of NGC 4395 (mass 100 times lower, accretion rate ∼20 times lower than for NGC 5548). We find that the UVW1 lags the X‐rays by ∼470 s. Simultaneous observations at 6 different ground based observatories also allowed the g‐band lag (∼800 s) to be measured. These observations are in agreement with X‐ray reprocessing but initial analysis suggests that, for NGC 4395, they do not differ markedly from the predictions of the standard thin disc model. (© 2016 WILEY‐VCH Verlag GmbH &amp; Co. KGaA, Weinheim)

Parsec-scale radio morphology and variability of a changing-look AGN: the case of Mrk 590

We investigate the origin of the parsec-scale radio emission from the changing-look active galactic nucleus (AGN) of Mrk 590, and examine whether the radio power has faded concurrently with the dramatic decrease in accretion rates observed between the 1990s and the present. We detect a compact core at 1.6 GHz and 8.4 GHz using new Very Long Baseline Array observations, finding no significant extended, jet-like features down to $\sim$1 pc scales. The flat spectral index ($\alpha_{1.6}^{8.4} = 0.03$) and high brightness temperature ($T_{\rm b} \sim 10^{8}\,\rm K$) indicate self-absorbed synchrotron emission from the AGN. The radio to X-ray luminosity ratio of ${\rm log}(L_{\rm R}/L_{\rm X}) \sim -5$, similar to that in coronally active stars, suggests emission from magnetized coronal winds, although unresolved radio jets are also consistent with the data. Comparing new Karl G. Jansky Very Large Array measurements with archival and published radio flux densities, we find $46\%$, $34\%$, and (insignificantly) $13\%$ flux density decreases between the 1990s and the year 2015 at 1.4 GHz, 5 GHz and 8.4 GHz respectively. This trend, possibly due to the expansion and fading of internal shocks within the radio-emitting outflow after a recent outburst, is consistent with the decline of the optical-UV and X-ray luminosities over the same period. Such correlated variability demonstrates the AGN accretion-outflow connection, confirming that the changing-look behaviour in Mrk 590 originates from variable accretion rates rather than dust obscuration. The present radio and X-ray luminosity correlation, consistent with low/hard state accretion, suggests that the black hole may now be accreting in a radiatively inefficient mode.

Long-term X-ray spectral variability in AGN from the Palomar sample observed bySwift

We present X-ray spectral variability of 24 local active galactic nuclei (AGN) from the Palomar sample of nearby galaxies, as observed mainly by Swift. From hardness ratio measurements, we find that 18 AGN with low accretion rates show hardening with increasing count rate, converse to the softer-when-brighter behaviour normally observed in AGN with higher accretion rates. Two AGN show softening with increasing count rate, two show more complex behaviour, and two do not show any simple relationship. Sufficient data were available for the spectra of 13 AGN to be summed in flux-bins. In 9 of these sources, correlated luminosity-dependent changes in the photon index ($\Gamma$) of a power-law component are found to be the main cause of hardness variability. For 6 objects, with a low accretion rate as a fraction of the Eddington rate (\.m$\mathrm{_{Edd}}$), $\Gamma$ is anticorrelated with \.m$\mathrm{_{Edd}}$, i.e. `harder-when-brighter' behaviour is observed. The 3 higher-\.m$\mathrm{_{Edd}}$-rate objects show a positive correlation between $\Gamma$ and \.m$\mathrm{_{Edd}}$. This transition from harder-when-brighter at low \.m$\mathrm{_{Edd}}$s to softer-when-brighter at high \.m$\mathrm{_{Edd}}$s can be explained by a change in the dominant source of seed-photons for X-ray emission from cyclo-synchrotron emission from the Comptonising corona itself to thermal seed-photons from the accretion disc. This transition is also seen in the `hard state' of black hole X-ray binaries (BHXRBs). The results support the idea that LINERs are analogues of BHXRBs in the hard state and that Seyferts are analogues of BHXRBs in either the high-accretion-rate end of the hard state or in the hard-intermediate state.

A degeneracy in DRW modelling of AGN light curves

Individual light curves of active galactic nuclei (AGNs) are nowadays successfully modelled with the damped random walk (DRW) stochastic process, characterized by the power exponential covariance matrix of the signal, with the power $\beta=1$. By Monte Carlo simulation means, we generate mock AGN light curves described by non-DRW stochastic processes ($0.5\leq\beta\leq 1.5$ and $\beta\neq1$) and show they can be successfully and well-modelled as a single DRW process, obtaining comparable goodness of fits. A good DRW fit, in fact, may not mean that DRW is the true underlying process leading to variability and it cannot be used as a proof for it. When comparing the input (non-DRW) and measured (DRW) process parameters, the recovered time scale (amplitude) increases (decreases) with the increasing input $\beta$. In practice, this means that the recovered DRW parameters may lead to biased (or even non-existing) correlations of the variability and physical parameters of AGNs if the true AGN variability is caused by non-DRW stochastic processes. The proper way of identifying the processes leading to variability are model-independent structure functions and/or power spectral densities and then using such information on the covariance matrix of the signal in light curve modelling.

A note on periodicity of long-term variations of optical continuum in active galactic nuclei

Abstract Graham et al. found a sample of active galactic nuclei (AGNs) and quasars from the Catalina Real-time Transient Survey (CRTS) that have long-term periodic variations in optical continuum. The nature of the periodicity remains uncertain. We investigate the periodic variability characteristics of the sample by testing the relations of the observed variability periods with AGN optical luminosity, black hole mass and accretion rates, and find no significant correlations. We also test the observed periods in several different aspects related to accretion discs surrounding single black holes, such as the Keplerian rotational periods of 5100 Å photon-emission regions and self-gravity dominated regions and the precessing period of warped discs. These tests shed new lights on understanding AGN variability in general. Under the assumption that the periodic behaviour is associated with supermassive black hole binary systems in particular, we compare the separations (${\scr {D}}_{\bullet }$) against characteristic radii of broad-line regions ($R_{\scriptscriptstyle {\rm BLR}}$) of the binaries and find ${\scr {D}}_{\bullet }\approx 0.05R_{\scriptscriptstyle {\rm BLR}}$. This interestingly implies that these binaries have only circumbinary BLRs.

Optimal strategies for observation of active galactic nuclei variability with Imaging Atmospheric Cherenkov Telescopes

Variable emission is one of the defining characteristic of active galactic nuclei (AGN). While providing precious information on the nature and physics of the sources, variability is often challenging to observe with time- and field-of-view-limited astronomical observatories such as Imaging Atmospheric Cherenkov Telescopes (IACTs). In this work, we address two questions relevant for the observation of sources characterized by AGN-like variability: what is the most time-efficient way to detect such sources, and what is the observational bias that can be introduced by the choice of the observing strategy when conducting blind surveys of the sky. Different observing strategies are evaluated using simulated light curves and realistic instrument response functions of the Cherenkov Telescope Array (CTA), a future gamma-ray observatory. We show that strategies that makes use of very small observing windows, spread over large periods of time, allows for a faster detection of the source, and are less influenced by the variability properties of the sources, as compared to strategies that concentrate the observing time in a small number of large observing windows. Although derived using CTA as an example, our conclusions are conceptually valid for any IACTs facility, and in general, to all observatories with small field of view and limited duty cycle.

A new approach to the variability characterization of active galactic nuclei

The normalized excess variance is a popular method used by many authors to estimate the variability of active galactic nuclei (AGNs), especially in the X-ray band. We show that this estimator is affected by the cosmological time dilation, so that it should be appropriately corrected when applied to AGN samples distributed in wide redshift intervals. We propose a formula to modify this estimator, based on the use of the structure function. To verify the presence of the cosmological effect and the reliability of the proposed correction, we use data extracted from the XMM-Newton Serendipitous Source Catalogue, data release 5 (XMMSSC-DR5), and cross-matched with the Sloan Digital Sky Survey quasar catalogue, of data release 7 and 12.

Ensemble spectral variability study of Active Galactic Nuclei from the XMM-Newton serendipitous source catalogue

The variability of the X-Ray spectra of active galactic nuclei (AGN) usually includes a change of the spectral slope. This has been investigated for a small sample of local AGNs by Sobolewska and Papadakis [1], who found that slope variations are well correlated with flux variations, and that the spectra are typically steeper in the bright phase (softer when brighter behaviour). Not much information is available for the spectral variability of high-luminosity AGNs and quasars. In order to investigate this phenomenon, we use data from the XMM-Newton Serendipitous Source Catalogue, Data Release 5, which contains X- Ray observations for a large number of active galactic nuclei in a wide luminosity and redshift range, for several different epochs. This allows to perform an ensemble analysis of the spectral variability for a large sample of quasars. We quantify the spectral variability through the spectral variability parameter β, defined by Trevese and Vagnetti [2] as the ratio between the change in spectral slope and the corresponding logarithmic flux variation. We find that the spectral variability of quasars has a softer when brighter behaviour, similarly to local AGNs.

QUASAR VARIABILITY IN THE MID-INFRARED

ABSTRACT The Decadal IRAC Boötes Survey is a mid-IR variability survey of the ∼9 sq. deg. of the NDWFS Boötes Field and extends the time baseline of its predecessor, the Spitzer Deep, Wide-Field Survey (SDWFS), from 4 to 10 years. The Spitzer Space Telescope visited the field five times between 2004 and 2014 at 3.6 and 4.5 μm. We provide the difference image analysis photometry for a half a million mostly extragalactic sources. In mid-IR color–color plane, sources with quasar colors constitute the largest variability class (75%), 16% of the variable objects have stellar colors and the remaining 9% have the colors of galaxies. Adding the fifth epoch doubles the number of variable active galactic nuclei (AGNs) for the same false positive rates as in SDWFS, or increases the number of sources by 20% while decreasing the false positive rates by factors of 2–3 for the same variability amplitude. We quantify the ensemble mid-IR variability of ∼1500 spectroscopically confirmed AGNs using single power-law structure functions (SFs), which we find to be steeper (index &gamma; &approx; 0.45 ?> ) than in the optical ( &gamma; &approx; 0.3 ?> ), leading to much lower amplitudes at short time-lags. This provides evidence for large emission regions, smoothing out any fast UV/optical variations, as the origin of infrared quasar variability. The mid-IR AGN SF slope γ seems to be uncorrelated with both the luminosity and rest-frame wavelength, while the amplitude shows an anti-correlation with the luminosity and a correlation with the rest-frame wavelength.