Quasi-periodicities at year-like timescales in Blazars
Extracted source text ยท arXiv 1512.04561
%%
%% Beginning of file 'sample.tex'
%%
%% Modified 2005 December 5
%%
%% This is a sample manuscript marked up using the
%% AASTeX v5.x LaTeX 2e macros.
%% The first piece of markup in an AASTeX v5.x document
%% is the \documentclass command. LaTeX will ignore
%% any data that comes before this command.
%% The command below calls the preprint style
%% which will produce a one-column, single-spaced document.
%% Examples of commands for other substyles follow. Use
%% whichever is most appropriate for your purposes.
%%
\documentclass[12pt,preprint]{aastex}
%% manuscript produces a one-column, double-spaced document:
%\documentclass[manuscript]{aastex}
%% preprint2 produces a double-column, single-spaced document:
%\documentclass[preprint2]{aastex}
%% Sometimes a paper's abstract is too long to fit on the
%% title page in preprint2 mode. When that is the case,
%% use the longabstract style option.
%% \documentclass[preprint2,longabstract]{aastex}
%% If you want to create your own macros, you can do so
%% using \newcommand. Your macros should appear before
%% the \begin{document} command.
%%
%% If you are submitting to a journal that translates manuscripts
%% into SGML, you need to follow certain guidelines when preparing
%% your macros. See the AASTeX v5.x Author Guide
%% for information.
\usepackage{graphicx}
\newcommand{\vdag}{(v)^\dagger}
\newcommand{\myemail}{skywalker@galaxy.far.far.away}
%% You can insert a short comment on the title page using the command below.
\slugcomment{Not to appear in Nonlearned J., 45.}
%% If you wish, you may supply running head information, although
%% this information may be modified by the editorial offices.
%% The left head contains a list of authors,
%% usually a maximum of three (otherwise use et al.). The right
%% head is a modified title of up to roughly 44 characters.
%% Running heads will not print in the manuscript style.
\shorttitle{Quasi-periodicities in Blazars}
\shortauthors{A. Sandrinelli et al.}
%% This is the end of the preamble. Indicate the beginning of the
%% paper itself with \begin{document}.
\begin{document}
%% LaTeX will automatically break titles if they run longer than
%% one line. However, you may use \\ to force a line break if
%% you desire.
\title{Quasi-periodicities at year-like timescales in Blazars}
%% Use \author, \affil, and the \and command to format
%% author and affiliation information.
%% Note that \email has replaced the old \authoremail command
%% from AASTeX v4.0. You can use \email to mark an email address
%% anywhere in the paper, not just in the front matter.
%% As in the title, use \\ to force line breaks.
\author{A. Sandrinelli} %\altaffilmark{1,2,3} and Ivan R. King\altaffilmark{1}}
\affil{
\textit{Universit\`a degli Studi dell'Insubria, Dipartimento di Scienza ed Alta Tecnologia, \\ Via Valleggio 11, I-22100 Como, Italy.\\
INAF - Istituto Nazionale di Astrofisica, Osservatorio Astronomico di Brera, \\Via Emilio Bianchi 46, I-23807 Merate, Italy.}
}
\author{S. Covino}
\affil{\textit{
INAF - Istituto Nazionale di Astrofisica, Osservatorio Astronomico di Brera, \\Via Emilio Bianchi 46, I-23807 Merate, Italy
}}
\author{M. Dotti}
\affil{\textit{
Universit\`a degli Studi di Milano Bicocca,
Dipartimento di Fisica G. Occhialini,\\ Piazza della Scienza 3, I-20126 Milano, Italy\\
INFN- Istituto Nazionale di Fisica Nucleare, \\Universit\`a degli Studi di Milano Bicocca,
Dipartimento di Fisica G. Occhialini,\\ Piazza della Scienza 3, I-20126 Milano, Italy
}}
\and
\author{A. Treves\altaffilmark{1}}
\affil{\textit{Universit\`a degli Studi dell'Insubria, Dipartimento di Scienza ed Alta Tecnologia,\\ Via Valleggio 11, I-22100 Como, Italy\\
INAF - Istituto Nazionale di Astrofisica, Osservatorio Astronomico di Brera, \\Via Emilio Bianchi 46, I-23807 Merate, Italy\\
INFN- Istituto Nazionale di Fisica Nucleare, Sezione Trieste - Udine, \\
Via Valerio 2, I-34127 Trieste, Italy}
}
\altaffiltext{1}{ICRA}
%% Notice that each of these authors has alternate affiliations, which
%% are identified by the \altaffilmark after each name. Specify alternate
%% affiliation information with \altaffiltext, with one command per each
%% affiliation.
%% Mark off your abstract in the ``abstract'' environment. In the manuscript
%% style, abstract will output a Received/Accepted line after the
%% title and affiliation information. No date will appear since the author
%% does not have this information. The dates will be filled in by the
%% editorial office after submission.
\begin{abstract}
We searched for quasi-periodicities on year-like timescales in the light curves of 6
blazars in the optical - near infrared bands and
we made a comparison with the high energy emission.
We obtained optical/NIR light curves from REM photometry plus archival SMARTS
data and we accessed the \textit{Fermi} light curves for the $\gamma$-ray data.
The periodograms often show strong peaks
in the optical and $\gamma$-ray bands, which in some cases
may be inter-related.
The significance of the revealed peaks is then discussed,
taking into account that the noise is frequency dependent.
Quasi-periodicities on a year-like timescale appear to occur often in blazars.
No straightforward model describing these possible periodicities
is yet available, but some plausible
interpretations for the physical mechanisms
causing periodic variabilities of these sources are examined.
\end{abstract}
\keywords{
Blazar objects: general $-$
Blazar objects: individual (PKS 0537-441, OJ 287, 3C 279, PKS 1510-089,
PKS 2005-489 and PKS 2155-304) $-$
galaxies: active $-$
method: photometry, statistics
}
%---------------------------------
\section{Introduction}
%---------------------------------
Blazars are active galactic nuclei exhibiting large variability
at all frequencies.
Their emission is dominated by a relativistic jet, the amplification
factor of the intensity is characterized by a power of the
Doppler factor \citep[e.g.][]{Ghisellini2013}.
The jet produces a non thermal spectrum where often one can distinguish
a synchrotron and a Compton component.
The energy source is assumed to be a combination of accretion on a supermassive
black hole (SMBH)
and extraction of its spin energy. In some cases the presence of an accretion
disk is suggested by the appearance in the spectral energy distribution
of a thermal component.
The observed variability is supposedly a consequence of the intrinsic change of the
accretion onto the SMBH, of the jet formation process, and specifically of its beaming.
Various processes originating in different regions and with different time
scales result in a complicated variability
leading to rather chaotic light curves.
The study of variability through auto- and cross-correlation procedures
has proven to be effective in constraining complex models of these sources.
The discovery of a periodicity in the variability could have
profound consequences in the global understanding of the sources, constituting a
basic block for models.
The effort for finding periodicities has been
substantial at all frequencies \citep[see e.g.][]{Falomo2014}. A rather robust claim
is that of a $\sim$12 year period in OJ 287 \citep[e.g.][]{Sillanpaa1988},
a BL Lac object possibly containing in its center a system of two SMBHs
\citep{Lehto1996}.
Note however that the periodicity and the picture are disputed for instance by \cite{Hudec2013}.
\cite{Graham2015} proposed a 1980 day optical period for the quasar PG1302-102,
which also appears reliable.
Models based on the presence of a
binary black hole \citep[e.g.][]{Sundelius1997,Kidger2000,Valtonen2012},
describe outbursts and flares
following a not strictly periodic cadence.
In these cases, convolution of various processes can lead to an apparently
unstable or variable period, i.e. to quasi-periodicities.
Search for quasi-periodicities in blazars is a valid and efficient diagnostic tool.
Some sources could exhibit long lasting quasi-periodic behaviors. The time scales
of the events and their persistence in the light curves may allows us to shed light on the physical
processes underlying these variations.
Among previous proposals of year-like quasi-periodicities
in blazars, we refer to \cite{Raiteri2001}, \cite{Gabanyi2007}, \cite{Rani2009},
\cite{Li2015}, and references therein.
In addition, recently \cite{Zhang2014} collecting photometric data of PKS 2155-304
published in the last 35 years, discovered a quasi-periodicity of $T_{0}= $ 317 d.
This was confirmed by our independent photometry \citep{Sandrinelli2014a,
Sandrinelli2014b}.
Moreover we showed that a periodicity appears also in the $\gamma$-rays
observed by the \textit{Fermi} mission at $T=2\cdot T_{0}$.
Because \textit{Fermi} has monitored continuously the sky since $\sim$ 6 years,
it is obvious that it is only now that one can combine optical and $\gamma$-ray
searches for year-like periodicities, a procedure which was
successful for PKS 2155-304.
Since the number of covered periods in the optical and in $\gamma$-rays
is limited, as in the case of PKS 2155-304,
it is difficult to assess the stability on a long-term basis of the inferred (quasi) periodicities.
The starting point of the present investigation are the VRIJHK photometric
observations obtained with the robotic
Rapid Eye Mounting telescope \citep[REM\footnote{REM data can be retrieved from
\texttt{http://www.rem.inaf.it}},][]{Zerbi2004, Covino2004}
at La Silla, which are described in detail in
\cite{Sandrinelli2014a}.
Among the blazar sources monitored by REM and described in the above mentioned
paper we consider here PKS 0537-441,
OJ 287, PKS 1510-089, PKS 2005-489 and
PKS 2155-304, because of the extensive coverage of the observations.
These data are available on-line\footnote{Photometric nightly averaged data
is tabled in
\texttt{http://vizier.cfa.harvard.edu/viz-bin/VizieR?-source\\=J/A+A/562/A79}}.
We add also the REM photometry of 3C 279, which is unpublished thus far.
%
We have then combined the REM data with those derived from the Small \& Moderate
Aperture Research Telescope System
archives \citep[SMARTS\footnote{\texttt{http://www.astro.yale.edu/smarts/glast/home.php}},][]{Bonning2012}.
The REM data on PKS 2155-304 were originally examined in \cite{Sandrinelli2014b}
and are those which led us to confirm the results of \cite{Zhang2014}.
In Table \ref{sample} we report a summary of the characteristics of the six sources,
of which two are flat spectrum radio quasars (FSRQs)
and the others are BL Lac objects.
The unpublished photometry of 3C 279 is described in Section \ref{photometry}.
The search for periodicities from both optical data and
\textit{Fermi} archives is presented in Section \ref{search}.
Discussion of results with a possible picture for the interpretation
of a $\sim$1 yr quasi-periodicity in blazars is reported in Section \ref{disc}
and conclusions in Section \ref{concl}.
%______________________________________________
\section{\label{photometry}REM photometry of 3C 279}
%______________________________________________
The analysis of the data follows closely the scheme described by
\cite{Sandrinelli2014a}. For the optical bands we used reference
stars from \cite{Gonzalez2001}, while for the NIR frames we referred to
the \textit{Two Micron All Sky Survey
Catalog }
\citep[2MASS\footnote{\texttt{http://www.ipac.caltech.edu/2mass/}},][]{Skrutskie2006}.
All images have been visually checked
eliminating those where the targets or the reference stars are
close to the borders of the frame, and where obvious biases
were present. In Table \ref{3c279lc_tab} we report our photometry of the source.
Some overall properties are given in Table \ref{3c279prop_tab}.
Comparing the fractional variability amplitude $\sigma_{rms}$
\citep[e.g.][]{Nandra1997,Edelson2002}
with those of the other 5 objects, 3C 279 appears highly variable.
The nightly averaged light curves in the 6 filters
are reported in Figure \ref{lc}.
%-----------------------------------------------------------------
\section{\label{search}Search for periodicities in the target sources}
%-----------------------------------------------------------------
The starting point of our analysis are the light curves obtained combining
REM (2005-2012) and SMARTS photometry (2008-2014) in V, R, J, K bands, see Figures
\ref{3lc_pks0537}-\ref{3lc_pks2155}
for the cases of R and K filters.
Note that contrary to our REM data, SMARTS' photometry is taken at face value,
as archived in the public web site.
\textit{Fermi} data\footnote{\texttt{http://fermi.gsfc.nasa.gov/ssc/data/access/lat/msl\_lc/}}
are also reported in Figure \ref{3lc_pks0537}-\ref{3lc_pks2155}
in the 100 MeV - 300 GeV band.
The procedures for constructing gamma ray curves are complex but rather
standardized\footnote{The data have been analyzed by using the standard
Fermi LAT ScienceTools software package, see\\
\texttt{http://fermi.gsfc.nasa.gov/ssc/data/analysis/\\
documentation/Cicerone/}},
and fully described e.g. in \cite{Abdo2010}. Note that these curves are
not corrected for background.
The search for (quasi) periodicities in the light curves of active galactic nuclei
is notoriously an arduous problem, as it was pointed out in the seminal paper
by \cite{Press1978}, who considered the X-ray light curve of 3C 273
and indicated a number of caveats, which should be taken into account,
before assessing the reality of a periodicity.
In our case four main points should be examined:
\begin{itemize}
\item Our optical sampling is rather irregular, as usual in ground based observations.
On the other hand $\gamma$-ray light curves are essentially evenly spaced.
\item The light curves are affected by frequency dependent \textit{red} noise.
\item The total duration $T_{tot}$ of the monitoring is $\leqslant$ 9 yr in the optical,
$\leqslant$ 6 yr in gamma rays, constraining the minimum frequency
that can be searched for.
\item Flares or periods of high activity can affect the analysis, requiring a careful check of the results.
\end{itemize}
The problem of the presence of red noise in evaluating a periodogram
has been discussed in detail in the case of X-ray light curves, mainly in the context
of galactic sources.
The case of time series not evenly spaced typically requires the use of interpolation techniques.
Their application implies that the interpolated data points are no longer independent
and may introduce a significant additional bias, which leads to an enhancement of low-frequency
components at the expense of higher frequency ones
\citep[e.g.][]{Schulz1997}.
The assessment of the reality of periodicities cannot be carried out without firm
procedures to measure a significance against the background
noise.
In particular the problem of the red noise modeling was examined
by, i.e., \cite{Israel1996}, \cite{Vaughan2005,Vaughan2010},
\cite{vanderKlis1989a,vanderKlis1989b}, \cite{Zhou2002}, \cite{Timmer1995},
although usually for the simpler case of evenly sampled time-series.
A procedure for the study of periodicities in non-equally spaced light curves and
affected by red noise was developed by \cite{Schulz2002}
with reference to paleoclimatic time
series\footnote{Details on the relevant software "REDFIT" can be found at
\texttt{http://www.geo.uni-bremen.de/geomod/staff/mschulz/\\\#software2}}.
In this procedure a first-order
autoregressive (AR1) process is applied to model the red-noise
\citep{Hasselmann1976}.
This avoids interpolation in the time domain with the introduced bias.
The AR1 technique can give a good description of rather smooth and
regular time-series, as it is the case of the Fermi light-curves we are here
considering. Our optical light-curves are, on the contrary, much more difficult
to model due to periods of high-activity and/or flares introducing order of
magnitudes variations together with a highly irregular sampling.
Nevertheless, the results of these analyses can provide useful hints about
the significance of a detected possible periodicity, in particular when these
periods are also present in the Fermi data.
We started with light curves with binning of 1 day in the optical bands,
and 1 week in the $\gamma$-rays (see Figure \ref{3lc_pks0537}-\ref{3lc_pks2155}).
We considered $\gamma$-rays bins with test statistics \citep{Mattox1996}
$TS>$ 4, corresponding to a $\sim$ 2$\sigma$ detection.
Note that for the large majority of cases the detections are much more significant than this limit.
We have chosen a maximum frequency for the period analysis
corresponding to $\sim$ 20 days.
In fact here we are not interested in the short time scale variability.
The procedure yields (see Figures \ref{pow_pks0537}-\ref{pow_pks2155}):
\begin{itemize}
\item Lomb and Scargle periodograms \citep{Scargle1982},
which account for the unevenly spaced photometry.
\item Modeling of red noise continuum.
\item Bias-corrected spectra and significance ($S$) level curves: we have chosen
$S=99.0\%$ (2.5$\sigma$) and $S= 99.7\%$ (3$\sigma$).
\end{itemize}
The significance is given by the comparison of the periodogram with that based on the
auto-regressive model (see above) to test the null hypothesis
that the observed time series can be fully accounted by the noise.
In Tables \ref{top_hits_pks0537}-\ref{top_hits_pks2155} we report the \textit{observed} periods
corresponding to peaks with $\gtrsim$ 99\% significance.
We evaluated the period uncertainty following \cite{Schwar1991},
searching for the Mean Noise Power Level (MNPL) in the vicinity of the
investigated period T.
The 1$\sigma$ confidence interval on T is the width
of the peak in the power spectrum at $p-$MNPL, where $p$ is the height
of the peak.
The sinusoidal artificial light-curves with the most significant periods calculated using
Vstar package\footnote{\texttt{http://www.aavso.org/vstar-overview}} are reported
in Figures \ref{3lc_pks0537}-\ref{3lc_pks2155}.
We checked for aliases derived from the interval sampling between observations or
the sampling rate,
which cause false peaks in the time analysis \citep[see e.g.][]{Deeming1975}.
We adopted the procedure in the Period Analysis
Software\footnote{\texttt{http://www.peranso.com}}
and found in
all the NIR-optical curves evidences
of an alias period of
$\sim$ 370 d, which is representative of the year length.
The corresponding peak is negligible with respect to the other periodicities
in all the sources with the exception of PKS 2005-489.
Applying the same procedure to $\gamma$-ray
light curves, no aliases have been detected, as expected,
except for a significant one at 7 d, denoting the sampling rate of the \textit{Fermi} data.
The peaks, which appear marginally revealed in the spectral analysis for
alias detection, are marked in Tables \ref{top_hits_pks0537} - \ref{top_hits_pks2155}.
%------------------------------------
We now examine the sources, separately, discussing the most important possible
periodicities detected by the analysis. As a general point, several periods
appear to have a formal significance close or better than 99\%.
However considering the many possible sources of uncertainty
(mainly in the optical/NIR) it may be not sufficient to provide a fully solid
statement about their reliability.
%------------------------------------
\textit{PKS 0537-441} --
In the R, J, K bands there are peaks at $T_1 \sim$ 150 d with a significance $S\gtrsim$ 98\%,
as reported in Table \ref{top_hits_pks0537},
and in the $\gamma$-rays a peak appears at 351 d
with $S$$\sim$ 90\%,
see Figure \ref{pow_pks0537}.
The most prominent peak
at 1668 d
is close to the total observing time $T_{tot}$ and it is non reported in Table \ref{top_hits_pks0537}
and Figure \ref{pow_pks0537}.
\textit{OJ 287} --
In the NIR-optical bands plotted in Figure \ref{3lc_oj287} there are peaks
at $T_1 \sim $435 d ($S\gtrsim$ 99.7\% in V, J and K, see Table \ref{top_hits_oj}), which
within the errors may be related with the $\gamma$-ray periodicity at $\sim$ 410 d ($S\sim$ 99\%).
\textit{3C 279} --
As it is apparent from
Figure \ref{3lc_3c279}, there are
significant peaks at
$T_{1} \sim 910$ d, which however is comparable with $T_{tot}$. $T_{1}$
is the most prominent one, which can be taken into account
at low frequencies in the $\gamma$-ray spectrum ($S\gtrsim$ 95 \%).
It is noticeable that peaks at $T_{1}$ and $T_{7} \sim$ 24 d are present both in the periodograms
of the $\gamma$-ray data and of the NIR-optical bands
(in K-band for $T_{1}$ $S$ is $>$99.7 \%), see Figure \ref{pow_3c279} and Table \ref{top_hits_3c279}.
In the NIR-optical bands the presence of peaks at $ T_{2}\sim$ 260 d, with
significance $\geq3\sigma$ in V and K, stands out.
In R and V bands the adopted procedure fails in producing a reliable red-noise profile
due to the prominent flare at $\sim$ MJD 53700.
In this case the search for periodicities
was performed splitting the light curve in 3 segments, which are overlapping for the 50\%
of their length \citep[Welch-overlapped-segment-averaging, WOSA, ][]{Welch1967}.
The spectral features are estimated from averaging the 3 resulting periodograms.
The obtained spectra return peaks at 256 d, which we can associate with
the one at $\sim 265$ d in K-band (see Figure \ref{pow_3c279} and Table \ref{top_hits_3c279}).
\textit{PKS 1510-089} --
The most significant peak in $\gamma$-ray band is at $T_5$=115 d ($S\sim99\%$), see
Figure \ref{pow_pks1510}.
This may be related to the optical peaks at $4\cdot T_5$ and $3\cdot T_5$,
with $S > 99.0\%$ (see Table \ref{top_hits_pks1510}).
We note that in this source the fit of the red noise continuum may be affected by a number
of short flares \citep{Sandrinelli2014a}.
\textit{PKS 2005-489} --
For this source only REM photometry is available.
A conspicuous peak at $\sim$680-720 d is present in all the NIR-optical
bands (Figure \ref{pow_pks2005} and Table \ref{top_hits_pks2005}). It may be related to
the peak at $\sim$ 370 d, which however is supposedly spurious (see above).
Because of the length of the light curve ($T_{tot}$=2500 d) the candidate periods
between 1000 d and 1500 d
are hardly significant and are not reported in Table \ref{top_hits_pks2005}.
\textit{PKS 2155-304} --
We confirm the results presented in \cite{Sandrinelli2014b} about a peak of $ T_3 \sim$ 315 d in
the NIR-optical bands and at $2 \cdot T_3$ in the $\gamma$-rays,
see Table \ref{top_hits_pks2155} and Figure \ref{pow_pks2155}.
In $\gamma$-rays and in K the significances are above $3\sigma$.
In K it is noticeable also the presence of peaks with $S \gtrsim99\%$ at 453 d and
151 d, which with $2 \cdot T_3$ and $T_{3}$ could be harmonics of the
same period $T_{7}=$76 d.
%------------------------------------------------
\section{\label{disc} Discussion}
%------------------------------------------------
The advent of robotic telescopes in the last decade on the one hand, and the
systematic monitoring of the $\gamma$-ray sky since 2008, gave an unprecedented
opportunity for exploring quasi-periodicities in blazars at year-like timescales.
The presence of possible associated periodicities in both the optical and $\gamma$-ray bands
may be indicative of their physical relevance.
The most convincing cases are PKS 2155-304, OJ 287 and 3C 279.
In the following we discuss some possible interpretations for the physical mechanisms
causing the periodic variability of these sources with a rest-frame year-like duration.
As already discussed in \cite{Sandrinelli2014b} and similarly to the
interpretation proposed by \cite{Lehto1996} for OJ 287,
the observed year-like timescale periodicity
could be related to the orbit of a perturbing object. This could
destabilize the accretion flow onto the primary SMBH, modulating the accretion
rate and, as a consequence, the luminosity of the active nucleus. Assuming
the mass of the active SMBH $M_1 \sim 10^9 M_{\odot}$ (typical of blazars) and
that the perturber is significantly less massive ($q=M_2/M_1<0.1$), the
observed periodicity implies a separation between the two bodies of $d \sim
100$ r$_{S}\approx 10^{-2}$ pc, where r$_{S}$ is the Schwarzschild radius of
the central SMBH. Currently available observations do not allow us to
constrain the nature of the perturber. Even assuming that this is a secondary
SMBH of mass $\sim 10^8 M_{\odot}$, the gravitational wave (GW) driven orbital
decay would be of the order of 1 Myr \citep{Peters1964}, making it hard to
observe any frequency drift due to the shrinking of the perturber orbit.
A second possibility is that the observed periodicity is related to the
precession period of the blazar jet. We can draw two distinct scenarios,
depending on whether the jet is forced to be aligned to the SMBH spin or
not. In this second case the jet would be aligned to the angular momentum of
the inner part of the accretion disc. In the case the inner disc is not lying
on the SMBH equatorial plane, it would undergo Lense-Thirring precession
around
the SMBH spin. The Lense Thirring precession period scales as:
\begin{equation}
T_{\rm Prec} = \frac{8 \pi G M_1}{c^3 a} \left(\frac{r}{r_S}\right)^3,
\end{equation}
where $a$ is the SMBH spin parameter. If the accretion flow is geometrically
thick, the inner disk precesses as a solid body, as demonstrated by
general relativistic magnetohydrodynamic numerical simulations
\citep[e.g.][]{Fragile2007, Dexter2011}\footnote{We note that a thick geometry is expected for radiatively
inefficient accretion, making this scenario particularly attractive for BL
Lacs and specifically for PKS 2155-304 \citep{Ghisellini2011, Sbarrato2012}.}.
In this case the observed period does provide an estimate of
the size of the inner precessing region. About 900 day correspond to
$\approx 8 r_S$, while $\approx 300$ day
would imply $\approx 5 r_S$ (both the estimates
assume $M_1=10^9 M_{\odot}$ and $a=0.9$).
If instead the jet is aligned with the SMBH spin, the observed periodicity has
to be related to the precession timescale of the spin itself. In order for
the
observed flux to be significantly affected by the precession, the angle
between the SMBH spin and the axis of precession (defined by the total angular
momentum of the system) has to be of the order of the jet opening angle $\sim
1^{\circ}$ or greater. Because of the large mass of the SMBH, the angular
momentum of the accreting gas within the inner $\sim 10 r_S$ cannot cause
such
a large displacement of the SMBH spin. A secondary SMBH orbiting at $\approx 8
r_S$ is required in this case. The secondary mass needed to cause a
significant variation in the observed flux has to be $M_{2} \gtrsim 10^6 M_{\odot}$.
This last scenario has a number of testable predictions:\\ $(i)$ The
lifetime of such a system would be determined by the GW driven orbital decay
of the binary. The timescale is $\gtrsim 8 \cdot 10^3$ yr for $M_{2} \gtrsim10^6
M_{\odot}$ and it scales with the inverse of $M_2$. For $M_2
\gtrsim 10^7 M_{\odot}$ a drift of the period toward smaller values could be
observable in the next decades;\\
$(ii)$ A faster variability, corresponding
to the orbital timescale of the binary, could be observed (with a period of
$\approx 24$ days assuming $q<0.1$).
%************************\\
In a different scenario,
the detected quasi-periodic oscillations in blazars
could be related to jet emissions.
Variability can be ascribed to
helical jets or helical structures in jets
\citep[e.g.][]{Camenzind1992}
which may be quite common in blazars \citep[][]{Villata1999}.
%
They could arise from hydrodynamical instabilities in magnetized
jets \citep{Hardee1999} or from variations in the jet engine, e.g., accretion
disc instabilities \citep[][and reference therein]{Godfrey2012},
also coupled with the interaction of the jet plasma with the surrounding medium.
The emitting flow moving around a helical path
could produce relatively long-term quasi-periodic changes in Doppler boosted flux \citep{Villata1999}.
In such a picture a rotating helical structure
was proposed to explain, e.g., both the quasi-periodic behavior \citep[$\sim$8 yr,][]{Raiteri2010}
of BL Lacertae and the occurrence and the mean shape of major radio-optical outbursts
in AO 0235+16 \citep[$\sim$5.7 yr,][]{Ostorero2004}.
%
Relativistic shocks \citep[e.g,][]{Marscher1985} exciting jet helical patterns can be also considered
\citep[e.g.,][]{Rani2009,Larionov2013}, as well as models based on dominant
turbulent cells in plasma jet \citep{Marscher1992,Marscher2014} driving short-lived
quasi-periodic oscillations behind a shock \citep[e.g.,][]{Rani2009}.
%
The presence of quasi-periodic peaks in radio-mapped jet emissions are, for instance,
discussed in \cite{Godfrey2012} for the case of PKS 0637-752,
where large-scale shocks in continuous flow are invoked
(re-confinement shocks, Kelvin-Helmholtz instabilities).
The discussion by \cite{Godfrey2012} proposes time scales of variability much larger
than those inferred in this paper, but the issue should be reconsidered in detail.
\section{\label{concl}Conclusions}
The significance of the detected periods is almost never very high, although in
several case good enough to deserve consideration. As discussed in the text,
a periodicity analysis of highly irregular and unevenly sampled time-series is
never an easy task. On the other hand, in some cases the detected periods are
also connected with periods derived from \textit{Fermi} data, where most of the source
of uncertainty do not apply. It is then possible that quasi-periodicities
of $\sim$ 1 yr in blazars are not rare.
Their origin is not easy to be traced, and detailed models are not yet available.
Progress in the field can come from the study of other blazars, which have
both a long optical monitoring, and are relatively bright in the \textit{Fermi} archives.
It is also obvious to search for confirmation of the periodicities in the X-rays where
large amount of sparse observations are archived. However, the X-rays could be produced
independently of the optical and $\gamma$-rays.
For the future, observations of the
SMARTS and REM type should hopefully be prosecuted for several years,
largely improving the robustness of the analysis for periodicities of several months / one year
length.
\acknowledgments
We are grateful to dr. G.L. Israel for discussions on periodicity search in light curves
affected by red noise. This paper made use of up-to-date SMARTS optical/near-infrared
light curves that are available on
line\footnote{\texttt{www.astro.yale.edu/smarts/glast/home.php}}.
We acknowledge the support of the Italian Ministry of Education
(grant PRIN-MIUR 2009, 2010, 2011).
This work was also supported by ASI grant I/004/11/0.
\clearpage
\begin{table*}
\centering
\caption{\label{sample} Blazar sample.}
\begin{tabular}{lcclcl}
\hline\hline
Source &ra\tablenotemark{a} &dec\tablenotemark{a}& Class\tablenotemark{b} & SED\tablenotemark{c} &Redshift\tablenotemark{a}\\
\hline\hline
&&&&&\\
PKS 0537-441 & 05:38:50 &$-$44:05:09 & BL Lac & LSP &0.896 \\
OJ 287 & 08:54.48 & +20.06:30 & BL Lac & LSP &0.3060 \\
3C 279 & 12:56:11 &$-$05:47:21 & FSRQ &LSP &0.536 \\
PKS 1510-089 & 15:12:50 &$-$09:05:59 & FSRQ & LSP &0.3599 \\
PKS 2005-489 & 20:09:25 &$-$48:49:53 & BL Lac & HSP &0.071 \\
PKS 2155-304 & 21:58:52 &$-$30:13:32 & BL Lac & HSP &0.117 \\
&&&&&\\
\hline
\end{tabular}
\tablenotetext{a}{From the Simbad archive (\texttt{http://simbad.u-strasg.fr/}).}
\tablenotetext{b}{From \cite{Massaro2012}.}
\tablenotetext{c}{LSP, ISP and HSP refer to low, intermediate and high synchrotron peaked
blazars \citep{Abdo2010}}
\end{table*}
\begin{table*}
\caption{\label{3c279lc_tab} REM photometry of 3C 279.
}
\centering
\begin{tabular}{cccc}
\hline\hline
Filter &Time of &Average & Magnitude \\
&observation &magnitude &error \\
&[MJD] &[mag] &[mag]\\
\hline\hline
&&&\\
V & 53467 & 15.61 &0.05\\
V & 53469 & 15.64 &0.06\\
V & 53474 & 15.40 &0.06\\
V & 53476 & 15.16 &0.04\\
V & 53492 & 15.80 &0.05\\
.... & .... & .... & .... \\
&&&\\
\hline
\end{tabular}
\tablecomments{
A full version of Table \ref{3c279lc_tab} is available
in electronic format. A portion is
shown here for guidance regarding its form and content.
}
\end{table*}
\begin{table*}
\caption{\label{3c279prop_tab} Properties of REM NIR-optical light curves of 3C 279. }
\centering
\begin{tabular}{lccccc}
\hline\hline
Filter &Mag. range & Mean mag. &Median &Flux range & $\sigma_{rms}$\tablenotemark{a}\\
&[mag] & [mag] &[mag] &[mJy] &[\%]\\
\hline
\hline
&&&&&\\
V & 13.15 - 16.83 & 15.13 & 15.14 & 0.73 - 21.77 & 72 $\pm$ 2\\
R & 12.70 - 16.58 & 14.83 & 14.95 & 0.83 - 27.37 & 83 $\pm$ 2\\
I & 12.13 - 15.93 & 14.27 & 14.32 & 1.08 - 35.68 & 86 $\pm$ 2\\
J & 10.90 - 15.03 & 13.04 & 13.07 & 1.55 - 69.30 & 82 $\pm$ 1\\
H & 10.08 - 14.24 & 12.20 & 12.18 & 2.13 - 97.74 & 82 $\pm$ 1\\
K & 9.58 - 13.30 & 11.18 & 11.18 & 3.33 - 99.81 & 66 $\pm$ 1\\
&&&&&\\
\hline
\end{tabular}
\tablenotetext{a}{Fractional variability amplitude.}
\end{table*}
\begin{table*}
\centering
\caption{\label{top_hits_pks0537}
Most prominent peaks and significances
($S \gtrsim99\%$) in PKS 0537-441 \textit{observed} power spectra.
The peaks marked
with (*) appear marginally revealed in our spectral analysis for alias detection.
We note that PKS 0537-441 is not observed by SMARTS in V band.
}
\begin{tabular}{|c|c|c|c|c|c||c||}
\hline\hline
Band & $T_{1}$ & $T_{2}$ & $T_{3}$ & $T_{4}$ & $T_{5}$ & $S \% \gtrsim$ \\
& [d] & [d] & [d] & [d] & [d] & \\
\hline\hline
R & 150$\pm$5 & & & & & 99.0 \\
& & 68$\pm$1* & & & & 99.0 \\
& & & & 61$\pm$1* & & 99.0 \\
& & & & & 58$\pm$1* & 99.0 \\
\hline
J & 153$\pm$3 & & & & & 99.0 \\
& & 68$\pm$1* & & & & 99.0 \\
& & & 64$\pm$1* & & & 99.0 \\
& & & & 61$\pm$1 & & 99.7 \\
& & & & & 53$\pm$1* & 99.0 \\
\hline
K & 152$\pm$4 & & & & & 98.0 \\
& & 68$\pm$1* & & & & 99.7 \\
& & & & 61$\pm$1 & & 99.7 \\
& & & & & 53$\pm$1* & 99.7 \\
\hline
\end{tabular}
\end{table*}
\begin{table*}
\centering
\caption{\label{top_hits_oj} The same of Table \ref{top_hits_pks0537} for OJ 287.
}
\begin{tabular}{|c|c|c|c||c||}
\hline\hline
Band & $T_{1}$ & $T_{2}$ & $T_{3}$ & $S \% \gtrsim$ \\
& [d] & [d] & [d] & \\
\hline\hline
100 MeV-300 Gev & 412$\pm$25 & & & 99.0 \\
\hline
V & 435$\pm$24 & & & 99.7 \\
\hline
R & 436$\pm$27 & & & 99.0 \\
\hline
J & 436$\pm$25 & & & 99.7 \\
& & 303$\pm$12* & & 99.0 \\
\hline
K & 438$\pm$22 & & & 99.7 \\
& & 296$\pm$10* & & 99.7 \\
& & & 203$\pm$5 & 99.7 \\
\hline
\end{tabular}
\end{table*}
\begin{table*}
\centering
\caption{\label{top_hits_3c279} The same of Table \ref{top_hits_pks0537} for 3C 279.
}
\begin{tabular}{|c|c|c|c|c|c|c|c||c||}
\hline\hline
Band & $T_{1}$ & $T_{2}$ & $T_{3}$ & $T_{4}$ & $T_{5}$ & $T_{6}$ & $T_{7}$ & $S\%\gtrsim$ \\
& [d] & [d] & [d] & [d] & [d] & [d] & [d] & \\
\hline\hline
100 MeV-300 GeV & & & & & 39$\pm$1 & & & 99.7 \\
& & & & & & & 24$\pm$1 & 99.7 \\
\hline
V & & 256$\pm$15 & & & & & & 99.7 \\
\hline
R & & 256$\pm$12 & & & & & & 99.0 \\
& & & 71$\pm$1 & & & & & 99.0 \\
& & & & 66$\pm$1 & & & & 99.0 \\
& & & & & & 29$\pm$1* & & 99.0 \\
& & & & & & & 24$\pm$1 & 99.7 \\
\hline
K & 931 $\pm$ 46 & & & & & & & 99.7 \\
& & 263$\pm$5 & & & & & & 99.7 \\
\hline
\end{tabular}
\end{table*}
\begin{table*}
\centering
\caption{\label{top_hits_pks1510} The same of Table \ref{top_hits_pks0537} for PKS 1510-089.}
\begin{tabular}{|c|c|c|c|c|c|c|c||c||}
\hline\hline
Band & $T_{1}$ & $T_{2}$ & $T_{3}$ & $T_{4}$ & $T_{6}$ & $S \% \gtrsim$ \\
\hline\hline
100 MeV-300 Gev & & & & & 115$\pm$5 & 99 \\
\hline
V & 490$\pm$34 & & & & & 99.0 \\
& & 325$\pm$16 & & & & 99.7 \\
\hline
R & 490$\pm$37 & & & & & 99.7 \\
& & 325$\pm$13 & & & & 99.0 \\
& & & & 206$\pm$9 & & 99.7 \\
\hline
J & 474$\pm$36 & & & & & 99.7 \\
& & 321$\pm$15 & & & & 99.0 \\
& & & & 203$\pm$9 & & 99.7 \\
\hline
K & 474$\pm$34 & & & & & 99.7 \\
& & 321$\pm$16 & & & & 99.0 \\
& & & 262$\pm$10* & & & 99.0 \\
& & & & 207$\pm$10 & & 99.7 \\
\hline
\end{tabular}
\end{table*}
\begin{table*}
\centering
\caption{\label{top_hits_pks2005} The same of Table \ref{top_hits_pks0537} for PKS 2005-489.
}
\begin{tabular}{|c|c|c|c|c||c||}
\hline\hline
Band & $T_{1}$ & $T_{2}$ & $T_{3}$ & $S\%\gtrsim $ \\
& [d] & [d] & [d] & \\
\hline\hline
V & 719$\pm$64* & & & 97 \\
& & 360$\pm$44* & & 99 \\
& & & 93$\pm$2* & 99.7 \\
\hline
R & 693$\pm$74* & & & 99 \\
& & 381$\pm$48* & & 99.7 \\
& & & 93$\pm$1* & 99 \\
\hline
J & 683$\pm$51* & & & 99 \\
& & 381$\pm$39* & & 99.7 \\
\hline
K & 722$\pm$47 & & & 99 \\
& & 361$\pm$28* & & 99 \\
\hline
\end{tabular}
\end{table*}
\begin{table*}
\centering
\small
\caption{\label{top_hits_pks2155} The same of Table \ref{top_hits_pks0537} for PKS 2155-304. }
\begin{tabular}{|c|c|c|c|c|c|c|c|c|c||c||}
\hline\hline
Band & $T_{1}$ & $T_{2}$ & $T_{3}$ & $T_{4}$ & $T_{5}$ & $T_{6}$ & $T_{7}$ & $T_{8}$ & $T_{9}$ & $S \% \gtrsim$ \\
& [d] & [d] & [d] & [d] & [d] & [d] & [d] & [d] & [d] & \\
\hline\hline
100 MeV-300 GeV & 642$\pm$59 & & & & & & & & & 99.7 \\
& & & & & & & & 61$\pm$1* & & 99.0 \\
& & & & & & & & & 52$\pm$1 & 99.0 \\
\hline
V & & & 318$\pm$14 & & & & & & & 99.0 \\
& & & & & & & 75$\pm$1 & & & 99.7 \\
& & & & & & & & 64$\pm$1* & & 99.0 \\
\hline
R & & & 318$\pm$14 & & & & & & & 99.0 \\
\hline
J & & & 310$\pm$15 & & & & & & & 99.0 \\
& & & & & & & 76$\pm$1 & & & 99.0 \\
& & & & & & & & 63$\pm$1* & & 99.0 \\
\hline
K & & 453$\pm$16 & & & & & & & & 99.7 \\
& & & 311$\pm$14 & & & & & & & 99.7 \\
& & & & 166$\pm$2 & & & & & & 99.7 \\
& & & & & 151$\pm$2 & & & & & 99.7 \\
& & & & & & 93$\pm$1* & & & & 99.0 \\
& & & & & & & 76$\pm$1 & & & 99.0 \\
\hline
\end{tabular}
\end{table*}
\clearpage
\begin{figure*}
\centering
\includegraphics[trim=0cm 2.5cm 0cm 0cm,clip,width=0.7\textwidth]{f01_3C279_COL+ck_waves}
\caption
{\label{lc} REM near-infrared and optical nightly averaged light curves of 3C 279.
The light curve of the check star is reported in J band (black points) with the indicated
displacements $\Delta$m.}
\end{figure*}
\begin{figure*}
\centering
\includegraphics[trim=0cm 1cm 0cm 0cm,clip,width=.8\textwidth]{f02_PKS0537_plotlc}
\caption{
\label{3lc_pks0537}
Weekly averaged {Fermi} $\gamma$-ray light curve in the 100
MeV - 300 GeV energy range (\textit{top panel}, yellow points).
Nightly averaged REM and SMARTS light curves in R and K bands
are also reported
(\textit{central and bottom panels}, yellow points).
Flux is
in photons $ \cdot $ s$^{-1}\cdot$ cm$^{2}$
for the $\gamma$-ray light curve and in Jy for the NIR-optical light curves.
Errors are omitted for readability.
The blue lines are the sinusoidal artificial models (see text).
The amplitudes are
A=0.001346 Jy in R (T=150 d), \
and A=0.00647 Jy in K band (T=152 d).
}
\end{figure*}
\begin{figure*}
\centering
\includegraphics[trim=0cm 1cm 0cm 0cm,clip,width=.8\textwidth]{f03_OJ287_plotlc_2_TS4}
\caption{ \label{3lc_oj287}
Same of Figure \ref{3lc_pks0537} for OJ 287.
The amplitudes of the sinusoidal curves are
A=3.60 $\cdot$ 10$^{-8}$ photons $\cdot$ s$^{-1}$ $\cdot$ cm$^{2}$ \ in 100 MeV-300 GeV (T=398 d), \
A=0.0023 Jy in R (T=436 d) \
and A=0.0179 Jy in K band (T=438 d).
}
\end{figure*}
\begin{figure*}
\centering
\includegraphics[trim=0cm 1cm 0cm 0cm,clip,width=.8\textwidth]{f04_3C279_plotlc}
\caption{ \label{3lc_3c279}
Same of Figure \ref{3lc_pks0537} for 3C 279.
The amplitude $A$ for the sinusoidal curve with a 263 d period in K-band is 0.0133 Jy.
}
\end{figure*}
\begin{figure*}
\centering
\includegraphics[trim=0cm 1cm 0cm 0cm,clip,width=.8\textwidth]{f05_PKS1510_plotlc_large}
\caption{ \label{3lc_pks1510}
Same of Figure \ref{3lc_pks0537} for PKS 1510-089. The prominent flare occurring
on $\sim$ 54960 MJD is partially cut for an easier visualization of the data.
Amplitudes of the sinusoidal curves are
A=0.000497 Jy in R (T=206 d), and A=0.0048 Jy in K band (T=207 d and T=474 d).
}
\end{figure*}
\begin{figure*}
\centering
\includegraphics[trim=0cm 1cm 0cm 0cm,clip,width=.8\textwidth]{f06_PKS2155_plotlc}
\caption{ \label{3lc_pks2155}
Same of Figure \ref{3lc_pks0537} for PKS 2155-304.
Amplitudes of the sinusoidal curves are
A=3.792 $\cdot$ 10$^{-7}$ photons $\cdot$ s$^{-1} \cdot$ cm$^{2}$
in 100 MeV-300 GeV (T=642 d),
A=0.00809 Jy in R (T=313 d),
and A=0.0194 Jy in K band (T=311 d).
}
\end{figure*}
\begin{figure*}
\centering
\includegraphics[trim=0cm 0cm 0cm 0.5cm,clip,width=0.45\textwidth]{f07_PKS0537_g1_1000_dos_r31_hifac1rm_B.eps}
\includegraphics[width=.45\textwidth]{f08_PKS0537_RSR_single_FLUXEXT_ave_hifac05}
\includegraphics[width=.45\textwidth]{f09_PKS0537_RSK_single_FLUXEXT_ave_hifac07}
\caption{ \label{pow_pks0537}
Bias-corrected power spectra (black line) of the blazar sample
in 100 MeV-300 GeV from the Fermi
$\gamma$-ray light curves, and
in K and R bands from REM + SMARTS photometry
(see also Table \ref{top_hits_pks0537}-\ref{top_hits_pks2155}).
The power is the output obtained by the procedure of \citet[REDFIT]{Schulz2002}
normalized with the variance.
Curves in each panel, starting from the bottom, are: the theoretical red-noise spectrum,
the 99.0\% (2.5$\sigma$) and 99.7\% (3$\sigma$) $\chi^{2}$
significance levels.
Periods in days corresponding to the prominent peaks are marked.
}
\end{figure*}
\begin{figure*}
\centering
\includegraphics[width=.45\textwidth]{f10_OJ287_g1_dos_r31_hifac07.eps}
\includegraphics[width=.45\textwidth]{f11_OJ287_R_dos_r31_sa_hifac06}
\includegraphics[width=.45\textwidth]{f12_OJ287_K_dos_r31_sa_hifac085}
\caption{\label{pow_oj287} The same of Figure \ref{pow_pks0537} for OJ 287.}
\end{figure*}
\begin{figure*}
\centering
\includegraphics[width=.45\textwidth]{f13_3C279_g1_dos_r31_hifac07_900dpeak}
\includegraphics[width=.45\textwidth]{f14_3C279_R_dos_r33_hifac05.eps}
\includegraphics[width=.45\textwidth]{f15_3C279_K_dos_r31_hifac05_900dpeak}
\caption{\label{pow_3c279} The same of Figure \ref{pow_pks0537} for 3C 279.}
\end{figure*}
\begin{figure*}
\centering
\includegraphics[width=.45\textwidth]{f16_PKS1510_g1_dos_r31_hifac07.eps}
\includegraphics[width=.45\textwidth]{f17_PKS1510_R_dos_r31_hifac05_new.eps}
\includegraphics[width=.45\textwidth]{f18_PKS1510_K_dos_r31_hifac05_new.eps}
\caption{\label{pow_pks1510} The same of Figure \ref{pow_pks0537} for PKS 1510-089.}
\end{figure*}
\begin{figure*}
\centering
\includegraphics[width=.45\textwidth]{f19_PKS2005_R_dos_r31_hifac05.eps}
\includegraphics[width=.45\textwidth]{f20_PKS2005_K_dos_r31.eps}
\caption{\label{pow_pks2005} The same of Figure \ref{pow_pks0537} for PKS 2005-489.
}
\end{figure*}
\begin{figure*}
\centering
\includegraphics[width=.45\textwidth]{f21_PKS2155_g1_dos_r31_hifac07_new.eps}
\includegraphics[width=.45\textwidth]{f22_PKS2155_R_dos_r31_hifac05_new.eps}
\includegraphics[width=.45\textwidth]{f23_PKS2155_K_dos_r31_hifac05_new.eps}
\caption{\label{pow_pks2155} The same of Figure \ref{pow_pks0537} for PKS 2155-304.}
\end{figure*}
%---------
%---------
\clearpage
\begin{thebibliography}{}
\bibitem[Abdo et al.(2010)]{Abdo2010} Abdo, A.~A., Ackermann,
M., Ajello, M., et al.\ 2010, \apj, 722, 520
\bibitem[\protect\citeauthoryear{Bonning et
al.}{2012}]{Bonning2012} Bonning E., et al., 2012, ApJ, 756, 13
%
\bibitem[Camenzind
\& Krockenberger(1992)]{Camenzind1992} Camenzind, M., \& Krockenberger, M.\ 1992, \aap, 255, 59
\bibitem[Covino et al.(2004)]{Covino2004} Covino, S., Stefanon,
M., Sciuto, G., et al.\ 2004, \procspie, 5492, 1613
\bibitem[\protect\citeauthoryear{Deeming}{1975}]{Deeming1975} Deeming T.~J., 1975, Ap\&SS, 36, 137
\bibitem[\protect\citeauthoryear{Dexter \& Fragile}{2011}]{Dexter2011}
Dexter J., Fragile P.~C., 2011, ApJ, 730, 36
\bibitem[\protect\citeauthoryear{Edelson et
al.}{2002}]{Edelson2002} Edelson R., Turner T.~J., Pounds K.,
Vaughan S., Markowitz A., Marshall H., Dobbie P., Warwick R., 2002, ApJ,
568, 610
\bibitem[\protect\citeauthoryear{Falomo, Pian,
\& Treves}{2014}]{Falomo2014} Falomo R., Pian E., Treves A., 2014, A\&ARv, 22, 73
\bibitem[Fragile et al. (2007)]{Fragile2007} Fragile P.~C., Blaes O.~M.,
Anninos P., Salmonson J.~D., 2007, ApJ, 668, 417
\bibitem[Gab{\'a}nyi et
al.(2007)]{Gabanyi2007} Gab{\'a}nyi, K.~{\'E}., Marchili, N., Krichbaum, T.~P., et al.\ 2007, \aap, 470, 83
\bibitem[Ghisellini et al. (2011)]{Ghisellini2011} Ghisellini G.,
Tavecchio F., Foschini L., Ghirlanda G., 2011, MNRAS, 414, 2674
\bibitem[Ghisellini(2013)]{Ghisellini2013} Ghisellini, G.\ 2013,
Lecture Notes in Physics, Radiative Processes in High Energy Astrophysics,
Berlin Springer Verlag, 873
\bibitem[Godfrey et al.(2012)]{Godfrey2012} Godfrey, L.~E.~H.,
Lovell, J.~E.~J., Burke-Spolaor, S., et al.\ 2012, \apjl, 758, L27
\bibitem[\protect\citeauthoryear{Gonz{\'a}lez-P{\'e}rez, Kidger,
\& Mart{\'{\i}}n-Luis}{2001}]{Gonzalez2001}
Gonz{\'a}lez-P{\'e}rez J.~N., Kidger M.~R., Mart{\'{\i}}n-Luis F., 2001, AJ, 122, 2055
\bibitem[Graham et al.(2015)]{Graham2015} Graham, M.~J.,
Djorgovski, S.~G., Stern, D., et al.\ 2015, \nat, 518, 74
\bibitem[\protect\citeauthoryear{Hardee
\& Rosen}{1999}]{Hardee1999} Hardee P.~E., Rosen A., 1999, ApJ, 524, 650
\bibitem[\protect\citeauthoryear{Hasselmann}{1976}]{Hasselmann1976}
Hasselmann K., Theory Tellus,1976, 28 (6), 473
\bibitem[\protect\citeauthoryear{Hudec et
al.}{2013}]{Hudec2013} Hudec R., Ba{\v s}ta M., Pihajoki P., Valtonen M., 2013, A\&A, 559, A20
\bibitem[Israel
\& Stella(1996)]{Israel1996} Israel, G.~L., \& Stella, L.\ 1996, \apj, 468, 369
\bibitem[Kidger(2000)]{Kidger2000} Kidger, M.~R.\ 2000, \aj, 119,
2053
\bibitem[\protect\citeauthoryear{Larionov et
al.}{2013}]{Larionov2013} Larionov V.~M., et al., 2013, ApJ, 768, 40
\bibitem[Lehto
\& Valtonen(1996)]{Lehto1996} Lehto, H.~J., \& Valtonen, M.~J.\ 1996, \apj, 460, 207
\bibitem[Li et al.(2015)]{Li2015} Li, H.~Z., Chen, L.~E., Yi,
T.~F., et al.\ 2015, \pasp, 127, 1
\bibitem[Marscher
\& Gear(1985)]{Marscher1985} Marscher, A.~P., \& Gear, W.~K.\ 1985, \apj, 298, 114
\bibitem[\protect\citeauthoryear{Marscher}{1992}]{Marscher1992} Marscher, A. P., Gear, W. K., \& Travis, J. P. 1992, in Variability of Blazars,
ed. E. Valtaoja \& M. Valtonen (Cambridge Univ. Press), 85
\bibitem[\protect\citeauthoryear{Marscher}{2014}]{Marscher2014}
Marscher A.~P., 2014, ApJ, 780, 87
\bibitem[Massaro et al. (2012)]{Massaro2012} Massaro, E., Giommi, P., Leto, C.;, et al\ 2012,
Multifrequency Catalogue of Blazars (3rd Edition)
(ARACNE Ed.)
\bibitem[Mattox et al.(1996)]{Mattox1996} Mattox, J.~R., Bertsch,
D.~L., Chiang, J., et al.\ 1996, \apj, 461, 396
\bibitem[\protect\citeauthoryear{Nandra et al.}{1997}]{Nandra1997}
Nandra K., George I.~M., Mushotzky R.~F., Turner T.~J., Yaqoob T., 1997,
ApJ, 476, 70
\bibitem[Ostorero et
al.(2004)]{Ostorero2004} Ostorero, L., Villata, M., \& Raiteri, C.~M.\ 2004, \aap, 419, 913
\bibitem[\protect\citeauthoryear{Peters}{1964}]{Peters1964}
Peters P.C., 1964, Phys. Rev. B, 136, 1224
\bibitem[Press(1978)]{Press1978} Press, W.~H.\ 1978, Comments on
Astrophysics, 7, 103
\bibitem[Raiteri et
al.(2001)]{Raiteri2001} Raiteri, C.~M., Villata, M., Aller, H.~D., et al.\ 2001, \aap, 377, 396
\bibitem[Raiteri et
al.(2010)]{Raiteri2010} Raiteri, C.~M., Villata, M., Bruschini, L., et al.\ 2010, \aap, 524, A43
\bibitem[Rani et al.(2009)]{Rani2009} Rani, B., Wiita, P.~J.,
\& Gupta, A.~C.\ 2009, \apj, 696, 2170
\bibitem[Sandrinelli et
al.(2014a)]{Sandrinelli2014a} Sandrinelli, A., Covino, S., \& Treves, A.\ 2014a, \aap, 562, A79
\bibitem[\protect\citeauthoryear{Sandrinelli et al.}{2014b}]
{Sandrinelli2014b} Sandrinelli A., Covino S., Treves A., 2014b, ApJ, 793, L1
\bibitem[\protect\citeauthoryear{Sbarrato et al.}{2012}]{Sbarrato2012}
Sbarrato T., Ghisellini G., Maraschi L., Colpi M., 2012, MNRAS, 421, 1764
\bibitem[Scargle(1982)]{Scargle1982} Scargle, J.~D.\ 1982, \apj,
263, 835
\bibitem[\protect\citeauthoryear{Schulz \& Statteger}{1997}]{Schulz1997}
Schulz, M. \& Statteger, K., 1997, Comput. Geosci., 23, 929 %-945
\bibitem[\protect\citeauthoryear{Schulz \& Mudelsee}{2002}]{Schulz2002}
Schulz, M. \& Mudelsee, M., 2002, Comput. Geosci., 28(3), 421
\bibitem[\protect\citeauthoryear{Schwarzenberg-Czerny}{1991}]{Schwar1991}
Schwarzenberg-Czerny A., 1991, MNRAS, 253, 198
\bibitem[Sillanp\"a\"a et al.(1988)]{Sillanpaa1988} Sillanp\"a\"a, A.,
Haarala, S., Valtonen, M.~J., Sundelius, B.,
\& Byrd, G.~G.\ 1988, \apj, 325, 628
\bibitem[\protect\citeauthoryear{Skrutskie et
al.}{2006}]{Skrutskie2006} Skrutskie M.~F., et al., 2006, AJ, 131,
1163
\bibitem[Sundelius et al.(1997)]{Sundelius1997} Sundelius, B., Wahde,
M., Lehto, H.~J., \& Valtonen, M.~J.\ 1997, \apj, 484, 180
\bibitem[\protect\citeauthoryear{Timmer
\& Koenig}{1995}]{Timmer1995} Timmer J., Koenig M., 1995, A\&A, 300, 707
\bibitem[van der Klis(1989a)]{vanderKlis1989a} van der Klis, M.\ 1989a,
NATO Advanced Science Institutes (ASI) Series C, 262, 27
\bibitem[Valtonen
\& Ciprini(2012)]{Valtonen2012} Valtonen, M., \& Ciprini, S.\ 2012, \memsai, 83, 219
\bibitem[\protect\citeauthoryear{van der
Klis}{1989b}]{vanderKlis1989b} van der Klis M., 1989b, ARA\&A, 27, 517
\bibitem[Vaughan(2005)]{Vaughan2005} Vaughan, S.\ 2005, \aap, 431, 391
\bibitem[Vaughan(2010)]{Vaughan2010} Vaughan, S.\ 2010, \mnras,
402, 307
\bibitem[Villata
\& Raiteri(1999)]{Villata1999} Villata, M., \& Raiteri, C.~M.\ 1999, \aap, 347, 30
\bibitem[\protect\citeauthoryear{Welch}{1967}]{Welch1967}
Welch, P.D., 1967, IEEE Transactions on Audio and Electroacoustics
15 (2), 70
\bibitem[Zerbi et al.(2004)]{Zerbi2004} Zerbi, F.~M., Chincarini,
G., Ghisellini, G., et al.\ 2004, \procspie, 5492, 1590
\bibitem[\protect\citeauthoryear{Zhang et al.}{2014}]{Zhang2014}
Zhang B.-K., Zhao X.-Y., Wang C.-X., Dai B.-Z., 2014, RAA, 14, 933
\bibitem[\protect\citeauthoryear{Zhou
\& Sornette}{2002}]{Zhou2002} Zhou W.-X., Sornette D., 2002, IJMPC, 13, 137
\end{thebibliography}
\end{document}