A&A 365, L162-L167 (2001)
XMM-Newton observations of Markarian 421![[*]](/icons/foot_motif.gif)
W. Brinkmann1 - S. Sembay2 - R. G. Griffiths2 - G. Branduardi-Raymont3 - M. Gliozzi1 - Th. Boller1 -
A. Tiengo4 - S. Molendi5 - S. Zane3
Send offprint request: W. Brinkmann
1 - Max-Planck-Institut für extraterrestrische Physik, Giessenbachstrasse,
85740 Garching, Germany
2 - X-Ray Astronomy Group, Department of Physics
and Astronomy, University of Leicester, LE1 7RH, UK
3 -
Mullard Space Science Laboratory, University College London, Holmbury St Mary,
Dorking, Surrey,
RH5 6NT, UK
4 - XMM-Newton Science Operation Centre,
Villafranca Satellite Tracking Station, 28080 Madrid, Spain
5 - Istituto di Fisica Cosmica ``G. Occhialini'', Via Bassini 15, 20133, Italy
Received 2 October 2000 / Accepted 30 October 2000
Abstract
The BL Lac object Mrk 421 was observed on May 25, 2000 during the
XMM-Newton Cal/PV phase. The high throughput of the X-ray telescopes and the spectral capabilities
of the instruments allow an uninterrupted temporal and spectral study of the
source with unprecedented time resolution. Mrk 421 was found at a relatively high state with a
2-6 keV flux of (1.3-1.9) 10-10 erg cm-2 s-1. The observed intensity variations by more than a factor of three at highest X-ray energies are accompanied by complex spectral variations with only a small time lag (
s) between the hard and soft photons. The (0.2-10) keV spectrum can be well fitted by a broken power law and no absorption structures are found in the source spectrum at the high spectral resolution of the transmission gratings.
Key words: BL Lac objects - general; galaxies: active - quasars; X-rays: general - radio sources: general
Author for correspondance: wpb@mpe.mpg.de
Mrk 421 is the brightest BL Lac object at X-ray and UV wavelengths and it
is the first extragalactic source discovered at TeV energies (Punch et al. 1992).
This nearby (z = 0.031) X-ray bright BL Lac has been observed by
essentially all
previous X-ray missions and shows remarkable X-ray variability
correlated with strong activity at TeV energies (e.g., Takahashi et al. 1996;
Maraschi et al. 1999).
BL Lacs are thought to be dominated by relativistic jets seen at small
angles to the line of sight (Urry & Padovani 1995),
and their radio-through-X-ray spectra are well fitted
by inhomogeneous jet models (Bregman et al. 1987).
However, the structure of the relativistic jets remains largely
unknown as the models are generally under-constrained by single epoch
spectra and the typical smooth and nearly featureless blazar
spectra can be reproduced by models with widely different assumptions
(e.g., Königl 1989).
Combining spectral and temporal information greatly constrains the
jet physics.
Time scales are related to the crossing times of the emission regions
which depend on wavelength and/or the time scales of micro-physical processes
like acceleration and radiative losses.
The measured lags between the light curves at different energies as well as
spectral changes during intensity variations allow to probe the micro-physics
of particle acceleration and radiation in the jet.
Thus XMM-Newton with its high sensitivity and broad
energy bandwidth is an ideal tool to study BL Lacs
as it allows spectroscopy with unprecedented time resolution,
uninterrupted by gaps because of the long period of the satellite orbit.
Mrk 421 was the first BL Lac object to be established as an X-ray source
(Ricketts et al. 1976; Cooke et al. 1978) and subsequent observations
indicated that the X-ray spectrum has a soft power law form (Mushotzky et al. 1978; Hall et al. 1981) which exhibits significant
variability (Mushotzky et al. 1979).
More detailed studies with IUE and EXOSAT showed that the variability
occurs on time scales of typically a day with an e-folding time scale
of
5 104 s (George et al. 1988). The source shows
a dichotomy of X-ray states: a low, soft state (
erg cm-2 s-1,
)
where
the source hardens when it brightens and a hard outburst state (
erg cm-2 s-1)
during which the spectral index remains at
.
In several Ginga observations, partly simultaneously with ROSAT,
Mrk 421 was found at intermediate fluxes of
erg cm-2 s-1
(Makino et al. 1992; Tashiro 1994).
The data indicated that the amplitude of the flux variations with time scales
of a few hours got larger with increasing energy and the
correlation between flux and spectral index was inconsistent with
that observed by EXOSAT. The quality of the spectral fits improved
considerably by using a broken power law or a power law with
exponential cut off and the Ginga spectrum was significantly
steeper than the simultaneous ROSAT spectrum.
Since its discovery as a TeV source several multi-wavelength campaigns
have been conducted to study possible time lags between the X-ray band
and TeV energies and to investigate the pronounced spectral evolution
during flares seen in X-rays with ASCA and BeppoSAX
(Macomb et al. 1995, 1996;
Takahashi et al. 1996; Fossati et al. 1998; Maraschi et al. 1999).
The source generally shows a complex behavior. While
Takahashi et al. (1996) found a lag of about 4000 s between the soft
(0.5-1.0 keV) photons and the hard band (2-7.5 keV),
which was interpreted as an effect of radiative cooling, recent ASCA
observations show both, positive and negative lags (Takahashi et al. 2000).
BeppoSAX observations of a flare in April 1998, simultaneously observed
at TeV energies, showed that the hard photons lag the soft ones by
2-3 ksec and that, while the light curve is symmetric at softest X-ray energies,
it becomes increasingly asymmetric at higher energies with the
decay being slower than the rise (Fossati et al. 2000).
Fitting the ASCA data by a simple power law Takahashi et al. (1996) find
that an absorbing column density considerably higher than the Galactic
value of
cm-2 (Elvis et al. 1989)
is required to obtain
acceptable fits. Fixing the absorption at the Galactic value a
broken power law model provides a better fit than a simple power
law, but the
is often un-acceptable. With these
models the break energy is at
1.5 keV, and the change of the
power law index at the break point is
.
With the wider energy range of BeppoSAX it became clear that these simple
models are not adequate descriptions of the downward curved Synchrotron spectra
(Fossati et al. 2000) and continuously curved shapes had to be employed
(Inoue & Takahara 1996; Tavecchio et al. 1998).
The Synchrotron peak energy varied between 0.4-1 keV, the spectral
index at an energy of 5 keV between
.
Both quantities are correlated with the X-ray flux: the peak energy
positively, the spectral slope inversely: with increasing
flux the synchrotron peak shifts to higher energies and the spectrum at
5 keV gets flatter.
Most of these results were obtained from data integrated over
wide time intervals (typically one satellite orbit) and from
giant flares with time scales of a day.
Uninterrupted data with high temporal and spectral resolution
can only be provided by XMM-Newton with its high sensitivity,
spectral resolving power, and broad
energy band.
Mrk 421 was observed with the European Photon Imaging Camera (EPIC) and
the RGS during orbit 84
(May 25, 2000) of the calibration and performance verification phase
(Cal/PV) of XMM-Newton. The EPIC-PN camera was first operated from
UT 3:53 - UT 10:11 in the fast timing mode, then, from UT 11:34 to UT 19:34, in
Small Window (SW) mode.
As the data analysis tools for the timing mode are not yet fully established
we will restrict our discussion mainly to the second part of the observation,
where the PN camera was operated in the SW mode.
This mode, with a frame integration time of
5.7 msec was chosen to
avoid photon pile-up in this strong source. Overall, a total of more
than 8.9 million counts were accumulated in SW mode in a net observation time of
33 ksec, predominantly originating from the source itself.
Two separate observations of Mrk 421 were performed with the MOS cameras,
one in timing mode and the other in partial window mode with a free
running readout (PRI PART W4). During both observations the medium filter
was selected. For the purposes of this paper, however, only data taken in
partial window mode is used. In PRI PART W4 mode only a fraction of the
central chip (
pixels) is read out giving a time resolution of 0.2 s. Exposure times of 7 ksec and 24 ksec were obtained by MOS1 and MOS2
respectively. A standard reduction of raw MOS event lists was performed,
using the XMM-Newton Science Analysis System (SAS) and only ``real''
X-ray events with a pattern 0-12 were used to create spectra and light
curves for further analysis.
 |
Figure 1:
Light curves of Mrk 421 during the XMM-Newton
observation
in the different instruments |
Open with DEXTER |
MOS data and PN imaging data only have a small overlap in time. Due to a radiation alert
the MOS cameras were switched off shortly after the PN SW mode
observation began and were not turned on again.
So the
30 ksec of available MOS time largely
precedes the PN observation; the actual overlap is only about
5 ksec.
Both RGS instrument chains also observed Mrk 421 in two
exposures for a total observing time of 64 ksec; thus the RGS coverage is
slightly longer than that of EPIC. The data were acquired in the RGS
standard spectroscopy mode, where the full spectrum is read-out every
6 s (for details on the RGS instrument see den Herder et al., this volume).
The data were processed with the SAS,
using the most up-to-date calibration parameters.
Figure 1 shows the light curves of the whole observation for the
different instruments.
The count rate plots provide only a qualitative picture
as the photons in the individual instruments are selected according
to different criteria, neither covering the same energy
ranges nor originating from identical spatial regions around the source.
This is most clearly seen in the PN data taken in Fast Timing and in
Small Window mode, which do not match in count rate
when the modes were changed.
The quantitatively correct conversion from count rates to fluxes
is detector and operation mode dependent and must be deferred
to a later paper.
No scientific data are available from the
Optical Monitor for this observation of Mrk 421.
 |
Figure 2:
The EPIC PN light curve of Mrk 421 during orbit 84.
The time is counted from the start of the SW mode observation. Only single
events were taken and a time binning of 100 seconds has been used.
Shown from the top are: the total 0.2-10 keV count rate; the count rate for
the soft photons keV, for the
hard photons E > 3 keV, and at the bottom the hardness ratio
between the hard and the soft photons |
Open with DEXTER |
Light curves were produced by using the PN photon event files produced
by the SAS. Only single events with energies
0.2 keV
keV were selected
from a circle of 9 detector pixels radius (corresponding to
)
around the source center;
the background was taken from the outer source-free
region of the detector and amounted to about 3%
of the source counts.
The photons were first binned into 10 s intervals which turned
out to be necessary as the accepted time intervals are frequently
disrupted by short gaps where the camera fell into counting mode.
Finally, the binned data were summed up into
typically 100 s bins providing an excellent signal to noise ratio.
Figure 2 shows the background
subtracted light curves in three energy intervals: the total 0.2-10 keV
count rate at the top, then the soft (
keV) and the
hard (E > 3 keV) light curves in the middle, all in 100 s time bins
and, at the bottom, the hardness ratios of the count rates
in the
bands.
The total light curve shows intensity variations of about a factor of 1.5;
at high energies the maximal variation is much stronger, about a factor
of 3.4 during the SW mode exposure.
The source gets harder when it flares, in accordance to previous observations.
The hard flux is considerably more variable than the soft flux as
quantified by
the excess variance (Nandra et al. 1997): for the soft band it is
while for the hard band it is
.
The source was found in a relatively high state. From the spectral fit to
the total data set (see Sect. 3) we obtain an average 2-6 keV flux of
(1.3 - 1.9) 10-10 erg cm-2 s-1 which is considerably
higher than the ``classical'' soft state, but still lower than in the
April 1998 observation of BeppoSAX (Fossati et al. 2000).
 |
Figure 3:
Spectral evolution during the two prominent flares
occuring at 3 ksec and 25 ksec after the start of the
observation.
Plotted are the hardness ratios as function of total count rate and
a few error bars at the beginning of the first peak.
Numbers indicate the temporal sequences of the data |
Open with DEXTER |
The light curve is rather complex and seems to consist of several,
partly overlapping
flares on top of smooth intensity variations. The
spectral evolution during the two prominent flares at
3 ksec and
25 ksec after the start of the observation shows a clockwise hysteresis,
as seen in Fig. 3. This effect was
already noted in previous ASCA observations (Takahashi et al. 1996) but
as a property of the long term (
1 day) intensity variations of
the source.
The flares seem to have nearly linear rise- and decay profiles which
are asymmetric with energy dependent time scales. In the total band
the first flare rises slower (
4.7 10-3 counts s-2)
than it decays (
8.8 10-3 counts s-2), but in
general an exact determination of these time scales is problematic
due to the occurrence of additional, overlapping smaller flares.
In Fig. 3 these lead to a break of the
clean intensity-hardness ratio correlation at the end of
the decaying part of the flares.
The flare profiles are energy dependent and we
quantified the time delays between the hard and soft photons
employing the Z-transformed discrete cross-correlation function algorithm
(ZDCF) of Alexander (1997). The soft photons lag the
hard by
265+116-102 s where the
errors were
determined by using a Monte-Carlo method described by Peterson et al. (1998).
Lags of a few 1000 s as claimed from previous observations
(Takahashi et al. 1996; Fossati et al. 2000) must thus be related to
long-term flux changes and not to individual flares.
A structure function analysis is a method of
quantifying time variability
without the problems encountered in the traditional Fourier
analysis technique in case of unevenly sampled data.
The first-order structure function (Simonetti et al. 1985) measures
the mean deviation for data points separated by a time
lag
,
.
It is commonly characterized in terms of its slope:
.
One of the most useful features
of the structure function is its ability to discern the
range of time scales that contribute to the variations in the data set. For lags shorter than the smallest correlation time scale and for lags longer
than the longest correlation time scale, the structure function displays two
plateau states (b=0) at different levels.
These regions are linked by a curve
whose slope depends on the nature of the intrinsic variation of the source
(e.g. flicker noise, shot noise, etc.).
Figure 4 shows the structure function of Mrk 421 obtained from
single events during the SW mode observation. The ``wavy'' structure of the
increasing part must be related to the few individual flares in the light curve.
Mrk 421 shows little variation at time scales lower than
.
For longer time scales the slope of the structure function is
which indicates that the nature of the variation
of the source can be ascribed to shot noise.
The current data do not allow to see the
roll-over at a lag of about half a day, found by ASCA (Takahashi et al. 2000).
The result is fully consistent with the findings of
Hughes et al. (1992), who obtained an average slope of
for a sample of 20 BL Lac objects.
The X-ray spectrum of Mrk 421 was repeatedly studied in the past and
fitted to a featureless power law or broken power law.
By taking the single events from the whole PN SW mode data set,
employing the latest versions of the detector response and CTE corrections,
we find that a power law with frozen Galactic absorption
does not fit the 0.2-10 keV data (
= 2.025),
while the fit improves
leaving the absorption column density
free (
cm-2,
at a
for 1246 d.o.f).
 |
Figure 5:
Broken power law fit to the total PN Small Window
mode data of Mrk 421. Only single events were used |
Open with DEXTER |
A broken power law with free absorption gives an acceptable
fit to the data (see Fig. 5).
The best fit parameters are
cm-2,
,
keV,
at a
for 1244 d.o.f.
There are still systematic (instrumental)
deviations in the residuals at low energies around 0.3 keV and 0.5 keV,
as clearly visible in Fig. 5,
leading to an inexact determination of the absorption column density.
Note the extremely high statistical significance in the data: at lower
energies every spectral bin contains more than 2000 counts!
With these spectral parameters the source average flux in the 2-6 keV
bands amounts to (1.3-1.9) 10-10 erg cm-2 s-1,
where the quoted uncertainty originates mainly from the
currently uncertain contributions of the double, triple, and unrecognized
events to the ``real'' source flux.
In the RGS spatial and spectral domains (i.e. in the dispersion vs. cross-dispersion, and dispersion vs. CCD pulse height,
event distributions) data cuts were made in order to extract the
source spectra in both first
and second order. Spectra of the background (which is
1% of the source
flux) were extracted from a spatial region offset from the source in the cross
dispersion direction, and as wide as that used for the source. Instrumental
responses were built for first and second order spectra for RGS1 and RGS2
(in 3000 spectral bins).
The model that best fits the Mrk 421 RGS data is a broken power law;
other models were tried (but no ``curved'' models), but none matched the
data better.
The RGS1 first order spectra for the two exposures
were fitted simultaneously with the same absorbed broken power law model,
keeping all parameters tied, except for the normalizations which were allowed
to vary independently; the low-energy absorption column density was
fixed at the Galactic value.
The best fit parameters are
,
,
and a reduced
.
Data from the second RGS1 exposure (strictly simultaneous with the
spectra obtained from the PN camera) were analyzed on their own, with
similar results to the total dataset. Further, we find no evidence for any discrete spectral feature (neither in absorption, nor in emission) in the Mrk 421 data.
The RGS spectra allow us to set a 90% upper limit of 0.3 eV (0.012 Å)
to the EW of a narrow, saturated OVII resonance absorption line at 0.574 keV
(21.6 Å; these values correspond to 0.557 keV, and 22.3 Å,
respectively, at the redshift of the galaxy). The presence of such an
absorption feature was suggested by
Guainazzi et al. (1999) from BeppoSAX observations.
A similar RGS upper limit applies to the EW of an OVIII
L
resonant absorption line at 0.654 keV.
The flare at the beginning of the SW mode observation was observed by
the MOS1 camera as well and could thus be fit with both, the PN
and the MOS instruments.
A broken power law model with the absorption fixed at the Galactic value
resulted for the MOS1 in an acceptable fit (
=
1.04 for 213 d.o.f.) with:
,
,
keV.
The same model (fixed Galactic absorption and broken power law) did not
yield an acceptable fit in the PN; only when the
-value was left
as a free parameter we obtained acceptable fits with
.
While the fitted slope of the high energy power law remained stable
(
)
and nearly identical to the MOS1 value
in all fits, the other parameters depended strongly on the actual lower
boundary of the energy range for the fit. The fitted
always remains lower than the Galactic value.
Similarly, broken power law fits at different intensity levels of
Mrk 421 tend to yield nearly identical slopes at higher energies, only
the low energy slopes and the break energies seem to vary.
In conclusion, the
above fit comparisons indicate that the cross calibration of the
spectral responses still needs improving.
X-ray observations have been used frequently to
determine the physical conditions of the central engines of BL Lac
objects.
In most cases integration times over typically one satellite
orbit inhibited the study of irregularities in the
variability patterns and the puzzling spectral behavior of the sources
on shorter time scales.
XMM-Newton with its high sensitivity and broad
energy bandwidth allows spectroscopy with unprecedented time resolution,
uninterrupted by gaps because of the long period of the satellite orbit.
In the extended XMM-Newton Cal/PV observations of Mrk 421 in May 25, 2000
the source was found in a relatively high state with intensity variations
by a factor of more than 3.
In the hard energy band the source is considerably more variable
than in the soft band, but in the time resolved flares we
find only a small time lag between the hard and the soft photons.
This result and the fact that the shape of individual pulses and their
spectral evolution are resolvable on time scales of
100 s,
puts strong constraints on time dependent radiation models for
BL Lac jets (Georganopoulos & Marscher 1997; Kirk et al. 1998; Chiaberge & Ghisellini 1999; Kataoka et al. 1999).
The spectra can be well fitted by broken power laws with break energies
around 1-2 keV. Flux variations seem to affect the low energy power law
slope and the value of the break energy; the higher energy slope
appears to be quite stable.
A quantitative analysis, however, will have to await a further
reduction of the remaining calibration uncertainties.
Acknowledgements
We thank I. Papadakis for his help with the cross-correlation analysis
and K. Dennerl, S. Zavlin, and F. Haberl for producing off-line
event files with currently unpublished CTE corrections.
The XMM-Newton project is supported by the Bundesministerium fuer
Bildung und Forschung/Deutsches Zentrum fuer Luft- und Raumfahrt
(BMBF/DLR), the Max-Planck Society and the Heidenhain-Stiftung.
The Mullard Space Science Laboratory and Leicester University
acknowledge financial support
from the UK Particle Physics and Astronomy Research Council.
Part of this work was done in the TMR research network ``Accretion
onto black holes, compact stars and protostars'' funded by
the European Commission under contract number ERBFMRX-CT98-0195.
- Alexander, T. 1997, in Astronomical Time Series,
ed. D. Maoz, A. Sternberg, & E. M. Leibowitz (Dordrecht, Kluwer), 163
In the text
-
Bregman, J., Maraschi, L., & Urry, C. M. 1987, in Exploring
the Universe with the IUE Satellite, ed. Y. Kondo
(Dordrecht: Reidel), 685
In the text
-
Chiaberge, M., & Ghisellini, G. 1999, MNRAS, 306, 551
In the text
NASA ADS
- Cooke, B. A., Ricketts, M. J., Maccacaro, T., et al.
1978, MNRAS, 182, 489
In the text
NASA ADS
- Elvis, M., Lockman F. J., & Wilkes, B. J. 1989, AJ, 97, 777
In the text
NASA ADS
- Fossati, G., Chiappetti, L, Celotti, A., et al. 1998, Nucl. Phys. B. Proc. Supp., 69/1-3, 423
In the text
- Fossati, G., Celotti, A., Chiaberge, M., et al. 2000 [astro-ph/0005066, astro-ph/0005067]
In the text
-
Georganopoulos, M., & Marscher, A. P. 1997, in Relativistic Jets in AGN, ed. M. Ostrowski, et al. (Krakow, Poland), 313
In the text
-
George, I. M., Warwick, R. S., & Bromage, G. E. 1988,
MNRAS, 232, 793
In the text
NASA ADS
- Guainazzi, M., Vacanti, G., Malizia, A., et al. 1999,
A&A, 342, 124
In the text
NASA ADS
- Hall, R., Ricketts, M. J., Page, C., & Pounds, K. A. 1981, Space Sci. Rev., 30, 47
In the text
NASA ADS
- Hughes, P. A., Aller, H. D., & Aller, M. F. 1992, ApJ, 396, 469
In the text
NASA ADS
- Inoue, S., & Takahara, F. 1996, ApJ, 463, 555
In the text
NASA ADS
-
Kataoka, J., Takahashi, T., Makino, F., et al. 1999, ISAS
Research Note, 687
In the text
-
Kirk, J. G., Rieger, F. M., & Mastichiades, A. 1998, A&A, 333, 452
In the text
NASA ADS
-
Königl, A. 1989, in BL Lac Objects, ed. L. Maraschi,
T. Maccacaro, & M.-H. Ulrich (Berlin: Springer-Verlag), 321
In the text
- Macomb, D. J., Akerlof, C. W., Aller, D. H., et al. 1995, ApJ, 449, L99
In the text
NASA ADS
- Macomb, D. J., Akerlof, C. W., Aller, D. H., et al. 1996, ApJ, 459, L111
In the text
NASA ADS
-
Makino, F., Fink, H. H., Clavel, J. 1992, in Frontiers of X-ray Astronomy,
ed. Y. Tanaka, & K. Koyama (Tokyo: Univ. Acad. Press), 543
In the text
-
Maraschi, L., Fossati, G., Tavecchio, F., et al. 1999, ApJ, 526, L81
In the text
NASA ADS
- Mushotzky, R. F., Boldt, E. A., Holt, S. S., et al. 1978, ApJ, 226, L65
In the text
NASA ADS
- Mushotzky, R. F., Boldt, E. A., Holt, S. S., & Serlemitsos, P. J. 1979, ApJ, 232, L17
In the text
NASA ADS
-
Nandra, K., George, I. M., Mushotzky, R. F., Turner, T. J., & Yakoob, T. 1997, ApJ, 476, 70
In the text
NASA ADS
- Peterson, B. M., Wanders, I., Horne, K., et al. 1998,
PASP, 110, 660
In the text
NASA ADS
- Punch, M., Akerlof, C. W., Cawley, M. F., et al. 1992, Nat, 358, 477
In the text
NASA ADS
- Ricketts, M. J., Cooke, B. A., & Pounds, K. A. 1976, Nat, 259, 546
In the text
NASA ADS
- Simonetti, J. H., Cordes, M. J., & Heeschen, D. S. 1985, ApJ, 296, 46
In the text
NASA ADS
- Takahashi, T., Tashiro, M., Madejski, G., et al. 1996, ApJ, 470, L89
In the text
NASA ADS
- Takahashi, T., Kataoka, J., Madejski, G., et al. 2000 [astro-ph/0008505]
In the text
- Tashiro, M. 1994, Ph.D. Thesis, Univ. of Tokyo; ISAS Research Note, 549
In the text
- Tavecchio, F., Maraschi, L., & Ghisellini, G. 1998, ApJ, 509, 608
In the text
NASA ADS
-
Urry, C. M., & Padovani, P. 1995, PASP, 107, 803
In the text
NASA ADS
© ESO 2001