The 28th International Cosmic Ray Conference
1
Hourly Spectral Variability of Mrk 421
F. Krennrich^^, I.H. Bond, P.J. Boyle, S.M. Bradbury, J.H. Buckley, D. Carter-
Lewis, O. Celik, W. Cui, M. Daniel, M. D'Vali, I.de la Calle Perez, C. Duke,
A. Falcone, D.J. Fegan, S.J. Fegan, J. P. Finley, L.F. Fortson, J. Gaidos, S. Gam-
mell, K. Gibbs, G.H. Gillanders, J. Grube, J. Hall, T.A. Hall, D. Hanna,
A.M. Hillas, J. Holder, D. Horan, A. Jarvis, M. Jordan, G.E. Kenny, M. Kertz-
man, D. Kieda, J. Kildea, J. Knapp, K. Kosack, H. Krawczynski, M.J. Lang,
S. LeBohec, E. Linton, J. Lloyd-Evans, A. Milovanovic, P. Moriarty, D. Muller,
T. Nagai, S. Nolan, R.A. Ong, R. Pallassini, D. Petry, B. Power-Mooney, J. Quinn,
M. Quinn, K. Ragan, P. Rebillot, P.T. Reynolds, H.J. Rose, M. Scliroedter,
G. Sembroski, S.P. Swordy, A. Syson, V.V. Vassiliev, S.P. Wakely, G. Walker,
T.C. Weekes, J. Zweerink
(1) Physics & Astronomy Department, Iowa State University, Ames, I A, 50011
(2) The VERITAS Collaboration-see S.P. Wakely 's paper "The VERITAS Proto-
type" from these proceedings for affiliations
Abstract
Mrk 421 is the first TeV blazar found to exhibit significant spectral vari-
ability during strong flaring activity, showing hardening of the TeV spectrum in
high emission states. Mrk 421 is also known to exhibit flux variability on time
scales as short as 15 minutes. In this paper we present studies of hourly spectral
variability of Mrk 421 in 2001 using data from the Whipple Observatory 10 m
gamma-ray telescope.
1. Introduction
The AGN phenomenon is generally associated with strongly varying fluxes
at all wavelengths on time scales of hours to several months and years. At TeV
gamma-ray energies, flux variability as short as 15 minutes have been observed for
Mrk 421 (Gaidos et al., 1996). Flux variations are useful information providing
the grounds for dynamical tests for particle acceleration and/or emission models,
e.g., constraining cooling time scales and physical parameters of the source such
as the magnetic field and size of the emission region.
To further the understanding of non-thermal emission from AGN jets,
multiwavelength observations are pursued by measuring the spectrum over many
orders of magnitude in energy and its variations. Correlations between X-ray and
TeV emission were found in various flaring episodes for Mrk 421 (Buckley et al.
pp. ©2003 by Universal Academy Press, Inc.
2
20
1 Q
3
IS
c
1 f,
lU
e
14
12
Q
10
•4— >
o
8
4
2
lU
-73
-1.8
-2
;h
o
-2.2
&
-2.4
-2.6
-2.8
-3
-3.2
* * *
+ ♦ ♦
^^^^
9 10
UTC (h)
Fig. 1. The lightcurve of a flare of Mrk 421 on March 19 2001 is shown together
with the differential spectral index. The probability that the spectral variability
is statistical is 3.1 x 10~^ (significance ~ S.Gcj). The lower plot shows a division
(dotted line) into three different stages of the flare into preflare, rising flare and
postflare.
1996; Maraschi et al. 1999; Jordan et al. 2001). A next step in testing jet models
can be provided by measurements of the spectral variations of an X-ray (Fossati
et al. 2000) and contemporaneous TeV gamma-ray flare. Results of spectral
variability of Mrk 421 were shown previously (Aharonian et al. 2002; Krennrich
et al. 2002) by the HEGRA and VERITAS collaborations. In this paper we
discuss spectral variability of flares occuring on two nights in March 2001.
2. Results:
Unusually intense and lasting flaring activity of Mrk 421 in 2001 gave
excellent statistics and detailed features of its energy spectrum have been derived:
Mrk 421 exhibits a curved spectrum that can be described by a power law with an
exponential cutoff around 4 TeV (Krennrich et al. 2001; Aharonian et al. 2002).
The energy spectral index varies as a function of flux when averaging over several
months of data, but no evidence for variability in the cutoff energy was found
when analyzing the entire 2001 data set. Therefore, in the following analysis of
spectral variability on hourly time scales, since it is based on spectra from small
subsets of the 2001 data with less statistics, we fit the data using a power law
3
20
1 s
1
1 o
16
1
14
12
o
•4— >
10
O
8
Si
Of
6
4
2
X
u
-1.8
-2
-2.2
o
pe
-2.4
-2.6
-2.8
-3
-3.2
* ♦
♦ ♦
5 6 7
9 10
UTC (h)
Fig. 2. The lightcurve of a flare of Mrk 421 on March 25 2001 is shown together with
the differential spectral index. The spectral index is constant despite the fact that
the flux changes by a factor of 2.5.
with a fixed exponential cutoff energy of 4.3 TeV (see also Krennrich et al. 2002).
Figure 1 shows the gamma-ray lightcurve for Mrk 421 on March 19 2001
(Jordan et al. 2001; Buckley 2001) for which the TeV observations provide almost
complete sampling of the rise and decay of a flare. The gamma-ray rate ranges
from a moderate level (~ 1 Crab) up to an eight fold flux increase in less than 4.5
hours. The lower part of figure 1 shows the spectral index measured at various
times during the flare. The hypothesis that the spectral index is constant during
the flare has a chance probability of 3.1 x 10~^, suggesting spectral variability
during this outburst of gamma-ray emission with a significance of ~ 3.6a.
When dividing the spectral index measurements into three different episodes,
preflare, rising flare and postflare, the hardest spectral index coincides with the
rise of the flare, whereas during the preflare and the postflare the spectrum ap-
pears to be softer. A similar correlation between hardening/softening of the en-
ergy spectrum and flux for Mrk 421 has been reported by the HEGRA collabo-
ration (Aharonian et al. 2002) for the nights of March 21/22 and 22/23 2001.
However, observations of flaring activity during other nights suggest that the re-
lation between spectral index and flux variability is more complex than suggested
by data from March 19, 21/22 and 22/23 2001. Representative of the complexity
of spectral characteristics during flares of other nights is a flare on March 25 2001
4
measured with the Whipple 10 m telescope. Figure 2 shows its lightcurve with
significant flux variations and a flux increase by a factor of 2.5 within 2 hours.
The spectral index during this flare is exceptionally hard {a = 1.82 ±0.04)
without any indications of spectral variability (hypothesis of constant spectral in-
dex has a probability of 78.5%.). Although substantial flux variations occured no
signiflcant spectral variations are detected, despite good statistics. It is notewor-
thy that the data from this flare do not include a preflare episode with fluxes of
the level of 1 Crab and do not include the postflare era. This may indicate the
spectral variations on hourly time scales may be dependent on the absolute flux.
3. Conclusions
In this paper we present two lightcurves for hourly flaring activity of
Mrk 421 together with spectral index variations. During a flare on March 19
2001 signiflcant spectral variability is suggested by the data showing a soft spec-
tral index in the preflare and postflare phase and a hard spectral index during
the rising phase of the flare. In contrast no spectral variations are observed for
the night of March 25 despite signiflcant flux variations.
One-component models that describe gamma-ray flares solely by an in-
crease in the maximum energy of the radiating particle distribution are ruled out
by this observation. A more comprehensive analysis involving a large number of
flares is beyond the scope of this paper but will be presented at the conference.
4. Acknowledgements
We acknowledge the technical assistance of E. Roache and J. Melnick.
This research is supported by grants from the U.S. Department of Energy, the
Enterprise Ireland and by PPARC in the UK.
5. References
Aharonian, F.A., et al. 1999, A&A, 351, 330
Aharonian, F.A., et al. 2002, A&A, 384, L23
Buckley, J.H., et al. 1996, ApJ, 472, L9
Buckley, J.H. 2001, AIP Conf. Proceedings 587, Gamma 2001, ed. S. Ritz, N.
Gehrels, C.R. Shrader, p 235
Fossati, G., et al. 2000, ApJ, 541, 166
Gaidos, J. A., et al. 1996, Nature, 383, 319
Jordan, M., et al. 2001, in Proc. 27th ICRC (Hamburg), OG2.3, 2691
Krennrich, F., et al. 2001, ApJ, 560, L45
Krennrich, F., et al. 2002, ApJ, 575, L9
Maraschi, L., et al. 1999, ApJ, 526, L81
5
