The Astrophysical Journal, 763:42 (11pp), 2013 January 20 doi:10.1088/0004-637X/763/1/42
C
2013. The American Astronomical Society. All rights reserved. Printed in the U.S.A.
X-RAY EMISSION FROM SUPERNOVAE IN DENSE CIRCUMSTELLAR MATTER
ENVIRONMENTS: A SEARCH FOR COLLISIONLESS SHOCKS
E. O. Ofek
1
, D. Fox
2
,S.B.Cenko
3
, M. Sullivan
4
,O.Gnat
5
, D. A. Frail
6
, A. Horesh
7
,A.Corsi
8
,R.M.Quimby
9
,
N. Gehrels
10
, S. R. Kulkarni
7
, A. Gal-Yam
1
,P.E.Nugent
11
, O. Yaron
1
, A. V. Filippenko
3
,M.M.Kasliwal
12
,
L. Bildsten
13,14
, J. S. Bloom
3
, D. Poznanski
15
, I. Arcavi
1
,R.R.Laher
16
, D. Levitan
7
, B. Sesar
7
, and J. Surace
16
1
Benoziyo Center for Astrophysics, Weizmann Institute of Science, 76100 Rehovot, Israel
2
Department of Astronomy and Astrophysics, Pennsylvania State University, University Park, PA 16802, USA
3
Department of Astronomy, University of California, Berkeley, CA 94720-3411, USA
4
Department of Physics, University of Oxford, Denys Wilkinson Building, Keble Road, Oxford OX1 3RH, UK
5
Racah Institute of Physics, The Hebrew University of Jerusalem, 91904 Jerusalem, Israel
6
National Radio Astronomy Observatory, P.O. Box O, Socorro, NM 87801, USA
7
Division of Physics, Mathematics, and Astronomy, California Institute of Technology, Pasadena, CA 91125, USA
8
LIGO Laboratory, Division of Physics, California Institute of Technology, MS 100-36, Pasadena, CA 91125, USA
9
Kavli IPMU, University of Tokyo, 5-1-5 Kashiwanoha, Kashiwa-shi, Chiba 277-8583, Japan
10
NASA-Goddard Space Flight Center, Greenbelt, MD 20771, USA
11
Lawrence Berkeley National Laboratory, 1 Cyclotron Road, Berkeley, CA 94720, USA
12
Observatories of the Carnegie Institution for Science, 813 Santa Barbara Street, Pasadena, CA 91101, USA
13
Kavli Institute for Theoretical Physics, Kohn Hall, University of California, Santa Barbara, CA 93106, USA
14
Department of Physics, Broida Hall, University of California, Santa Barbara, CA 93106, USA
15
School of Physics and Astronomy, Tel Aviv University, Tel Aviv, Israel
16
Spitzer Science Center, California Institute of Technology, MS 314-6, Pasadena, CA 91125, USA
Received 2012 May 30; accepted 2012 October 18; published 2013 January 4
ABSTRACT
The optical light curve of some supernovae (SNe) may be powered by the outward diffusion of the energy deposited
by the explosion shock (the so-called shock breakout) in optically thick (τ 30) circumstellar matter (CSM).
Recently, it was shown that the radiation-mediated and radiation-dominated shock in an optically thick wind
must transform into a collisionless shock and can produce hard X-rays. The X-rays are expected to peak at late
times, relative to maximum visible light. Here we report on a search, using Swift/XRT and Chandra,forX-ray
emission from 28 SNe that belong to classes whose progenitors are suspected to be embedded in dense CSM.
Our sample includes 19 Type IIn SNe, one Type Ibn SN, and eight hydrogen-poor superluminous SNe (SLSN-I
such as SN 2005ap). Two SNe (SN 2006jc and SN 2010jl) have X-ray properties that are roughly consistent with
the expectation for X-rays from a collisionless shock in optically thick CSM. However, the X-ray emission from
SN 2006jc can also be explained as originating in an optically thin region. Thus, we propose that the optical light
curve of SN 2010jl is powered by shock breakout in CSM. We suggest that two other events (SN 2010al and
SN 2011ht) were too X-ray bright during the SN maximum optical light to be explained by the shock-breakout
model. We conclude that the light curves of some, but not all, SNe IIn/Ibn are powered by shock breakout in CSM.
For the rest of the SNe in our sample, including all of the SLSN-I events, our X-ray limits are not deep enough and
were typically obtained too early (i.e., near the SN maximum light) for definitive conclusions about their nature.
Late-time X-ray observations are required in order to further test whether these SNe are indeed embedded in dense
CSM. We review the conditions required for a shock breakout in a wind profile. We argue that the timescale, relative
to maximum light, for the SN to peak in X-rays is a probe of the column density and the density profile above the
shock region. In SNe whose X-ray emission slowly rises, and peaks at late times, the optical light curve is likely
powered by the diffusion of shock energy in a dense CSM. We note that if the CSM density profile falls faster than a
constant-rate wind-density profile, then X-rays may escape at earlier times than estimated for the wind-profile case.
Furthermore, if the CSM has a region in which the density profile is very steep relative to a steady wind-density
profile, or if the CSM is neutral, then the radio free–free absorption may be sufficiently low for radio emission to be
detected.
Key words: stars: mass-loss – supernovae: general – supernovae: individual (SN 2006jc, SN 2010jl)
Online-only material: machine-readable table
1. INTRODUCTION
Circumstellar matter (CSM) around supernova (SN) progeni-
tors may play an important role in the emission and propagation
of energy from SN explosions. The interaction of the SN ra-
diation with optically thin CSM shells may generate emission
lines, with widths that are representative of the shell velocity or
the velocity of the shocked cool gas in t he post-shock region, as
in Type IIn SNe (SNe IIn; Schlegel 1990; Kiewe et al. 2012;see
Filippenko 1997 for a review of SN classification). The interac-
tion of SN ejecta with the CSM can power the light curves of
SNe by transformation of the SN kinetic energy into photons.
In cases where a considerable amount of optically thin (and
ionized) material is present around the exploding star, syn-
chrotron, and free–free radiation can emerge, and inverse-
Compton scattering can generate X-ray photons (e.g., Chevalier
& Fransson 1994; Horesh et al. 2012; Krauss et al. 2012).
For the Type IIn SN PTF 09uj, Ofek et al. (2010) suggested
that a shock breakout can take place in an optically thick wind
(see also Grassberg et al. 1971; Falk & Arnett 1977; Chevalier
1
The Astrophysical Journal, 763:42 (11pp), 2013 January 20 Ofek et al.
&Irwin2011; Balberg & Loeb 2011). This will happen if
the Thomson optical depth within the wind profile is c/v
sh
,
where c is the speed of light and v
sh
is the shock speed. Ofek
et al. (2010) showed that shock breakout in wind environments
produces optical displays that are brighter and have longer
timescales than those from the surfaces of red supergiants (e.g.,
Colgate 1974; Matzner & McKee 1999; Nakar & Sari 2010;
Rabinak & Waxman 2011; Couch et al. 2011). Chevalier &
Irwin (2011) extended this picture. Specifically, they discussed
CSM with a wind profile in which the wind has a cutoff at a
distance R
w
. If the optical depth at R
w
is c/v
s
, then the SN light
curve will have a slow decay (e.g., SN 2006gy; Ofek et al. 2007;
Smith et al. 2007). If the optical depth at R
w
is c/v
s
, then it will
have a faster decay (e.g., SN 2010gx; Pastorello et al. 2010a;
Quimby et al. 2011b). Moriya & Tominaga (2012) investigated
shock breakouts in general wind-density profiles of the form
ρ ∝ r
−w
. They suggested that, depending on the power-law
index w, shock breakouts in wind environments can produce
bright SNe without narrow emission lines (e.g., SN 2008es;
Gezari et al. 2009; Miller et al. 2009).
Recently Katz et al. (2011) and Murase et al. (2011)showed
that if the progenitor is surrounded by optically thick CSM, then
a collisionless shock is necessarily formed during the shock
breakout. Moreover, they argued that the energy emitted from
the collisionless shock in the form of high-energy photons and
particles is comparable to the shock-breakout energy. Further-
more, this process may generate high-energy (1 TeV) neutri-
nos. Although Katz et al. (2011) predicted that the photons are
generated with energy typically above 60 keV, it is reasonable
to assume that some photons will be emitted with lower energy
due to reprocessing of photons (see below) and the continuum
nature of the radiation. Chevalier & Irwin (2012) s howed that
Comptonization and inverse-Compton scattering of the high-
energy photons is likely to play an important role, and that the
high-energy photons will be absorbed.
Svirski et al. (2012) discuss the X-ray emission from colli-
sionless shocks. They show that at early times the X-rays will
be processed into the optical regime by the Compton process.
Therefore, at early times, the optical emission will be about
10
4
times stronger than the high-energy emission. With time, the
X-ray emission will become stronger, while the optical emis-
sion will decay. They conclude that for a CSM with a steady
wind profile (w = 2), X-ray emission may peak only at late
times, roughly 10–50 times the shock-breakout timescale. The
shock-breakout timescale, t
br
, is roughly given by the diffusion
timescale at the time of shock breakout. This timescale is also
equivalent to the radius at which the shock breaks out (r
br
)di-
vided by the shock velocity (v
s
;Weaver1976). If the main source
of optical photons is due to diffusion of the shock-breakout en-
ergy, the SN optical light rise time, t
rise
, will be equivalent to the
shock-breakout timescale. Therefore, X-ray flux measurements
and spectra of SNe embedded in dense CSM starting from the
explosion until months or years after maximum light are able to
measure the properties of the CSM around the SN progenitors
and the progenitor mass-loss history. This unique probe into the
final stages of massive star evolution has been only partially
exploited, at best.
Herein, we analyze the X-ray data for 28 SNe with light curves
that may be powered by a shock breakout from dense CSM,
and for which Swift/XRT (Gehrels et al. 2004) observations
exist. We use this sample to search for X-ray signatures of
collisionless shocks—emission at late times (months to years
after peak optical luminosity). We suggest that these signals
were observed in several cases, most notably in SN 2010jl
(Chandra et al. 2012b). Finally, we review the conditions for
a shock breakout in CSM with a wind profile, and we discuss
the importance of bound-free absorption and the possibility of
detecting radio emission from such SNe.
The structure of this paper is as follows. In Section 2 we
present the SN sample, while Section 3 presents the X-ray
observations. We review and discuss the model in Section 4,
and consider the observations in the context of the model in
Section 5. We summarize our conclusions in Section 6.
2. SAMPLE
Our sample is based on SNe found by amateur astronomers
and several surveys, including the Lick Observatory Supernova
Search (Li et al. 2000; Filippenko et al. 2001), the Catalina
Real-time Transient Survey (Drake et al. 2009a), Pan-STARRS1
(PS1; Kaiser et al. 2002), and the Palomar Transient Factory
17
(PTF; Law et al. 2009; Rau et al. 2009). Two SNe, PTF 09drs
and PTF 10tel, are reported here for the first time. We note that
many of the nearby or luminous SNe found by PTF are also
observed by Swift.
We selected a s ample of SNe in which the main source of
energy may be explained by diffusion of the explosion shock
energy through optically thick CSM around the progenitor. First,
we include SNe IIn within 200 Mpc. Objects that belong to
this class show relatively narrow (intermediate width) hydrogen
emission lines, an indication of the presence of optically
thin material somewhere around the progenitor. However, it
is unlikely that all SNe showing relatively narrow hydrogen
emission lines in their spectra are powered mainly by the
diffusion of shock energy in an optically thick environment.
One reason is that some SNe IIn show X-ray emission near
maximum optical light, which is not expected when optically
thick CSM is present (see Section 4). Furthermore, Moriya
& Tominaga (2012) suggest that not all SNe powered by
interaction of the ejecta with slowly moving material will
necessarily have narrow emission lines in their spectrum. We
note that some of the SNe IIn in our sample are peculiar (e.g.,
SN 2010jp/PTF 10aaxi; Smith et al. 2012).
Another relevant, but rare, class of objects are SNe Ibn. This
class is defined by the lack of hydrogen lines and the presence
of narrow helium emission lines. The only SN of this type in
our sample is SN 2006jc (Nakano et al. 2006; Foley et al. 2007;
Pastorello et al. 2008; Smith et al. 2008).
The third class of SNe we investigate here is the small group of
hydrogen-poor superluminous SN (SLSN-I; see review in Gal-
Yam 2012). Quimby et al. (2011b) used spectra of several such
events found by PTF, at intermediate redshift (z ≈ 0.5), to show
that these events, as well as SCP 06F6 (Barbary et al. 2009) and
SN 2005ap (Quimby et al. 2007), are spectroscopically similar.
This group of SNe continues to grow with new discoveries
(e.g., Chomiuk et al. 2011; Leloudas et al. 2012), and their
hosts were studied by Neill et al. (2011). Although the nature
of these events is not understood (e.g., Kasen & Bildsten 2010),
Quimby et al. (2011b) suggested that they may be powered by
a pulsational pair-instability SN (Rakavy et al. 1967; Woosley
et al. 2007). According to this hypothesis, the SN ejecta interact
with a dense shell of material, enriched with intermediate-mass
elements, that was expelled by the progenitor during previous
explosions (see also Ginzburg & Balberg 2012). This model is
tentatively supported by observations of SN 2006oz (Leloudas
17
http://www.astro.caltech.edu/ptf/
2
The Astrophysical Journal, 763:42 (11pp), 2013 January 20 Ofek et al.
Tab le 1
SN Sample
Name Type α
J2000
δ
J2000
t
rise
M
R
zt
peak
N
H
L
X
L
X
L
opt
FAP
min
(deg) (deg) (day) (mag) (MJD) (10
20
cm
−2
)(ergs
−1
)
PTF 09atu SLSN-I 247.60229 +23.64029 30: −22.5 0.501 55060 4.05 <1.9 × 10
44
0.7 1.00
PTF 09cnd SLSN-I 243.03725 +51.48782 50 −22.8 0.258 55080 1.67 <8.0 × 10
42
0.02 1.00
PTF 09cwl/SN 2009jh SLSN-I 222.29200 +29.41983 50 −22.5: 0.349 55060 1.51 <1.1 × 10
44
0.4 1.00
SN 2010gx/PTF 10cwr SLSN-I 171.44448 −8.82810 20 −21.7: 0.231 55280 3.78 <9.9 × 10
42
0.07 1.00
PTF 10hgi SLSN-I 249.44601 +6.20898 50 −20.3 0.096 55370 6.06 <5.1 × 10
42
0.1 1.00
PTF 11dij SLSN-I 207.74069 +26.27856 40: −21.1: 0.143 55690 1.21 <4.6 × 10
42
0.06 1.00
PTF 11rks SLSN-I 24.93962 +29.92417 20 −21.0 0.20 55945 5.27 <5.4 × 10
42
0.07 1.00
PS 1-12fo SLSN-I 146.55379 +19.84131 >14 −21.0: 0.175 55956 2.79 <1.8 × 10
43
0.2 1.00
SN 2006jc Ibn 139.36667 +41.90889 <15 −17.8 0.006 54020 1.00 ∼1.5 × 10
41
0.04 0.00
PTF 09drs IIn 226.62567 +60.59427 40: −17.8: 0.045 55210 1.72 <4.4 × 10
42
1.2 1.00
SN 2010jl/PTF 10aaxf IIn 145.72221 +9.49494 15..23 −20.6 0.011 55500 3.05 ∼1.8 × 10
41
0.004 0.00
SN 2010jp/PTF 10aaxi IIn 94.12770 −21.41001 <19 −14.6: 0.01 55520 11.0 <1.2 × 10
40
0.06 0.04
SN 2010jj/PTF 10aazn IIn 31.71774 +44.57156 15..53 −18.0: 0.016 55530 9.38 <1.2 × 10
41
0.03 1.00
SN 2010bq/PTF 10fjh IIn 251.73066 +34.15964 15..45 −18.5 0.032 55310 1.79 <1.2 × 10
42
0.2 1.00
PTF 11iqb IIn 8.52015 −9.70498 10: −18.4 0.013 55780 2.79 ∼7.9 × 10
40
0.01 0.00
SN 2007bb IIn 105.28108 +51.26592 <15 −17.6: 0.021 54192 7.04 <2.8 × 10
41
0.09 0.07
SN 2007pk IIn 22.94613 +33.61503 <14 −17.3: 0.017 54423 4.72 <1 × 10
41
0.04 0.00
SN 2008cg IIn 238.56313 +10.97361 30..60 −19.4: 0.036 54583: 3.65 <2.6 × 10
41
0.02 1.00
SN 2009au IIn 194.94167 −29.60208 ··· −16.5: 0.009 54901: 6.42 <3.5 × 10
40
0.03 0.15
SN 2010al IIn 123.56629 +18.43839 <35 −16.0: 0.0075 55268: 3.92 ∼2.2 × 10
41
0.3 0.00
SN 2011ht IIn 152.04413 +51.84917 50 −16.8 0.004 55880 0.78 ∼7.2 × 10
39
0.05 0.00
SN 2011hw IIn 336.56058 +34.21642 ··· −19.1: 0.023 55883: 10.2 <5 .1 × 10
40
0.004 1.00
PTF 10tel IIn 260.37782 +48.12983 17 −18.5 0.035 55450 2.34 <7.2 × 10
41
0.1 1.00
SN 2011iw IIn 353.70083 −24.75044 <40 −18.1: 0.023 55895: 1.61 <1.0 × 10
41
0.02 1.00
SN 2005db IIn 10.36163 +25.49767 <18 −16.8: 0.0153 53570: 4.17 ∼5.3 × 10
40
0.04 0.00
SN 2005av IIn 311.15658 −68.75294 <19 −17.8: 0.0104 53453: 4.85 <8.1 × 10
39
0.02 1.00
SN 2003lo IIn 54.27133 −5.03814 ··· −15.8: 0.0079 53005: 4.87 <6.3 × 10
39
0.01 1.00
SN 2002fj IIn 130.18792 −4.12736 <90 −18.5 0.0145 52532 3.12 <4.9 × 10
40
0.007 1.00
Notes. The sample of SNe with Swift X-ray observations. Type refer to SN type, α
J2000
and δ
J2000
are the J2000.0 right ascension and declination, respectively. t
rise
is the approximate rise time of the SN optical light curve. The rise time is deduced from various sources including PTF and Katzman Automatic Imaging Telescope
(KAIT; Filippenko et al. 2001) photometry and the literature listed in the references. The colon sign indicates an uncertain value. M
R
is the approximate absolute
R-band magnitude at maximum light (ignoring K-corrections). z refers to the object redshift. If the galaxy is nearby and has a direct distance measurement in the
NASA Extragalactic Database (NED), we replaced the observed redshift by the redshift corresponding to the luminosity distance of the galaxy. t
peak
is the MJD of
maximum light and N
H
is the Galactic neutral hydrogen column density for the source position (Dickey & Lockman 1990). L
X
is the X-ray luminosity or the 2σ
upper limit on the X-ray luminosity in the 0.2–10 keV band. L
X
/L
opt
is the ratio between the X-ray measurements or limit (L
X
) and the peak visible-light luminosity.
Finally, FAP
min
is the minimum false-alarm probability value over all the observations of the source. Sources with FAP
min
< 0.01 indicate a possible detection of an
X-ray source at the position of the SN. As discussed in the text, some of these possible detections are chance coincidences or due to emission from the SN host galaxy.
References. PTF 09atu: Quimby et al. (2011b). PTF 09cnd: Quimby et al. (2011b); Chandra et al. (2009); Chandra et al. (2010). PTF 09cwl: SN 2009jh; Quimby
et al. (2011b); Drake et al. (2009b). SN 2010gx: PTF 10cwr; Mahabal et al. (2010); Quimby et al. (2010a); Pastorello et al. (2010b); Quimby et al. (2011). PTF 10hgi:
Quimby et al. (2010b). PTF 11dij: Drake et al. (2011a); Drake et al. (2011b); Quimby et al. (2011c). PTF 11rks: Quimby et al. (2011a). PS1-12fo: Drake et al. (2012);
Smartt et al. (2012); Maragutti et al. (2012). PTF 09drs: Reported here for the first time. SN 2010jl: PTF 10aaxf; Newton & Puckett (2010); Stoll et al. (2011).
SN 2010jp: PTF 10aaxi; A peculiar SN IIn; Maza et al. (2010); Challis et al. (2010b); Smith et al. (2012). SN 2010jj: PTF 10aazn; Rich (2010b); Silverman et al.
(2010a). SN 2010bq: PTF 10fjh; Duszanowicz (2010); Challis et al. (2010a). PTF 11iqb: Parrent et al. (2011); Horesh et al. (2011). SN 2007bb: Joubert & Li (2007);
Blondin et al. (2007). t
peak
and t
rise
are based on unpublished KAIT photometry. SN 2007pk: A peculiar SN IIn (Parisky & Li 2007). t
peak
and t
rise
are based on
unpublished KAIT photometry. SN 2008cg: Drake et al. (2008); Blondin & Calkins (2008); Filippenko et al. (2008); Spectrum is similar to SN 1997cy (Filippenko
et al. 2008). SN 2009au: Pignata et al. (2009); Stritzinger et al. (2009). SN 2010al: Spectrum is similar to SN 1983K with He ii,Niii,andHi emission lines; Rich
(2010a); Stritzinger et al. (2010); Silverman et al. (2010b). SN 2011ht: Boles et al. (2011); Prieto et al. (2011); Roming et al (2012). SN 2011hw: Dintinjana et al.
(2011). PTF 10tel: Reported here for the first time. SN 2011iw: Mahabal et al. (2011). SN 2005db: Blanc et al. (2005); Monard (2005b);Kieweetal.2012). SN 2005av:
Monard (2005a); Salvo et al. (2005). SN 2003lo: Puckett et al. (2004); Matheson et al. (2004). SN 2002fj: Monard & Africa (2002); Hamuy (2002).
et al. 2012) that may show a dip in the light curve followed
by rebrightening. Moriya & Maeda (2012) interpret the dip
as an increase in the opacity due to ionization of the massive
shell/CSM as it interacts with the ejecta.
Our sample, presented in Table 1, consists of eight SLSN-I
objects, 19 SNe IIn, and a single SN Ibn. We note that the
spectra of SNe having PTF names, as well as some other SNe,
are available online from the WISeREP
18
Web site (Yaron &
Gal-Yam 2012).
18
Weizmann Interactive Supernova (data) REPository;
http://www.weizmann.ac.il/astrophysics/wiserep/.
3. OBSERVATIONS
For each Swift/XRT image of an SN, we extracted the number
of X-ray counts in the 0.2–10 keV band within an aperture of 7.
2
(3 pixels) radius centered on the SN position. We chose a small
aperture in order to minimize any host-galaxy contamination.
We note that this aperture contains ∼37% of the source flux
(Moretti et al. 2004). The background count rates were estimated
in annuli around each SN, with an inner (outer) radius of
50
(100
). For each SN that has Swift/XRT observations, we
searched for Chandra observations. The Chandra observations
were analyzed in a similar manner with an extraction aperture
3
The Astrophysical Journal, 763:42 (11pp), 2013 January 20 Ofek et al.
Tab le 2
Swift/XRT X-Ray Measurements
Name t − t
peak
Exp. CR ΔCR
−
ΔCR
+
CR
UL
FAP L
X
(Start (End
Day) Day) (s) (counts ks
−1
) (counts ks
−1
) (counts ks
−1
) (counts ks
−1
)(ergs
−1
)
PTF 09atu 2.7 ··· 4858.0 −0.02 ··· ··· 0.62 1.00 <1.9 × 10
44
PTF 09cnd −17.6 ··· 3441.8 −0.02 ··· ··· 0.87 1.00 <5.4 × 10
43
−13.7 ··· 3557.2 −0.01 ··· ··· 0.84 1.00 <5.3 × 10
43
−10.5 ··· 3273.1 −0.02 ··· ··· 0.92 1.00 <5.7 × 10
43
−5.9 ··· 3997.8 −0.02 ··· ··· 0.75 1.00 <4.7 × 10
43
−2.2 ··· 2232.1 −0.05 ··· ··· 1.34 1.00 <8.4 × 10
43
4.5 ··· 2980.6 −0.03 ··· ··· 1.01 1.00 <6.3 × 10
43
18.3 ··· 2026.8 −0.01 ··· ··· 1.48 1.00 <9.2 × 10
43
27.9 ··· 1910.6 −0.02 ··· ··· 1.57 1.00 <9.8 × 10
43
−17.6 −2.2 16501.9 −0.02 ··· ··· 0.18 1.00 <1.1 × 10
43
4.527.9 6917.9 −0.02 ··· ··· 0.43 1.00 <2.7 × 10
43
−17.627.9 23419.8 −0.02 ··· ··· 0.13 1.00 <8.0 × 10
42
Notes. Summary of all the 305 Swift/XRT flux measurements of the 28 SNe in our sample. For each SN we list the observation date (t − t
peak
) relative to the t
peak
listed in Table 1. Rows that list both a start day and end day give “superepoch” measurements as described in the main text. Exp. is the exposure time. CR, ΔCR
−
,
ΔCR
+
are the source count rate, lower error, and upper error, respectively. CR
UL
is the 2σ upper limit on the source count rate, FAP is the false-alarm probability
(see the text), and L
X
is the source luminosity, or the 2σ upper limit on the luminosity, in the 0.2–10 keV band. If FAP > 0.01 then we provide the 2σ upper limits,
otherwise the measurements are given along with the uncertainties.
(This table is available in its entirety in a machine-readable form in the online journal. A portion is shown here for guidance regarding its form and content.)
of 2
and background annuli with an inner (outer) radius of
15
(40
). All of the Swift/XRT X-ray measurements are listed
in Table 2 (the full table is available in the online version).
In addition, in Table 2, for each object we give the count rate
in up to four superepochs: (1) all of the observations taken
prior to maximum light, or discovery date if time of maximum
light is not known; (2) all of the observations taken between
maximum light and 300 days after maximum light; (3) all of the
observations taken more than 300 days after maximum light;
and (4) all of the observations at the position of the SN.
In each epoch, and superepoch, we also provide an estimator
for the false-alarm probability (FAP), which is the probability
that the X-ray counts are due to the X-ray background rather than
a source. This probability is estimated as 1 minus the Poisson
cumulative distribution to get a source count rate smaller than
the observed count rate, assuming the expectancy value of the
Poisson distribution equal to the background counts. We note
that the median background counts in our background annuli
is six counts (within an exposure). However, the background
annuli area is about 340 times larger than the photometric
aperture area. Therefore, the contribution of the errors in the
background to the uncertainty in the source counts is negligible.
We note that in some cases X-ray emission from the host galaxy
will tend to produce some seemingly significant, but actually
“false alarm” detections under these assumptions. Furthermore,
many nuclear X-ray sources like active galactic nuclei (AGNs),
as well as X-ray sources in general, tend to significantly vary
with time (e.g., Gonz
´
alez-Mart
´
ın & Vaughan 2012). This fact
makes it very hard to distinguish between X-ray emission from
an SN and its host galaxy. An example for such confusion is
demonstrated in the case of SN 2007pk discussed in this paper.
In cases in which FAP 0.01, we also estimated the 2σ
upper limit on the count rate (Gehrels 1986). The count-rate
measurements or upper limits are converted to luminosity in the
0.2–10 keV band using the PIMMS
19
web tool and assuming
that (1) the aperture in which we extracted the source photometry
19
http://cxc.harvard.edu/toolkit/pimms.jsp
contains 37% of the source photons (Moretti et al. 2004);
(2) Galactic neutral hydrogen column density at the position
of the sources as listed in Table 1 (Dickey & Lockman 1990);
(3) a spectrum of N
ph
(E) ∝ E
−0.2
, where N(E) has units
of photons cm
−2
s
−1
(motivated in Section 4); and (4) the
luminosity distance to each SN calculated using the redshift
listed in Table 1 together with H
0
= 70.4kms
−1
Mpc
−1
,
Ω
m
= 0.268, Ω
Λ
= 0.716, and Ω
K
= 0.986 (the third-year
WMAP+SNLS cosmology; Spergel et al. 2007).
Several objects listed in Table 2 show count rates which may
deviate from zero. Here we discuss the observations of all seven
sources that have FAP 0.01 in at least one of the epochs or
superepochs. We note that Table 2 contains 305 epochs and
superepochs; therefore, we expect about three random false
detections. Interpretation of these observations is discussed in
Section 5.
SN 2010jl/PTF 10aaxf (Figure 1). This SN has a large num-
ber of Swift/XRT observations, as well as Chandra/ACIS-S
observations in five epochs (Chandra et al. 2012b), of
which three are public (PIs: D. Pooley; Tremonti). The
host, SDSSJ094253.43+092941.9, is an irregular star-forming
galaxy. The binned Swift/XRT and Chandra light curves, as
well as the PTF R-band light curve, are presented in Figure 1.
SN 2006jc (Figure 2). This is the only SN in our sample
that belongs to the rare class of Type Ibn SNe. SN 2006jc
has a large number of Swift/XRT and Chandra observations.
The SN is detected on multiple epochs and its X-ray light
curve is presented in Figure 2 (see also Immler et al. 2008).
It was detected in X-rays soon after maximum optical light and
reached a maximum X-ray luminosity of about 1.5×10
40
erg s
−1
at ∼100 days after maximum optical light, 6 times the SN
rise time. SN 2006jc was observed by Chandra on several
occasions. We reduced a 55 ks Chandra observation with
306 photons at the SN location taken 87 days after maximum
light. Given the limited number of photons we did not attempt
to fit complex models. We found that the spectrum is well fitted
by an N
ph
(E) ∝ E
−0.2
power law and with negligible absorbing
column density. The spectra and the best-fit model are presented
4
The Astrophysical Journal, 763:42 (11pp), 2013 January 20 Ofek et al.
0 50 100 150 200
1
2
3
4
5
6
7
8
9
10
11
x 10
−13
Time since maximum light [day]
Flux [erg cm
−2
s
−1
]
Swift
Chandra
Optical/100
0.5
1
1.5
2
2.5
x 10
41
Luminosity [erg s
−1
]
Figure 1. Swift (circles) and Chandra (squares) X-ray light curves extracted at
the position of SN 2010jl. The unabsorbed flux was calculated using PIMMS
assuming a Galactic column density of N
H
= 3.05 × 10
20
cm
−2
,anda
N
ph
(E) ∝ E
−0.2
power-law spectrum. We note that the Chandra observations
show a possible extended source near the SN location. This additional source
may contaminate the Swift/XRT measurements and can explain the small
discrepancy between Chandra and Swift/XRT. Alternatively, the discrepancy
between the Chandra and Swift light curves can be explained if the X-ray
spectrum is harder or N
H
is larger than we assumed. We note that for N
H
which
is a factor of 1000 larger than the Galactic value, the unabsorbed Swift (Chandra)
flux will be about 5.2 (7.2) times larger. For reference, the gray circles show the
PTF R-band luminosity of this SN scaled by 0.01. The PTF R-band luminosity
was calibrated using the method described by Ofek et al. (2012a) and calibration
stars listed by Ofek et al. (2012b).
in Figure 3. Regardless of the exact spectral shape, the spectrum
is hard. Marginalizing over all the free parameters, we find a 2σ
upper limit of N
H
< 1.26 × 10
21
cm
−2
, in excess of the Galactic
column density.
SN 2011ht (Figure 4). This SN took place about 21
from
the center of UGC 5460. It was observed on multiple epochs
using Swift/XRT, and Roming et al. (2012) reported a detection
of an X-ray source at the SN position. The binned Swift/XRT
X-ray light curve of this SN is shown in Figure 4. Apparently
the light curve rises, peaks ∼40 day after maximum optical
light, and then declines. However, the uncertainties in the flux
measurements are large and the light curve is consistent with
being flat (i.e., a best-fit flat model gives χ
2
/dof = 1.23/4).
Moreover, recently Pooley (2012) reported on a 9.8 ks Chandra
observation
20
of this SN. He argued that the emission detected
by Roming et al. (2012), which is shown in Figure 4, is from a
nearby source found 4.
7 from the SN location.
SN 2010al (Figure 5). This SN was found 12
from the center
of the spiral galaxy UGC 4286. It was observed on multiple
epochs using Swift/XRT with a total integration time of 35 ks,
and is detected in the combined image with a mean luminosity
of ∼7 × 10
39
erg s
−1
. Figure 5 presents the binned Swift/XRT
light curve. Although the light curve peaks around 30 days after
maximum light, it is consistent with being flat (i.e., a best-fit flat
model gives χ
2
/dof = 0.15/2).
SN 2005db. This SN was observed on three epochs, about
two years after maximum light (676–695 days), using Swift/
XRT. The combined image, with an exposure time of 13.6 ks,
shows a faint source (five photons) with an FAP of 2 × 10
−4
per
trial. However, given the fact that we have 305 observations,
the FAP over all the trials is only about 0.06. Using the
20
At the time of the writing this paper, this observation was proprietary.
0
0.5
1
1.5
2
2.5
3
x 10
−13
Flux [erg cm
−2
s
−1
]
Swift
Chandra
0
0.5
1
1.5
2
x 10
40
Luminosity [erg s
−1
]
0 100 200 300 400 500
1
3
5
Time since maximum light [day]
<E> [keV]
Figure 2. Upper panel: the Swift (circles) and Chandra (squares) X-ray light
curve extracted at the position of SN 2006jc. The unabsorbed flux was calculated
using PIMMS assuming a Galactic column density of N
H
= 1.0 × 10
20
cm
−2
,
and a N
ph
(E) ∝ E
−0.2
power-law spectrum. Lower panel: the mean photon
energy of the Swift/XRT X-ray observations in the 0.2–10 keV band as a
function of time. A version of this plot is shown by Immler et al. (2008).
Galactic column density (Table 1) and assuming a power-law
spectrum N
ph
(E) ∝ E
−0.2
, the unabsorbed flux is (1.97
+1.7
−0.72
) ×
10
40
erg s
−1
. Chandra observed this target twice at 722.7
and 1051.4 days after maximum light (PI: D. Pooley), with
integrations of 3.0 and 5.0 ks, respectively. Using the same
assumptions as above, we put a 2σ upper limit on the unabsorbed
flux of 1.4 × 10
40
erg s
−1
for both epochs. To conclude, given
the uncertainties, the Chandra upper limits are consistent with
the possible Swift detection. However, given that this source has
a single detection, we cannot firmly conclude that the detection
is real.
PTF 11iqb. This SN has multi-epoch Swift/XRT observations
taken between about −13 and 28 days relative to maximum light.
The SN is detected in a single epoch ∼24 days after maximum
light, with an FAP of 1.4 × 10
−3
per trial. However, given that
we are reporting 305 individual X-ray observations, the FAP
over all the trials is 0.45. The SN was not detected at the last
epoch, 28 days after maximum light.
SN 2007pk (Figure 6). This SN has a large number of
Swift/XRT observations, as well as a Chandra observation (PI:
D. Pooley). Immler et al. (2007) reported a tentative detection
in the images taken between MJD 54417.09 and 54420.04.
The light curve of the source extracted at the SN position
is shown in Figure 6. The light curve shows a brightening,
with a peak around MJD 54461, and a full width at half-
maximum intensity of about 40 days. However, the SN is about
7
from the center of the spiral host galaxy, NGC 579, and the
centroids of the Swift/XRT positions in individual exposures
are clustered around the galaxy nucleus, rather than at the SN
position. We note that emission from the center of NGC 579
is clearly detected in the Chandra observation and that there
is some emission at the source position. However, the latter
may be due to diffuse emission from NGC 579. Therefore,
without conclusive evidence that the emission is from the SN,
here we assume that the observed flare as well as the quiescent
X-ray emission from the position of the source is due to AGN
activity in NGC 579. In Table 1, we adopted an upper limit
on the X-ray luminosity of SN 2007pk, which is based on the
5
