Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.stsci.edu/~sholland/papers/mnras_424_1297.pdf Äàòà èçìåíåíèÿ: Tue Aug 28 17:01:09 2012 Äàòà èíäåêñèðîâàíèÿ: Mon Feb 4 21:45:18 2013 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: magellan project

Mon. Not. R. Astron. Soc. 424, 1297­1306 (2012)

doi:10.1111/j.1365-2966.2012.21311.x

Multiwavelength observations of the Type IIb supernova 2009mg
S. R. Oates,1 A. J. Bayless,2 M. D. Stritzinger,3,4,5 T. Prichard,2,6 J. L. Prieto,7 S. Immler,8,9,10 P. J. Brown,11 A. A. Breeveld,1 M. De Pasquale,12 N. P. M. Kuin, M. Hamuy,13 S. T. Holland,14 F. Taddia4 and P. W. A. Roming2,6
1 2

1

Mullard Space Science Laboratory, University College London, Holmbury St. Mary, Dorking, Surrey RH5 6NT Space Science and Engineering, Division 15, Southwest Research Institute, Building 263, 6220 Culebra Road, San Antonio, TX 78238, USA 3 Department of Physics and Astronomy, Aarhus University, Ny Munkegade 120, DK-8000 Aarhus C, Denmark 4 Department of Astronomy, The Oskar Klein Centre, Department of Astronomy, Stockholm University, AlbaNova, 10691 Stockholm, Sweden 5 Carnegie Observatories, Las Campanas Observatory, La Serena, Chile 6 Department of Astronomy and Astrophysics, Pennsylvania State University, 104 Davey Laboratory, University Park, PA 16802, USA 7 Department of Astrophysical Sciences, Princeton University, Peyton Hall, Princeton, NJ 08544, USA 8 Department of Astronomy, University of Maryland, College Park, MD 20742, USA 9 Astrophysics Science Division, Code 660.1, NASA Goddard Space Flight Centre, 8800 Greenbelt Road, Greenbelt, MD 20771, USA 10 Centre for Research and Exploration in Space Science and Technology Code 668.8 8800, Greenbelt Road Goddard Space Flight Centre, Greenbelt, MD 20771, USA 11 Department of Physics & Astronomy, University of Utah, 115 South 1400 East #201, Salt Lake City, UT 84112, USA 12 Department of Physics and Astronomy, University of Nevada, Las Vegas 4505 S. Maryland Parkway, Las Vegas, NV 89154, USA 13 Departamento de Astronom´ Universidad de Chile, Santiago, Chile ia, 14 Space Telescope Science Institute, 3700 San Martin Drive, Baltimore, MD 21218, USA

Accepted 2012 May 15. Received 2012 May 1; in original form 2012 February 17

ABSTRACT

We present Swift Ultra-Violet Optical Telescope and X-Ray Telescope (XRT) observations, and visual wavelength spectroscopy of the Type IIb supernova (SN) 2009mg, discovered in the Sb galaxy ESO 121-G26. The observational properties of SN 2009mg are compared to the prototype Type IIb SNe 1993J and 2008ax, with which we find many similarities. However, minor differences are discernible including SN 2009mg not exhibiting an initial fast decline or u-band upturn as observed in the comparison objects, and its rise to maximum is somewhat slower leading to slightly broader light curves. The late-time temporal index of SN 2009mg, determined from 40 d post-explosion, is consistent with the decay rate of SN 1993J, but inconsistent with the decay of 56 Co. This suggests leakage of -rays out of the ejecta and a stellar mass on the small side of the mass distribution. Our XRT non-detection provides an upper limit on the mass-loss rate of the progenitor of M < 1.5 â 10-5 M yr-1 . Modelling of the SN light curve indicates a kinetic energy of 0.15+0..02 â 1051 erg, an ejecta mass of -0 13 0.56+0..10 M and a 56 Ni mass of 0.10 ± 0.01 M . -0 26 Key words: supernovae: individual: SN 2009mg.

1 I NTR O DUCTION Core collapse supernovae (SNe) are the death throws of massive stars. These stellar explosions are classified into Type II and

This paper includes data gathered with the 6.5-m Magellan telescopes located at Las Campanas Observatory, Chile, and the Gemini Observatory, Cerro Pachon, Chile (Gemini Programmes GS-2009B-Q-9, GS-2009B-Q40 and GS-2009B-Q-67). E-mail: sro@mssl.ucl.ac.uk Hubble, Carnegie-Princeton Fellow.
C 2012 The Authors Monthly Notices of the Royal Astronomical Society

Type Ib/c subtypes depending on their spectroscopic characteristics. The spectra of Type II SNe (SNe II) contain prevalent hydrogen lines, while those of SNe Ib lack hydrogen lines, but display lines of helium. The spectra of SNe Ic, on the other hand, lack both hydrogen and helium lines (see Filippenko 1997, for further details and references). The lack of hydrogen lines in SNe Ib/c is thought to be due to the hydrogen envelope being stripped before the final pre-SNe stage is reached. In the case of SNe Ic, the helium envelope is also thought to be stripped or 56 Ni is not sufficiently mixed into any helium layer and thereby prevents the excitation of helium lines. The progenitors of SNe Ib/c are therefore thought to be stars in which the majority of their hydrogen-rich envelope is stripped via stellar winds and/or transferred to a secondary star via Roche

C

2012 RAS


1298

S. R. Oates et al.
or an unrelated star (Arcavi et al. 2011; Bietenholz et al. 2012; Soderberg et al. 2012). Evidently, further observations are required after SN 2011dh has faded to determine if the candidate star has disappeared or is indeed still present.

lobe overflow (e.g. Wheeler & Levreault 1985; Woosley, Langer & Weaver 1993, 1995). In 1987, hints appeared of a new subclass of SNe II from the spectral evolution of the peculiar SN 1987K (Filippenko 1988). Around maximum light SN 1987K exhibited spectral characteristics reminiscent of a normal SN II event including the presence of a broad H absorption feature. However, nearly five months later, when the object reappeared from behind the sun, its spectrum resembled more closer that of an SN Ib. This peculiar spectral metamorphosis provided a hint that normal SNe II maybe linked to the hydrogen-stripped SN Ib subclass. Similarly, the well-observed SN 1993J exhibited a clear SN II-like spectrum just after explosion, but over several weeks its spectrum evolved to resemble that of a classic SN Ib event with dominant He I features (Filippenko, Matheson & Ho 1993; Swartz et al. 1993). This led to the introduction of the Type IIb subclass (Filippenko 1988). SNe IIb progenitors are able to retain a small amount (0.01 M ) of hydrogen at the time of explosion, and this residual hydrogen shell is manifested in the optical spectrum obtained just after explosion (Nomoto et al. 1993; Podsiadlowski et al. 1993). Recently, it has been proposed that SNe IIb may be further classified into Type eIIb and cIIb subclasses, dividing SNe IIb into extended or compact progenitors, respectively (Chevalier & Soderberg 2010). This subclassification depends on the radius and mass-loss history of the progenitor star, which may be determined from X-ray and radio observations. Examples of SNe eIIb are SN1993J and 2001dg, while examples of SNe cIIb are 1996cb, 2001ig and 2008ax (Chevalier & Soderberg 2010). Over the past 20 years, approximately 721 (Barbon et al. 1999) SNe IIb have been discovered, but only a handful have been well observed including, amongst others, SNe 1987K (Filippenko 1988), 1993J (Schmidt et al. 1993), 1996cb (Qiu et al. 1999), 2003bg (Hamuy et al. 2009), 2008ax (Pastorello et al. 2008; Roming et al. 2009a; Chornock et al. 2011; Taubenberger et al. 2011) and most recently 2011dh (Arcavi et al. 2011; Maund et al. 2011; Soderberg et al. 2012). The exact number of SNe IIb is uncertain due to the evolution of their spectra. If spectral observations of SNe IIb are not obtained early enough, then we may miss SNe II-like features, such as prevalent hydrogen lines, and so misidentify SNe IIb as SNe Ib. Conversely, if late-time spectral observations are not obtained or are too poor for line identification, then the development of SN Ib-like features, such as numerous He I lines, may not be detected and so the SN could be misidentified as an SN II. Currently, little is known about the spread in mass and energy of SNe IIb, the exact manner of mass-loss, nor exactly how they relate SNe II and SNe Ib/c. It is important that an expanded sample of wellobserved SNe IIb is constructed as these objects with their small hydrogen envelope may bridge the gap between normal SNe II and SNe Ib/c. Direct information can be gathered on the progenitors of SN from pre-explosion images, but it is difficult to identify candidate progenitors without high-resolution Hubble Space Telescope images. Candidate progenitor stars have been directly detected in pre-explosion images of the Type IIb SNe 1993J and 2011dh. In the case of SN 1993J, a K-type supergiant in a binary system was determined to be its progenitor (Maund et al. 2004). For SN 2011dh, a mid-Ftype yellow supergiant is observed at the location of the SN (Maund et al. 2011; Van Dyk et al. 2011), but it is currently too soon to determine whether this is the star that exploded, the binary companion,

1.1 Supernova 2009mg In this paper, we present broad-band ultraviolet (UV) and optical light curves and visual-wavelength spectroscopy of the Type IIb SN 2009mg. SN 2009mg was discovered on the 7.9 December 2009 UT (Monard 2009) in the Sb galaxy ESO 121-G26. With J2000 co. ordinates of RA = 6h 21m 44s 86 Dec. = -59 44 26 , SN 2009mg was 46 arcsec east and 4 arcsec south from the centre of the host galaxy. The redshift of the host galaxy is z = 0.0076 ± 0.0001 (Koribalski et al. 2004). The distance to the host galaxy is therefore 32.7 Mpc and the distance modulus is = 33.01 ± 0.48 mag, both of which were retrieved from the NASA/IPAC Extragalactic Database.2 As commonly found for SNe IIb, the classification was continually revised as the object evolved. The initial classification was that of an SN Ia (Prieto 2009); however, soon after this it was changed to a broad-line SN Ic (Roming, Prieto & Milne 2009b). Days later as the spectrum evolved, it was clear that this object was an SN IIb (Stritzinger 2010), as its early spectra strongly resembled that of the prototypical Type IIb SNe 1993J and 1996cb. SN 2009mg was not the first SN to be detected in ESO 121-G26. In the previous year, the Type II SN 2008M was also discovered in this galaxy (Green 2008). The locations of both SNe 2009mg and 2008M are indicated in Fig. 1. The organization of this paper is as follows. In Section 2, we provide the main observations of this SN, and describe the data reduction and analysis methods used. The main results are presented in Section 3, while the discussion and conclusions follow in Section 4 and Section 5, respectively. Unless stated, all uncertainties throughout this paper are quoted at 1 . We have adopted the Hubble parameter H 0 = 70 km s-1 Mpc-1 and density parameters = 0.73 and m = 0.27.

2 O BSER VATIONS 2.1 Swift imaging X-ray and optical/UV observations were performed simultaneously with the X-Ray Telescope (XRT; Burrows et al. 2005) and the Ultra-Violet Optical Telescope (UVOT; Roming et al. 2000, 2004, 2005), the two narrow-field instruments onboard Swift (Gehrels et al. 2004). The analysis of 57.5 ks of X-ray data, see Fig. 2, does not reveal a source at the position of the SN. In Section 3.4, the XRT observations are used to place limits on the mass-loss rate of the progenitor. Optical/UV observations of both SNe 2008M and 2009mg were obtained with Swift. SN 2008M was observed over seven epochs from 28.9 January 2008 UT until 5.3 February 2008 UT. These images were obtained from the Swift archive and used to construct deep template images, which were then used to subtract away galaxy

2

1

http://heasarc.gsfc.nasa.gov/W3Browse/star-catalog/asiagosn.html

http://ned.ipac.caltech.edu/ The NASA/IPAC Extragalactic Database (NED) is operated by the Jet Propulsion Laboratory, California Institute of Technology, under contract with the National Aeronautics and Space Administration.
C 2012 The Authors, MNRAS 424, 1297­1306 Monthly Notices of the Royal Astronomical Society C 2012 RAS


Type IIb SN 2009mg

1299

Figure 1. Two panels displaying the Swift UVOT images of the galaxy ESO 121-G26 and SNe 2009mg (left) and 2008M (right), respectively. The RGB images were constructed from u, b and v images.

Figure 2. The three panels display the Swift XRT images of the galaxy ESO 121-G26 at three different epochs. The first panel consists of data from a 2 d period displaying the X-ray emission from SN 2008M. The middle panel consists of data from a 4 d period, and shows that the X-ray emission from SN 2008M has faded. The final panel shows the position of SN 2008M and SN 2009mg. No X-ray emission is detected from SN 2009mg.

emission at the position of SN 2009mg in each of the science images. Swift UVOT observations of SN 2009mg commenced on 10.9 December 2009 UT and concluded on 14.4 February 2010 UT. 22 epochs of broad-band imaging was performed using three optical (u, b and v ) and three UV (uvw1, uv m2and uvw2) filters. Photometry of SN 2009mg was computed following the method described in Brown et al. (2009). The SN+galaxy photometry was extracted using a 3 arcsec radius source region and a background region positioned away from the galaxy in a source-free location. The galaxy count rate was determined from the summed exposures of SN 2008M, using the same background and source regions. This value was then subtracted from the SN+galaxy photometry determined from the images of SN 2009mg. The SN photometry was then aperture corrected to 5 arcsec in order to be compatible with UVOT calibration. The analysis pipeline used software HEADAS 6.10 and UVOT calibration 20111031. Count rates were converted into magnitudes using the UVOT zero-points presented by Breeveld et al. (2011). The resulting light curves are presented in Fig. 3 and the photometry is provided in Table 1. For some of the UV data points, several epochs have been co-added to provide detections or deep upper limits. 2.2 Optical spectroscopy Nine epochs of low-resolution and one epoch of high-resolution optical spectroscopy were obtained with facilities located at the Las Campanas Observatory and the Gemini-South (GEMINIS) telescope. A journal of the spectroscopic observations is provided in Table 2. Our spectroscopic series covers the flux
C 2012 The Authors, MNRAS 424, 1297­1306 Monthly Notices of the Royal Astronomical Society

evolution from -1.2 to +91.6 d relative to B-band maximum. All spectra were reduced in the standard manner using 3 IRAF scripts, and in most cases, telluric corrections were not performed. 3 A N A LY S I S 3.1 Reddening According to the Schlegel, Finkbeiner & Davis (1998) infrared (IR) dust maps, the Galactic reddening component in the direction of SN 2009mg is E (B - V )MW = 0.045 mag. In order to estimate the colour excess attributed to dust in the host galaxy of SN 2009mg, we turn to two methods. First, from close examination of our high-resolution MIKE spectrum obtained on 24.2 December 2009 UT (551 89.24 MJD) we identify Na I D1 5931.9 and Na I D2 5937.9, shown in Fig. 4, at the redshift of the host galaxy. From this, we measure an equivalent width (EW) for Na I D1 of 0.21 å. Following the Munari & Zwitter (1997) relation between E(B - V ) and EW of Na I D1 implies a host galaxy colour excess of E(B - V )host = 0.07 ± 0.02 mag. We have also estimated the host galaxy colour excess by determining the offset of the b - v colour curve of SN 2009mg with respect to a template colour curve (Drout et al. 2011; Stritzinger et al., in preparation). This template colour curve was constructed from a sample of six unreddened SNe Ib/c and SNe IIb, which were
3 IRAF is distributed by the National Optical Astronomy Observatories, which are operated by the Association of Universities for Research in Astronomy, Inc., under cooperative agreement with the National Science Foundation.

C

2012 RAS


1300

S. R. Oates et al.

Figure 3. Upper panel: observed UV and optical light curves of SN 2009mg. Middle panel: (b - v )and (u - b) colour curves, corrected for reddening. Lower panel: K -correction evolution for the UVOT v filter during the course of UVOT observations.

observed by the Carnegie Supernova Project (Stritzinger et al., in preparation). From this method, we estimate an E(B - V )host = 0.18 ± 0.04 mag. This estimate is about a factor of 2 higher than that obtained from the Na I D1 lines. In what follows, we adopt the weighted mean of our two host colour excess estimates, corresponding to an E(B - V )host = 0.09 ± 0.02 mag. When combined with the Galactic colour excess, we obtain a total colour excess of E(B - V )tot = 0.14 ± 0.02 mag. For an RV = 3.1, this corresponds to an AV = 0.43 mag. 3.2 Photometric evolution 3.2.1 Results Our broad-band photometric observations of SN 2009mg, which began 3 d after discovery, are plotted in Fig. 3, and extend over a duration of 47 d. There are detections in all bands, but there are fewer detections in the three UV bands compared to the three optical bands. The uv m2 band has only one detection from summed observations. The particular brightness of the uvw2 band, relative to the uv m2 band, is likely to be due to red leak in the former (Brown et al. 2010). Comparing the peak colours of this SN, which are derived below, with the colours of a number of sources (Brown et al. 2010, see their table 12), we find most similarity with a 4000 K blackbody source. This blackbody source required red-leak corrections of -2.25 mag for uvw2and -0.55 mag for uvw1. Observations of the SN light curve began 11 d prior to B-band maximum. The rise to maximum is best observed in the ubv -band light curves, which show a tendency for the SN to rise faster and peak earlier in the bluest filters. After reaching maximum, the light curves decay at different rates depending on wavelength with the quickest decays observed in the bluest filters. This is consistent with both the observations of most SNe Ib/c and SNe IIb, including SNe 1993J and 2008ax (Schmidt et al. 1993; Lewis et al. 1994;

Richmond et al. 1994; Pastorello et al. 2008; Roming et al. 2009a; Taubenberger et al. 2011), and with theoretical predictions (Arnett 1982). For each filter, the observed peak magnitude (mpeak ) and peak time (tpeak ), listed in Table 1, were determined using an average of 105 Monte Carlo simulations of a cubic spline fit to the light curves. The errors for these parameters were taken to be the standard deviation of the simulated distributions. The Monte Carlo simulations were performed on each filter using the first 12 observations; upper limits were not used. Since uv m2 has only a single detection, we are not able to use a cubic spline fit to determine tpeak and mpeak , and we therefore do not provide values for these parameters for this filter. To verify our estimates of tpeak and mpeak for the other filters, we repeated the fitting using a simple polynomial and compared the results to those values derived using the Monte Carlo cubic spline fitting. The values of mpeak estimated from the polynomial fits are consistent with cubic spline fits to within 1 for the optical bandpasses and 2 for the UV bandpasses. For values of tpeak , the polynomial fits are consistent with the cubic spline fits, in all filters, to within 1 . The determination of the peak times of the separate bands (see Table 1) confirms our earlier observation that the bluest filters peak the earliest. For each filter, we also calculated the absolute peak magnitude (M peak ) using the observed peak magnitude for each filter and the distance modulus. The resulting values are given in Table 1 and have been corrected for extinction and K -correction. To determine the K -corrections and extinction corrections, we de-reddened and redshifted to the rest frame a template spectrum from a similar Type IIb SN (SN1993J; Jeffery et al. 1994). The template spectrum of the Type IIb SN, 1993J, was used because the spectral range covers the full bandpass of the UVOT (i.e. 1600­8000 å). To the spectral template, we applied the E(B - V ) values corresponding to the Galactic and host extinction using the Milky Way and Small Magellanic Cloud extinction laws (Cardelli, Clayton & Mathis 1989; Pei
C 2012 The Authors, MNRAS 424, 1297­1306 Monthly Notices of the Royal Astronomical Society C 2012 RAS


Type IIb SN 2009mg
Table 1. Observed UVOT magnitudes and 3 upper limits, peak times (tpeak ), peak magnitudes (mpeak ) and absolute peak magnitudes (M peak ). For some of the UV data points, several epochs have been co-added to provide detections and deeper upper limits. The resulting summed exposures are indicated with an in the time column. Absolute peak magnitudes include extinction and K -corrections. Time (MJD-550 00+) 175.92 178.13 179.26 180.40 182.11 184.29 186.47 188.98 189.82 190.66 192.73 193.71 194.68 196.74 197.58 198.41 200.50 201.47 202.44 204.44 205.60 206.76 208.86 209.69 210.53 222.57 226.56 230.80 231.98 238.44 241.32 mpeak tpeak M peak ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 0.03 0.04 1.17 0.04 0.07 2.25 0.07 0.04 0.88 0.04 0.03 1.01 0.03 0.03 0.87 0.04 0.04 1.01 0.04 0.03 1.21 0.07 0.04 0.87 0.04 0.04 0.07 0.11 9.44 0.10 0.10 v 16.67 ± 0.07 16.40 ± 0.06 ­ 16.14 ± 0.06 16.03 ± 0.05 ­ 15.83 ± 0.05 15.76 ± 0.05 ­ 15.76 ± 0.06 15.89 ± 0.05 ­ 15.90 ± 0.05 16.14 ± 0.06 ­ 16.15 ± 0.06 16.55 ± 0.08 ­ 16.67 ± 0.07 16.85 ± 0.07 ­ 17.01 ± 0.07 17.08 ± 0.08 ­ 17.16 ± 0.08 17.62 ± 0.13 17.81 ± 0.13 17.68 ± 0.12 ­ 17.88 ± 0.13 17.85 ± 0.11 15.75 ± 0.05 190.27 ± 1.76 -17.68 ± 0.48 UVOT observed magnitudes u uvw 1 18.17 ± 0.13 17.43 ± 0.08 ­ 17.25 ± 0.08 16.96 ± 0.06 ­ 16.84 ± 0.07 17.02 ± 0.07 ­ 17.12 ± 0.09 17.46 ± 0.08 ­ 17.93 ± 0.10 18.16 ± 0.12 ­ 18.43 ± 0.16 18.75 ± 0.20 ­ 19.79 ± 0.35 19.59 ± 0.30 ­ 19.56 ± 0.26 20.19 ± 0.54 ­ 19.68 ± 0.34 19.76 ± 0.45 19.50 ± 0.31 >19.24 ­ 20.03 ± 0.50 19.63 ± 0.27 16.78 ± 0.07 184.65 ± 0.94 -16.95 ± 0.49 >19.49 19.34 ± 0.22 ­ 18.68 ± 0.15 18.98 ± 0.15 ­ 18.72 ± 0.15 19.15 ± 0.18 ­ 18.84 ± 0.19 19.39 ± 0.21 ­ 19.52 ± 0.24 19.75 ± 0.28 ­ >19.55 ­ 20.61 ± 0.49 ­ ­ 20.59 ± 0.42 ­ ­ >20.46 ­ ­ ­ ­ >20.42 ­ ­ 18.56 ± 0.12 183.82 ± 3.64 -15.30 ± 0.50 uv m2 >20.42 ­ >20.58 ­ ­ 21.52 ± 0.58 ­ ­ >20.84 ­ ­ >20.96 ­ ­ >20.56 ­ ­ >21.18 ­ ­ >21.37 ­ ­ >20.92 ­ ­ ­ ­ >21.52 ­ ­ ­ ­ ­ uvw 2 >22.91 >22.77 ­ 23.18 ± 0.40 22.99 ± 0.27 ­ 22.79 ± 0.27 22.90 ± 0.27 ­ 23.34 ± 0.52 >22.94 ­ 23.46 ± 0.45 23.50 ± 0.54 ­ 23.51 ± 0.56 ­ >23.40 ­ ­ >23.62 ­ ­ >23.50 ­ ­ ­ ­ >23.90 ­ ­ 22.50 ± 0.23 187.47 ± 4.32 -11.37 ± 0.53

1301

b 17.68 ± 0.07 17.08 ± 0.06 ­ 16.86 ± 0.06 16.70 ± 0.05 ­ 16.48 ± 0.05 16.52 ± 0.05 ­ 16.61 ± 0.06 16.68 ± 0.05 ­ 16.84 ± 0.06 17.17 ± 0.07 ­ 17.35 ± 0.07 17.76 ± 0.08 ­ 17.92 ± 0.08 18.10 ± 0.08 ­ 18.36 ± 0.09 18.52 ± 0.11 ­ 18.28 ± 0.09 18.62 ± 0.13 18.73 ± 0.13 18.91 ± 0.14 ­ 19.01 ± 0.16 19.10 ± 0.13 16.46 ± 0.04 187.40 ± 1.11 -17.10 ± 0.48

Table 2. Journal of spectroscopic observations. Date of observation (MJD-55000+) December 2009 December 2009 December 2009 December 2009 January 2010 January 2010 January 2010 January 2010 February 2010 March 2010 186.2 189.3 189.3 193.2 202.3 203.2 204.2 214.1 229.1 279.0 Epocha (d) -1.2 +1.8 +1.8 +5.8 +14.8 +15.8 +16.8 +26.6 +41.6 +91.6 Telescope Instrument Spectral range (å) 3900­9300 3900­9300 3330­9150 3976­6362; 6489­10051 4201­8429 4199­8426 3632­9516 4201­8428 3976­6357; 6485­10053 3700­9500 Resolution (å) 7 7 0.15 9 8 8 7 8 9 7 Exposure time (s) 2 â 300 2 â 300 1800 600 1600 1800 600 1800 600 900

(UT) 21.2 24.3 24.3 28.2 06.3 07.2 08.2 18.1 02.1 24.0
a

CLAY CLAY CLAY BAADE GEMINI-S GEMINI-S CLAY GEMINI-S BAADE DU PONT

LDSS3 LDSS3 MIKE IMACS GMOS GMOS LDSS3 GMOS IMACS WFCCD

Days relative to B-band maximum.

1992), respectively. The extinction corrections were computed via the subtraction of synthetic magnitudes of an unreddened and reddened template spectrum in the observed frame. The K -corrections were computed through the subtraction of synthetic magnitudes from the unreddened template spectrum in the rest and observed frames. The resulting extinction and K -correction factors are proC 2012 The Authors, MNRAS 424, 1297­1306 Monthly Notices of the Royal Astronomical Society

vided in Table 3. Correcting for extinction and K -correction gives a value for the absolute peak v magnitude of SN 2009mg of -17.68 ± 0.48, which is consistent with other SNe IIb (Richardson, Branch & Baron 2006; Drout et al. 2011), including SN 1993J (-17.57 ± 0.24) and SN 2008ax (-17.61 ± 0.43) (Taubenberger et al. 2011).

C

2012 RAS


1302

S. R. Oates et al.
maximum, both colour curves evolve towards the red, reaching 1 mag, and then cease evolving. The spectra for SN 2009mg have a spectral range covering the UVOT v bandpass; we were therefore able to compute optical v -band K -corrections. The results are plotted in the lower panel of Fig. 3 at epochs corresponding to when spectroscopic observations were obtained. The evolution of the K -correction for the v filter tends to follow the (b - v ) colour curve evolution. Interestingly, this behaviour mimics what is observed in thermonuclear SNe (Nugent, Kim & Perlmutter 2002).

3.2.2 Comparison with other SNe IIb The brightening to maximum and the subsequent evolution of SN 2009mg's optical light curves follows the classic behaviour of a normal SN IIb. In the top panel of Fig. 5, we compare the v band light curve of SN 2009mg to those of SNe 1993J (Lewis et al. 1994) and 2008ax (Roming et al. 2009a). Note that no extinction corrections have been applied to the photometry and cosmological corrections are ignored since the redshift of each object is low. Here the light curves of SNe 1993J and 2008ax have been scaled to the peak brightness of SN 2009mg. Since the time post-explosion is uncertain for SN 2009mg, the time for this SN has been scaled to the well-constrained explosion date of SN 2008ax. Scaling the time axis of SN 2009mg to that of SN 2008ax allows us to estimate the explosion epoch to be 22 ± 2 d prior to v -band maximum, which corresponds to 3 December 2009. Fig. 5 reveals that the v -band light curve of each SN is strikingly similar, particularly from maximum light and onwards. Prior to maximum, the v -band light curve of SN 2009mg appears to be slightly broader than the comparison objects. The broader rise for SN 2009mg may be an indication that the ejecta mass of SN 2009mg

Figure 4. The high-resolution CLAY­MIKE spectrum displaying the Na D1 5931.9 and Na I D2 5937.9 lines. Table 3. Total extinction and K -correction factors for the six UVOT bandpasses in magnitudes. UVOT filter Extinction K -correction v -0.42 0.00 b -0.55 -0.01 u -0.67 -0.05 uvw 1 -0.81 -0.04 uv m2 -1.09 -0.04 uvw 2 -0.99 -0.03

I

Plotted in the middle panel of Fig. 3 are the (u - b) and (b - v ) colour curves. From the start of observations, with (u - b) 0.84 mag and (b - v ) 0.37 mag, the colour curves are observed to evolve from the red to the blue until the time of photometric maximum, at which point they reach minimum values of 0.52 and 0.14 mag for (u - b) and (b - v ), respectively. After photometric

Figure 5. Comparison of the v -band light curves of SN 2009mg to those of the prototypical Type IIb SNe 2008ax and 1993J. The top panel shows the v -band light curve of SN 2009mg compared with those of SNe 1993J and 2008ax. As the explosion date of SN 2009mg is unknown, the peak time has been shifted to match that of SN 2008ax. The lower panel shows the (b - v ) colour evolution for each of these objects, corrected for extinction.
C 2012 The Authors, MNRAS 424, 1297­1306 Monthly Notices of the Royal Astronomical Society C 2012 RAS


Type IIb SN 2009mg
is greater than for SNe 1993J and 2008ax, or the ejecta velocity of SN 2009mg is lower than the comparison objects. The largest difference between the three objects is observed at the earliest epochs. In SN 1993J, there is an initial drop in luminosity, which is related to a cooling phase that follows shock breakout, while for SN 2008ax, there is a marginal detection of such a decline in the u band (Roming et al. 2009a). No post-shock breakout cooling phase is documented in the v light curve of SN 2009mg, nor at bluer wavelengths (see Fig. 3). This is a strong indication that the observations of SN 2009mg did not commence early enough to catch the initial upturn in luminosity and may also be an indication that the explosion date for SN 2009mg is earlier than the 22 d estimated from comparison with the light curve of SN 2008ax. From 10 d post-explosion, the light curves of all three objects exhibit a similar overall behaviour, a rise to a peak, followed by an exponential decay, which then flattens to a shallower linear decay from 40 d. After the peak, they all appear to decay at the same rate. Given the comparison objects are nearly identical to SN 2009mg from maximum light onwards, we attempt to quantify their similarities by measuring the m15 diagnostic for the b- and v -band light curves. Here m15 is taken to be the magnitude difference between the peak SN brightness and its brightness measured 15 d post-peak (Phillips 1993). For the v light curves, we obtain m15 (v ) values of 0.948 ± 0.004, 1.19 ± 0.07 and 1.17 ± 0.08 mag for SNe 1993J, 2008ax and 2009mg, respectively. These values of m15 (v ) are consistent with, but towards the steeper end of, the range of values determined for a sample of SNe Ib/c V -band light curves (Drout et al. 2011), where the mean m15 (V ) for their sample is 0.87 ± 0.25 mag. However, m15 (v ) for SN 2008ax is slightly faster than the V -band decline rate found by Taubenberger et al. (2011) and Drout et al. (2011). For the b-band light curve, m15 (b) for the three objects is 1.67 ± 0.01 (SN 1993J), 1.57 ± 0.06 (SN 2008ax) and 1.46 ± 0.09 mag (SN 2009mg), respectively. Note that the value of m15 (b) for SN 2008ax is consistent with the value of m15 (B) obtained by Taubenberger et al. (2011). SN 2009mg declines more slowly in both the v and b bands in comparison to 1993J, while it is more consistent with the evolution of SN 2008ax. The lower panel of Fig. 5 displays the (b - v ) evolution for SNe 1993J, 2008ax and 2009mg. We have applied extinction corrections, with values of Av = 0.94 and Ab = 1.24 for SN 2008ax (Roming et al. 2009a), and AV = 0.58 and AB = 0.78 for SN 1993J (Lewis et al. 1994). Although we have not corrected for any systematic offset between the colour curves of each object due to the different photometric systems, these differences are expected to be minimal with (B - V ) (b - v ) 0.01 mag (Poole et al. 2008). From 20 d post-explosion, the (b - v ) colour curves exhibit a similar evolution for all three SNe IIb. During the initial 20 d post-explosion, the colour evolution of SN 2009mg is most similar to SN 2008ax. Observations of SN 2009mg commenced at 7 d post-explosion at which (b - v ) for SN 2009mg and SN 2008ax is 0.83 and 0.56, respectively. From this point, the colour curves of SNe 2009mg and 2008ax are observed to dip towards the blue, reaching minimum values of (b - v ) of 0.52 and 0.28, respectively. The colour curves then evolve back towards the red, finally levelling off at (b - v ) 1 before observations ceased. The behaviour of the (b - v ) colour curve for SNe 2008ax and 2009mg appear to be a common feature for Type Ib/c and Type IIb SNe (e.g. Roming et al. 2009a; Stritzinger et al. 2009; Drout et al. 2011). To constrain and compare the late-time temporal behaviour, we computed the decay rate during the shallow decline phase, from 40 d, for the u, b and v light curves of SN 2009mg, and for the
C 2012 The Authors, MNRAS 424, 1297­1306 Monthly Notices of the Royal Astronomical Society C

1303

V -band light curve of 1993J. The decay rate was computed using a linear least-squares fit. As the observations of SN 2009mg ceased at 70 d post-explosion, we fit the linear function to the u, b and v light curves from 40 d post-explosion to the end of observations at 70 d. This resulted in decay rates of -0.004 ± 0.011, 0.022 ± 0.004 and 0.025 ± 0.003 mag d-1 for the u, b and v light curves, respectively. The decay rate of SN 1993J in the V band during the same period is 0.0213 ± 0.0002 mag d-1 , which is consistent with the decay rate of the v filter of SN 2009mg. The v decay rate of SN 2009mg is inconsistent, at 3 confidence, with the decay rate of 56 Co to 56 Fe, which is 0.0097 mag d-1 . The u-band light curve of SN 2009mg appears to cease decaying at late times, with the decay rate consistent with being constant. Similar behaviour is also observed in the u-band light curves of the Type Ib/IIb SNe 2007Y and 2008aq, which are observed not to decay at late times, but to increase in brightness between 20 and 90 d post-peak (Stritzinger et al. 2009). The u light curve of SN 2008aq is observed to decrease in brightness again after 90 d. The u-band light curve of SN 1993J was also reported to remain almost constant from 50 to 125 d past maximum (Lewis et al. 1994). 3.3 Optical spectroscopy The spectroscopic sequence of SN 2009mg, shown in Fig. 6, reveals the typical evolution of an SN IIb. At early phases, H and absorption features associated with Fe II and Ca II dominate the spectrum. In addition, during the earliest observed epochs features attributed to He I and Na I are also discernible and, over time, grow in strength.

Figure 6. Spectroscopic sequence of SN 2009mg ranging from -1to +92 d relative to B-band maximum. Each spectrum has been corrected to the rest frame of the host galaxy adopting the redshift z = 0.0076, and for presentation purposes shifted in flux by an arbitrary constant.

2012 RAS


1304

S. R. Oates et al.
luminosity of <5.0 â 1039 erg s-1 for an adopted thermal plasma spectrum with a temperature of kT = 10 keV, a Galactic foreground column density of N H = 3.5 â 1020 cm-2 (Dickey & Lockman 1990) and a distance of 33 Mpc. The upper limit derived for the luminosity is consistent with values determined for other stripped core-collapse SNe including SN 2008ax (Immler et al. 2006, 2007; Roming et al. 2009a). In core-collapse SNe, the likely cause of X-ray emission is the interaction of the SN blastwave with circumstellar material (Chevalier 1982; Immler, Aschenbach & Wang 2001; Immler et al. 2007). We use the X-ray non-detection to place an upper limit on the mass-loss rate of the progenitor. Following the methodology described in Immler et al. (2007), we derive a mass-loss rate of <1.5 â 10-5 M yr-1 (w /10 km s-1 ), assuming the speed of the blastwave as s = 10 000 km s-1 and scaled for a stellar wind speed of w = 10 km s-1 . The velocity of the Fe II 5169 feature is representative of the velocity of the photosphere (Dessart & Hillier 2005). Therefore, we are able to assume a value of 10 000 km s-1 for the blastwave velocity, since the peak velocity of Fe II 5169 feature is approximately 10 000 km s-1 at maximum light. The blastwave velocity is proportional to the mass-loss rate and so assuming a blastwave velocity of 10 000 km s-1 , rather than using an average value from across the duration of the XRT observations, gives us a conservative upper limit to the mass-loss rate. The resulting 3 upper limit lies within the range of the mass-loss rate determined for SN 1993J (10-5 ­10-4 ; Immler et al. 2007), and is also consistent with SN 2008ax ((9 ± 3) â 10-6 ; Roming et al. 2009a) and Type IIP SNe (10-6 ­10-5 ; Chevalier, Fransson & Nymark 2006; Immler et al. 2007), which have also been shown to have low mass-loss rates. 4 D ISCUSSION As we have shown in the previous section, the observation properties of SN 2009mg closely resemble those of other normal SNe IIb. We now endeavour to place limits on the explosion properties of this object. Using the method described by Richardson et al. (2006), we proceed to model the v -band light curve of SN 2009mg to provide reasonable estimates on the kinetic energy (Ek ), ejected mass (Mej ) and ejected nickel mass (MNi ). This simple analytic method uses the model of Arnett (1982) to describe the peak of the light curve while the linear tail is best represented by the model of Jeffery (1999). To constrain these three parameters, a Monte Carlo simulation of the v -band light curve of SN 2009mg was computed with 106 realizations. This analysis is best performed using the UVOIR light curve, but there are no IR observations of this SN to our knowledge. Since the v -band light curve is a reasonable proxy for the bolometric light curve, we use this as a substitute for the UVOIR light curve and apply a Bolometric correction of -1.48 mag, as described in Richardson et al. (2006). In this simulation, the routine switches from the Arnett model to the Jeffery model at day 31 post-explosion. The best 2 fit ( 2 /DOF = 1.1) results in Ek = 0.15+0..02 â 1051 erg, -0 13 Mej = 0.56+0..10 M and MNi = 0.10 ± 0.01 M . The best-fitting -0 26 model is shown in Fig. 8. The resultant parameters are similar to, but lower than, the values calculated by Richardson et al. (2006) for the Type IIb SNe 1996cb and 1993J. Since the Jeffery model assumes that the 56 Ni has decayed and the ejecta is optically thin, it should ideally be fit to data between 100 and 150 d. We therefore verified our result by repeating the simulation, fixing the time at which the simulation transfers from the Arnett to the Jeffery model to be 50 d post-explosion. The resulting values for Ek , Mej and MNi
C 2012 The Authors, MNRAS 424, 1297­1306 Monthly Notices of the Royal Astronomical Society C 2012 RAS

Figure 7. Time evolution of line velocities for H , He I 7065 and Fe II 5169. For comparison, the evolution of same line velocities for SN 1993J is plotted (Pastorello et al. 2008).

As is common for SNe IIb, the strength of the He I lines compared to H increased as the SN evolved. In the case of SN 2009mg, the relative strength does not appear to grow as quickly as in SN 2008ax. For SN 2008ax, H and He I 5876 were of comparable strength 30 d after the explosion, while the H absorption was negligible by +56 d post-explosion (Pastorello et al. 2008). For SN 2009mg, assuming an explosion date of 3 December 2009, the signal-to-noise ratio of these two features is comparable in strength by day 45 post-explosion. Our series of spectra do not allow us to place confirm constraints on the epoch in which H completely disappears; however, H is still strong 60 d after the explosion, and is absent in our next spectrum taken on 110 d post-explosion. In Fig. 7, we plot the time evolution of the blueshifts of the absorption features due to H ,Fe II 5169 and He I 7065. Included in this figure is also the time evolution of these lines determined from the spectroscopic sequence of SN 1993J (Pastorello et al. 2008). These measurements were determined from the minima of the P Cygni features. The evolution of the He I line is similar in both objects, while for H , the velocity is similar at the start of the observations, but unlike SN 1993J, the strength of this feature in SN 2009mg does not drop steeply after day 25, but remains consistently higher. The Fe II line velocity of SN 1993J is found to be higher at the start of observations in comparison to SN 2009mg, but declines at a faster rate. It is interesting to compare Fig. 7 with the equivalent figure for SN 2008ax in Pastorello et al. (2008, see their fig. 6). The evolution of the line velocities of H ,He I 7065 and Fe II 5169 for both SNe 2008ax and SN 2009mg behaves very similarly, although the H line in SN 2008ax remains at a slightly higher velocity at late times compared to SN 2009mg and Fe II 5169 drops off at slightly a faster rate for SN 2008ax. This may suggest that the H envelope mass and the blastwave evolution are similar for the two SNe. 3.4 X-ray observations and mass-loss rate Analysis of the X-ray observations shows no significant detection for SN 2009mg. Using a standard 10-pixel radius (24.5 arcsec) source region and extracting the background from a local, sourcefree region to account for sky background and for diffuse emission from the host galaxy, the 3 upper limit to the XRT net count rate is 5.3 â 10-4 counts s-1 . This corresponds to an unabsorbed (0.2­10 keV band) X-ray flux of <2.5 â 10-14 erg cm-1 s-1 , and a


Type IIb SN 2009mg
A C KNO W LEDGMENTS

1305

We thank the referee for their useful comments. We also thank N. Morrell for performing spectroscopic observations. This research has made use of data obtained from the High Energy Astrophysics Science Archive Research Center (HEASARC) and the Leicester Database and Archive Service (LEDAS), provided by NASA's Goddard Space Flight Center and the Department of Physics and Astronomy, Leicester University, UK, respectively. SRO, AAB and NPMK acknowledge the support of the UK Space Agency. JLP acknowledges support from NASA through Hubble Fellowship Grant HF51261.01-A awarded by STScI, which is operated by AURA, Inc. for NASA, under contract NAS 5-2655. MH acknowledges support by CONICYT through Centro de Astrofisica FONDAP 15010003, Centro BASAL CATA (PFB 06) and by Iniciativa Cientifica Milenio through the Millennium Center for Supernova Science (P10-064-F).
Figure 8. Thebestfittothe v -band light curve following the methodology of Richardson et al. (2006). The dot­dashed line shows the fit using the Arnett model, while the dashed line shows the fit of the Jeffery model. The overall combined best-fitting model is given by the solid line.

REFERENCES
Arcavi I. et al., 2011, ApJ, 742, L18 Arnett W. D., 1982, ApJ, 253, 785 Barbon R., Buond´ V., Cappellaro E., Turatto M., 1999, A&AS, 139, 531 i Bietenholz M. F., Brunthaler A., Soderberg A. M., Krauss M., Bartel N., Chomiuk L., Rupen M. P., 2012, ApJ, 751, 125 Breeveld A. A., Landsman W., Holland S. T., Roming P., Kuin N. P. M., Page M. J., 2011, in McEnery J. E., Racusin J. L., Gehrels N., eds, Proc. AIP Conf. Ser. Vol. 1358, Gamma Ray Bursts 2010. Am. Inst. Phys., New York, p. 373 Brown P. J. et al., 2009, AJ, 137, 4517 Brown P. J. et al., 2010, ApJ, 721, 1608 Burrows D. N. et al., 2005, Space Sci. Rev., 120, 165 Cardelli J. A., Clayton G. C., Mathis J. S., 1989, ApJ, 345, 245 Chevalier R. A., 1982, ApJ, 259, 302 Chevalier R. A., Soderberg A. M., 2010, ApJ, 711, L40 Chevalier R. A., Fransson C., Nymark T. K., 2006, ApJ, 641, 1029 Chornock R. et al., 2011, ApJ, 739, 41 Claeys J. S. W., de Mink S. E., Pols O. R., Eldridge J. J., Baes M., 2011, A&A, 528, A131 Dessart L., Hillier D. J., 2005, A&A, 439, 671 Dickey J. M., Lockman F. J., 1990, ARA&A, 28, 215 Drout M. R. et al., 2011, ApJ, 741, 97 Filippenko A. V., 1988, AJ, 96, 1941 Filippenko A. V., 1997, ARA&A, 35, 309 Filippenko A. V., Matheson T., Ho L. C., 1993, ApJ, 415, L103 Gehrels N. et al., 2004, ApJ, 611, 1005 Green D. W. E., 2008, Cent. Bureau Electron. Telegrams, 1227, 3 Hamuy M. et al., 2009, ApJ, 703, 1612 Immler S., Aschenbach B., Wang Q. D., 2001, ApJ, 561, L107 Immler S. et al., 2006, ApJ, 648, L119 Immler S. et al., 2007, ApJ, 664, 435 Iwamoto K., Young T. R., Nakasato N., Shigeyama T., Nomoto K., Hachisu I., Saio H., 1997, ApJ, 477, 865 Jeffery D. J., 1999, preprint (arXiv:e-prints) Jeffery D. J. et al., 1994, ApJ, 421, L27 Koribalski B. S. et al., 2004, AJ, 128, 16 Lewis J. R. et al., 1994, MNRAS, 266, L27 Maund J. R., Smartt S. J., Kudritzki R. P., Podsiadlowski P., Gilmore G. F., 2004, Nat, 427, 129 Maund J. R. et al., 2011, ApJ, 739, L37 Monard L. A. G., 2009, Cent. Bureau Electron. Telegrams, 2071, 1 Munari U., Zwitter T., 1997, A&A, 318, 269 Nomoto K., Suzuki T., Shigeyama T., Kumagai S., Yamaoka H., Saio H., 1993, Nat, 364, 507 Nugent P., Kim A., Perlmutter S., 2002, PASP, 114, 803 Pastorello A. et al., 2008, MNRAS, 389, 955 Pei Y. C., 1992, ApJ, 395, 130

are consistent with the values determined with the transfer at 31 d post-explosion. The favoured progenitors, of SNe IIb, are massive stars in binary systems (Smith et al. 2011; Claeys et al. 2011, see references within). Models of stripped helium stars in binary systems, are provided by Shigeyama et al. (1990) and Podsiadlowski et al. (1993). Shigeyama et al. (1990) conclude that smaller helium stars undergo more extensive mixing and eject smaller masses than larger helium stars and so form light curves with steeper tails. This is because the ejecta is more transparent when the ejecta mass is small and mixed (Shigeyama et al. 1990; Iwamoto et al. 1997), and thus allows more -rays to escape before they can be thermalized. When -rays are fully trapped the light curves decay slowly at a rate consistent with the decay rate of 56 Co. Therefore, one indication of the degree of mixing in the progenitor is the decay rate of the tail of the light curve. The decay rate of the tail of SN 2009mg in the v band is 0.025 ± 0.003 mag d-1 , which is steeper, at 3 confidence, than the decay rate of 56 Co to 56 Fe, which decays at a rate of 0.0097 mag d-1 (see Fig. 5). This suggests that there is leakage of -rays and hence some degree of mixing of the progenitor (Shigeyama et al. 1990). This would indicate that the progenitor is likely to be on the small side of the mass distribution. This is consistent with the light-curve modelling, which suggests values of Ek , Mej and MNi similar to, but lower than, those of 1993J and 1996cb.

5 C ONCLUSIONS SN 2009mg appears to be a normal SN IIb and exhibits properties similar to other normal, well-observed SNe IIb. Modelling the v -band light curve, we find best-fitting parameters for kinetic energy (Ek ) of 0.15+0..02 â 1051 erg, an ejecta mass (Mej )of0.56+0..10 M -0 13 -0 26 and a nickel mass (MNi ) of 0.10 ± 0.01 M . The decline rate of the light-curve tail starting from 40 d past B-band maximum is inconsistent, at 3 confidence, with the decline rate of 56 Co. This indicates that there is leakage of -rays out of the ejecta and suggests that the progenitor star was on the lower end of the stellar mass distribution.
C 2012 The Authors, MNRAS 424, 1297­1306 Monthly Notices of the Royal Astronomical Society

C

2012 RAS


1306

S. R. Oates et al.
Schlegel D. J., Finkbeiner D. P., Davis M., 1998, ApJ, 500, 525 Schmidt B. P. et al., 1993, Nat, 364, 600 Shigeyama T., Nomoto K., Tsujimoto T., Hashimoto M.-A., 1990, ApJ, 361, L23 Smith N., Li W., Filippenko A. V., Chornock R., 2011, MNRAS, 412, 1522 Soderberg A. M. et al., 2012, ApJ, 752, 78 Stritzinger M., 2010, Cent. Bureau Electron. Telegrams, 2158, 1 Stritzinger M. et al., 2009, ApJ, 696, 713 Swartz D. A., Clocchiatti A., Benjamin R., Lester D. F., Wheeler J. C., 1993, Nat, 365, 232 Taubenberger S. et al., 2011, MNRAS, 413, 2140 Van Dyk S. D. et al., 2011, ApJ, 741, L28 Wheeler J. C., Levreault R., 1985, ApJ, 294, L17 Woosley S. E., Langer N., Weaver T. A., 1993, ApJ, 411, 823 Woosley S. E., Langer N., Weaver T. A., 1995, ApJ, 448, 315

Phillips M. M., 1993, ApJ, 413, L105 Podsiadlowski P., Hsu J. J. L., Joss P. C., Ross R. R., 1993, Nat, 364, 509 Poole T. S. et al., 2008, MNRAS, 383, 627 Prieto J., 2009, Cent. Bureau Electron. Telegrams, 2087, 1 Qiu Y., Li W., Qiao Q., Hu J., 1999, AJ, 117, 736 Richardson D., Branch D., Baron E., 2006, AJ, 131, 2233 Richmond M. W., Treffers R. R., Filippenko A. V., Paik Y., Leibundgut B., Schulman E., Cox C. V., 1994, AJ, 107, 1022 Roming P. W. A. et al., 2000, in Flanagan K. A., Siegmund O. H. W., eds, Proc. SPIE Vol. 4140, X-Ray and gamma-Ray Instrumentation for Astronomy XI. SPIE, Bellingham, p. 76 Roming P. W. A. et al., 2004, in Flanagan K. A., Siegmund O. H. W., eds, Proc. SPIE Vol. 5165, X-Ray and gamma-Ray Instrumentation for Astronomy. SPIE, Bellingham, p. 262 Roming P. W. A. et al., 2005, Space Sci. Rev., 120, 95 Roming P. W. A. et al., 2009a, ApJ, 704, L118 Roming P. W. A., Prieto J., Milne P. A., 2009b, Cent. Bureau Electron. Telegrams, 2093, 1

A This paper has been typeset from a TEX/L TEX file prepared by the author.

C 2012 The Authors, MNRAS 424, 1297­1306 Monthly Notices of the Royal Astronomical Society C 2012 RAS