A Relativistic Type Ibc Supernova Without a Detected C-Ray Burst
Total Page:16
File Type:pdf, Size:1020Kb
Vol 463 | 28 January 2010 | doi:10.1038/nature08714 LETTERS A relativistic type Ibc supernova without a detected c-ray burst A. M. Soderberg1, S. Chakraborti2, G. Pignata3, R. A. Chevalier4, P. Chandra5, A. Ray2, M. H. Wieringa6, A. Copete1, V. Chaplin7, V. Connaughton7, S. D. Barthelmy8, M. F. Bietenholz9,10, N. Chugai11, M. D. Stritzinger12,13, M. Hamuy3, C. Fransson14, O. Fox4, E. M. Levesque1,15, J. E. Grindlay1, P. Challis1, R. J. Foley1, R. P. Kirshner1, P. A. Milne16 & M. A. P. Torres1 Long duration c-ray bursts (GRBs) mark1 the explosive death of some GRBs preferentially discovered at larger distances. Further VLA massive stars and are a rare sub-class of type Ibc supernovae. They are observations of SN 2009bb revealed a power-law flux decay, 21.4 13 distinguished by the production of an energetic and collimated relati- Fn,8.46 GHz < t , in line with the radio afterglow evolution seen vistic outflow powered2 by a central engine (an accreting black hole or for the nearest c-ray burst, namely GRB 980425 at a similar distance neutron star). Observationally, this outflow is manifested3 in the pulse of d < 38 Mpc (Fig. 1). of c-rays and a long-lived radio afterglow. Until now, central-engine- driven supernovae have been discovered exclusively through their 1030 c-ray emission, yet it is expected4 that a larger population goes un- GRB 980425 GRB 031203 detected because of limited satellite sensitivity or beaming of the col- SN 2009bb limated emission away from our line of sight. In this framework, the GRB 060218 1029 recovery of undetected GRBs may be possible through radio searches5,6 for type Ibc supernovae with relativistic outflows. Here we report the discovery of luminous radio emission from the seemingly ordinary ) type Ibc SN 2009bb, which requires a substantial relativistic outflow –1 1028 powered by a central engine. A comparison with our radio survey of Hz type Ibc supernovae reveals that the fraction harbouring central –1 engines is low, about one per cent, measured independently from, but consistent with, the inferred7 rate of nearby GRBs. Indepen- 1027 dently, a second mildly relativistic supernova has been reported8. On 2009 March 21.1 UT, the Chilean Automatic Supernova Search Program (CHASE; ref. 9) discovered10 a bright optical transient 1026 through repeated imaging of the nearby spiral galaxy NGC 3278 at a s Radio luminosity (erg distance d < 40 Mpc. The new object was offset 22 arcsec (4.2 kpc) from the centre of the galaxy and located within its star-forming disk. 11 Local Optical spectroscopy obtained on March 28.1 UT revealed that the 1025 transient was a young type Ibc supernova (hereafter SN 2009bb) lack- type Ibc supernovae ing evidence for hydrogen in the explosion debris. On the basis of the previous non-detection of SN 2009bb on March 19.2 UT, we tightly constrain the supernova explosion date to be March 19 6 1 UT 1024 2 3 (Supplementary Information). 11010 10 Time since explosion (days) Using the Very Large Array (VLA) on April 5.2 UT, we discovered a coincident radio counterpart at right ascension a(J2000) 5 10 h 31 min Figure 1 | Radio observations of the nearest massive star explosions. The 33.87 s and declination d(J2000) 5239u 579 30.10 (60.7 arcsec in each 8.46 GHz radio emission from SN 2009bb (red) is more luminous than any of the coordinate) and with a flux density F 5 24.53 6 0.06 mJy, at fre- other142local(d = 200 Mpc)typeIbcsupernovaeobserved(ref.12andreferences n 7,13,16 quency n 5 8.46 GHz. This corresponds to a spectral radio luminosity within) to date on a comparable timescale (Dt = 100 days), and is consistent 28 21 21 with the radio afterglow luminosities of the nearest GRBs discovered through their of Ln < 5 3 10 erg s Hz at Dt < 17 days after explosion, more 5,6,12 c-ray signal within a similar volume (black). Local type Ibc supernovae with well- luminous than any other type Ibc supernova observed on studied radio emission (grey) exhibit lower luminosities and peak at later times, a comparable timescale. Instead, the radio properties of SN 2009bb indicating smaller sizes and lower mean expansion velocities. The radio emission are consistent with the sample of nearby (redshift z = 0.1) GRBs, from most local type Ibc supernovae is below our current detection threshold; we observed to consistently yield7 lower relativistic energies than ‘classic’ include them here as upper limits (3s;greytriangles). 1Harvard-Smithsonian Center for Astrophysics, 60 Garden Street, MS-51, Cambridge, Massachusetts 02138, USA. 2Tata Institute of Fundamental Research, Mumbai 400 005, India. 3Departamento de Astronomi’a, Universidad de Chile, Casilla 36-D, Santiago, Chile. 4University of Virginia, Department of Astronomy, PO Box 400325, Charlottesville, Virginia 22904, USA. 5Royal Military College of Canada, Kingston, Ontario, Canada K7K 7B4. 6Australia Telescope National Facility, CSIRO, Epping 2121, Australia. 7University of Alabama, Huntsville, Alabama 35899, USA. 8NASA Goddard Space Flight Center, Greenbelt, Maryland 20771, USA. 9Department of Physics and Astronomy, York University, Toronto, Ontario, Canada M3J 1P3. 10Hartebeestehoek Radio Observatory, PO Box 443, Krugersdorp, 1740, South Africa. 11Institute of Astronomy, RAS, Pyatnitskaya 48, Moscow 119017, Russia. 12Las Campanas Observatory, Carnegie Observatories, Casilla 601, La Serena, Chile. 13Dark Cosmology Centre, Niels Bohr Institute, University of Copenhagen, Juliane Maries Vej 30, 2100 Copenhagen Ø, Copenhagen, Denmark. 14Department of Astronomy, Stockholm University, AlbaNova, SE-106 91 Stockholm, Sweden. 15Institute for Astronomy, University of Hawaii, 2680 Woodlawn Drive, Honolulu, Hawaii 96822, USA. 16Steward Observatory, University of Arizona, 933 North Cherry Avenue, Tucson, Arizona 85721, USA. 513 ©2010 Macmillan Publishers Limited. All rights reserved LETTERS NATURE | Vol 463 | 28 January 2010 Unlike the optical emission from supernovae, which traces only the 1030 14 slowest explosion debris, radio observations uniquely probe the Nearby GRBs fastest ejecta as the expanding blast wave (velocity, v) shocks and Type Ibc supernova accelerates electrons in amplified magnetic fields. The resulting syn- ) 1029 SN 2009bb chrotron emission is suppressed by self-absorption (‘synchrotron self- –1 Hz absorption’, SSA), producing a low frequency radio turnover that –1 defines the spectral peak frequency, np. Combining our observations from the VLA and the Giant Meterwave Radio Telescope (GMRT), the 1028 radio spectra of SN 2009bb are well described by an SSA model across multiple epochs (Fig. 2). From our earliest spectrum on April 8 UT 27 (Dt < 20 days), we infer np < 6 GHz and a spectral peak luminosity, 10 28 21 21 Ln,p < 3.6 3 10 erg s Hz . Making the conservative assumption that the energy of the radio R/Δt = c emitting material is partitioned equally into accelerating electrons and 26 Peak radio luminosity (erg s Peak radio luminosity (erg 10 amplifying magnetic fields (equipartition), the properties of the SSA t = 0.1c 13,14 R/Δ radio spectrum enable a robust estimate of the blast-wave radius, R/Δt = 0.01c 16 28 9/19 21 R < 2.9 3 10 (Ln,p/10 ) (np/5) cm. (Here Ln,p is in units of 25 21 21 10 erg s Hz , and np is in units of GHz.) Luminous synchrotron 1 10 100 Δt v sources with a low spectral peak frequency thus require larger sizes ( ) ( p/5) (Fig. 3). For SN 2009bb, we infer R < 4.4 3 1016 cm at Dt < 20 days Figure 3 | Radio properties of the nearest massive star explosions directly and thus the mean expansion velocity is R/Dt 5 0.85 6 0.02 c, where c reveal the blast-wave velocities. We compare the peak radio luminosities for is the speed of light. The transverse expansion speed, Cbc 5 R/Dt type Ibc supernovae (red circles) and nearby GRBs (z =0.1; blue squares) as indicates that the blast wave is relativistic, C > 1.3, at this time (here observed at the spectral peak frequency, n (in GHz), and at time Dt (in days). 2 21/2 p the bulk Lorentz factor C 5 (1 2 b ) , with b 5 v/c). This is a lower These observed properties are tightly related14 to the blast-wave radius. The limit on the initial velocity, as the radio evolution indicates that the average velocities are reasonably estimated as R/Dt (dashed grey lines). For blast wave decelerated early on. We further find that the radio emis- type Ibc supernovae we infer typical velocities of R/Dt < 0.1c, while SN sion requires a minimum energy, E 5 (1.3 6 0.1) 3 1049 erg, coupled 2009bb (yellow star) and the nearest GRBs show R/Dt < c. Error bars, 1s. 7,13,15,16 to the relativistic outflow and comparable to the values inferred energetic and relativistic outflow from SN 2009bb was powered by from the radio afterglows of nearby GRBs (Fig. 4). 14,17 another energy reservoir, a central engine. Until now, engine-driven These conclusions are robust; the blast-wave velocity is insensitive supernovae have been discovered exclusively through their c-ray emis- to deviations from equipartition while the relativistic energy can only be 18 sion, making SN 2009bb the first to be identified by its long-wavelength higher . In view of these constraints, we note that shock-acceleration in signal. some type Ibc supernovae may19 couple a minute fraction (= 0.01%) of Motivated by our discovery of an engine-driven relativistic outflow, the total energy, E , to material with a trans-relativistic velocity. tot we searched for a c-ray counterpart in temporal and spatialcoincidence However, this scenario would require an exceedingly high total energy with SN 2009bb. During our bracketed explosion date estimate, the all- for SN 2009bb, E > 1053 erg, a factor of 102 higher than the total explo- tot sky Interplanetary Network (IPN; ref.