Discovery of a Cosmological, Relativistic Outburst via its Rapidly Fading Optical Emission
S. Bradley Cenko, S. R. Kulkarni, Assaf Horesh, Alessandra Corsi, Derek B. Fox, John Carpenter, Dale A. Frail, Peter E. Nugent, Daniel A. Perley, D. Gruber, Avishay Gal-Yam, Paul J. Groot, G. Hallinan, Eran O. Ofek, Arne Rau, Chelsea L. MacLeod, Adam A. Miller, Joshua S. Bloom, Alexei V. Filippenko, Mansi M. Kasliwal, Nicholas M. Law, Adam N. Morgan, David Polishook, Dovi Poznanski, Robert M. Quimby, Branimir Sesar, Ken J. Shen, Jeffrey M. Silverman, Assaf Sternberg
aa r X i v : . [ a s t r o - ph . C O ] A p r Discovery of a Cosmological, Relativistic Outburst via its RapidlyFading Optical Emission
S. Bradley Cenko , S. R. Kulkarni , Assaf Horesh , Alessandra Corsi , , Derek B. Fox ,John Carpenter , Dale A. Frail , Peter E. Nugent , Daniel A. Perley , , D. Gruber ,Avishay Gal-Yam , , Paul J. Groot , , G. Hallinan , Eran O. Ofek , Arne Rau , ChelseaL. MacLeod , Adam A. Miller , Joshua S. Bloom , Alexei V. Filippenko , MansiM. Kasliwal , Nicholas M. Law , Adam N. Morgan , David Polishook , DoviPoznanski , Robert M. Quimby , Branimir Sesar , Ken J. Shen , , , JeffreyM. Silverman , , and Assaf Sternberg [email protected] ABSTRACT
We report the discovery by the Palomar Transient Factory (PTF) of the tran-sient source PTF11agg, which is distinguished by three primary characteristics: Department of Astronomy, University of California, Berkeley, CA 94720-3411, USA Cahill Center for Astrophysics, California Institute of Technology, Pasadena, CA 91125, USA LIGO Laboratory, California Institute of Technology, MS 100-36, Pasadena, CA 91125, USA Physics Department, George Washington University, 725 21st St, NW Washington, DC 20052 Department of Astronomy & Astrophysics, Pennsylvania State University, University Park, PA 16802,USA National Radio Astronomy Observatory, P.O. Box O, Socorro, NM 87801, USA Lawrence Berkeley National Laboratory, Berkeley, CA 94720, USA Hubble Fellow Max Planck Institute for Extraterrestrial Physics, Giessenbachstrasse, Postfach 1312, 85748, Garching,Germany Benoziyo Center for Astrophysics, Weizmann Institute of Science, 76100 Rehovot, Israel Kimmel Investigator Department of Astrophysics/IMAPP, Radboud University Nijmegen, 6500 GL, Nijmegen, The Nether-lands Physics Department, United States Naval Academy, 572c Holloway Roadd, Annapolis, MD 21402 Observatories of the Carnegie Institution for Science, 813 Santa Barbara St., Pasadena, CA, 91101, USA Dunlap Institute of Astronomy & Astrophysics, University of Toronto, 50 St. George Street, Toronto,ON M5S 3H4, Canada Department of Earth, Atmospheric, and Planetary Sciences, Massachusetts Institute of Technology,Cambridge, MA 02139, USA School of Physics and Astronomy, Tel-Aviv University, Tel Aviv 69978, Israel Kavli IPMU, University of Tokyo, 5-1-5 Kashiwanoha, Kashiwa City, Chiba 277-8583, Japan Einstein Fellow Department of Astronomy, University of Texas, Austin, TX 78712-0259, USA Minerva Fellow, Max Planck Institute for Astrophysics, Karl Schwarzschild St. 1, 85741 Garching, Ger-many R peak = 18 . R = 4 mag in ∆ t = 2 d) opticaltransient emission; (2) a faint ( R = 26 . ± . g ′ − R = 0 . ± .
29 mag)quiescent optical counterpart; and (3) an associated year-long, scintillating radiotransient. We argue that these observed properties are inconsistent with anyknown class of Galactic transients (flare stars, X-ray binaries, dwarf novae), andinstead suggest a cosmological origin. The detection of incoherent radio emis-sion at such distances implies a large emitting region, from which we infer thepresence of relativistic ejecta. The observed properties are all consistent withthe population of long-duration gamma-ray bursts (GRBs), marking the firsttime such an outburst has been discovered in the distant universe independentof a high-energy trigger. We searched for possible high-energy counterparts toPTF11agg, but found no evidence for associated prompt emission. We thereforeconsider three possible scenarios to account for a GRB-like afterglow withouta high-energy counterpart: an “untriggered” GRB (lack of satellite coverage),an “orphan” afterglow (viewing-angle effects), and a “dirty fireball” (suppressedhigh-energy emission). The observed optical and radio light curves appear incon-sistent with even the most basic predictions for off-axis afterglow models. Thesimplest explanation, then, is that PTF11agg is a normal, on-axis long-durationGRB for which the associated high-energy emission was simply missed. However,we have calculated the likelihood of such a serendipitous discovery by PTF andfind that it is quite small ( ≈ . > Subject headings: stars: flare – stars: gamma-ray burst: general – stars: super-novae
1. Introduction
From accreting stellar-mass black holes in our Galaxy to distant active galactic nuclei(AGNs) and gamma-ray bursts (GRBs), outflow velocities approaching the speed of light arecommon in nature. Indeed, the number of known sources capable of generating relativisticejecta has expanded in recent years to include a core-collapse supernova without an accom-panying GRB (SN 2009bb; Soderberg et al. 2010), as well as the presumed tidal disruptionof a star by a supermassive black hole (Levan et al. 2011; Bloom et al. 2011; Zauderer et al. 4 –2011; Burrows et al. 2011; Cenko et al. 2012). With revolutionary new time-domain facili-ties slated to come online in the coming decade, even more exotic examples will surely beuncovered.Time-variable high-energy emission (X-rays and γ -rays) tends to be the hallmark ofsuch relativistic outflows. Yet there is good reason to expect that some relativistic outburstsmay lack a detectable high-energy signature. In the case of GRBs, for example, the mostmundane possibility is a lack of sky coverage: the most sensitive high-energy GRB detectorscover only a fraction of the sky at any given time. But other, more interesting possibilitiesexist, including viewing-angle effects (Rhoads 1997; Perna & Loeb 1998; Nakar et al. 2002)and some physical process suppressing the high-energy emission entirely (Dermer et al. 2000;Huang et al. 2002; Rhoads 2003). The search at longer wavelengths for these “orphan” (i.e.,off-axis) afterglows or “dirty fireballs” has remained one of the most sought-after goals inthe GRB field for more than a decade.In this work, we report the discovery by the Palomar Transient Factory of PTF11agg, arapidly fading optical transient associated with a year-long, scintillating radio counterpart.The detection of a faint, blue, quiescent optical source at the transient location suggestsa cosmological origin for the transient (i.e., well beyond the Milky Way and any nearbygalaxies). At such distances, the observed radio emission requires the presence of relativisticejecta.Throughout this work, we adopt a standard ΛCDM cosmology with H = 71 km s − Mpc − , Ω m = 0 .
27, and Ω Λ = 1 − Ω m = 0 .
73 (Spergel et al. 2007). All quoted uncertaintiesare 1 σ (68%) confidence intervals unless otherwise noted, and UT times are used throughout.Reported optical magnitudes are in the AB system (Oke & Gunn 1983). We have correctedthe reported optical and near-infrared (NIR) photometry for a foreground Galactic extinctionof E ( B − V ) = 0 .
044 mag (Schlegel et al. 1998), using the extinction law from Cardelli et al.(1989).
2. Discovery and Basic Analysis2.1. Optical/Near-Infrared
Regular monitoring observations of field 100033 (centered at α = 08 h m . s , δ =+21 ◦ ′ . ′′
5, with a total on-sky area of 7.2 deg ) were obtained with the Palomar 48 inchOschin telescope (P48) equipped with the refurbished CFHT12k camera (Rahmer et al. 5 – P48 R−band2011 Jan 30.22 Keck/LRIS g−band2011 Sep 26.60 5"Keck/LRIS g−band2011 Sep 26.6030"
Fig. 1.—
Optical imaging of the field of PTF11agg. The P48 discovery ( R -band) image is shown inthe left panel. Follow-up Keck/LRIS g -band observations, obtained on 2011 September 26, are displayedin the center (wider field) and right (zoomed in) panels. The location of PTF11agg, as determined fromour P48 imaging, is indicated with a solid circle (1 ′′ radius; note that this is significantly larger than theastrometric uncertainty in our alignment between the Keck/LRIS and P48 images, which is ∼
50 mas in eachcoordinate). A faint, unresolved source consistent with the location of PTF11agg is detected in both our g -band and R -band (not shown) images. All images are oriented with North facing up and East to the left. R -band filter, which is similar to the r ′ filter from the Sloan Digital Sky Survey(SDSS; Aihara et al. 2011), but offset by ∼
27 ˚A redward (Ofek et al. 2012).In an image beginning at 5:17:11 on 2011 January 30, we detected a bright but short-lived optical flare at the (J2000.0) location α = 08 h m . s , δ = +21 ◦ ′ . ′′
26, witha 1 σ astrometric uncertainty of 70 mas in each coordinate (Figure 1). This source wassubsequently dubbed PTF11agg by our automated discovery and classification pipeline(Bloom et al. 2012). Our P48 photometry of PTF11agg, calculated with respect to nearbypoint sources from SDSS, is presented in Table 1.The peak observed magnitude, obtained in our first image of the field on 2011 January30, was measured to be R = 18 . ± .
05 mag. In the next ten P48 images of the field, allobtained on 2011 January 30, the source is seen to decay by 1.2 mag in the R band. A faintdetection is also obtained by coadding all P48 images from 2011 February 1 ( R = 22 . ± . R -band light curve is plotted in Figure 2. All subsequent P48 imagesresult in nondetections at this location.Examining our pre-outburst (i.e., before 2011 January 30) P48 imaging, we find noevidence for emission at this location in any individual frames (extending back in time to2009 November). The typical limiting magnitude for an individual P48 image is R &
20 mag.Stacking all frames from 2011 January 29 (i.e., the day preceding discovery), we limit theoptical emission at the location of PTF11agg to
R > . -2 -1 Time Since Outburst (d) R - b a n d M a g n i t u d e PTF11agg < 1234> 5 R e d s h i f t Fig. 2.—
Optical light curve of PTF11agg, compared with a representative sample of afterglows of long-duration GRBs discovered by the
Swift satellite (Cenko et al. 2009). The
Swift
GRBs are color-coded byredshift; small black points indicate GRBs with unknown distance. The observed power-law decline fromPTF11agg ( α = 1 .
66) is consistent with GRB afterglow observations at ∆ t ≈ t ≈ . ∼ Swift events have a comparable R -band magnitude at ∆ t ≈ σ upper limits. pre-outburst P48 images results in a nondetection with R > . t > g ′ - and R -band filters), and the Inamori-Magellan ArealCamera and Spectrograph (IMACS; Dressler et al. 2011) mounted on the 6.5 m Magellan-Baade telescope at Las Campanas Observatory ( I -band filter).In our deepest epoch of post-outburst optical imaging (2011 September 26 with Keck/LRIS, 7 –or ∆ t = 240 d), we identify a faint, unresolved (in 0 . ′′ g ′ and R at (J2000.0)coordinates α = 08 h m . s , δ = +21 ◦ ′ . ′′
26 (Figure 1). Given the uncertainty inthe astrometric tie between the Keck/LRIS and P48 imaging (50 mas in each coordinate),the observed 90 mas radial offset is not statistically significant (null probability of 0.17).Coadding Keck/LRIS images of the field of PTF11agg from several individual nights withless ideal conditions (2011 March 4, March 12, and April 27), we can recover an object at thislocation with similar brightness in both g ′ and R . No emission is detected at this locationin the I -band IMACS images to I > . t = 30 d). A total exposure time of 2246 s was obtained simultaneously in the J , H , and K s filters. Raw data files were processed using standard NIR reduction methodsvia PAIRITEL Pipeline III (C. Klein et al. , in preparation), and resampled using SWarp(Bertin et al. 2002) to create 1 . ′′ − images for final photometry.We also observed the location of PTF11agg with the Wide-Field Infrared Camera(WIRC; Wilson et al. 2003) mounted on the 5 m Hale telescope at Palomar Observatory.Images were obtained in the K s filter on 2012 March 28 (∆ t = 423 d) for a total exposuretime of 1200 s. The individual frames were reduced using a custom pipeline within the IRAFenvironment (Tody 1986). Both the PAIRITEL and WIRC images were calibrated withrespect to bright field stars from the Two Micron All-Sky Survey (2MASS; Skrutskie et al.2006).No emission was detected at the location of PTF11agg in any of the NIR bandpasses.The most constraining limits come from the WIRC observations ( K s > . We fit the observed P48 detections on 2011 January 30 and February 1 to a power-lawmodel of the form f ν = f ( t − t ) − α , where f ν is the flux density (in µ Jy), t is the time ofthe outburst onset, α is the power-law index, and f is the flux density at a fiducial time( t + 1 s). We find best-fit values of α = 1 . ± .
35 and t = 23:34 UT ( ± . t occurs 16.6 hr after the precedingP48 nondetection on 2011 January 29 (Table 1). 8 –In the event that PTF11agg is a bona fide GRB-like afterglow ( § R = 18 .
26 at discovery implies an ageof ∆ t . . R ≈
18 approximately 12 hours after the onset of the high-energyemission. Together with the P48 nondetection on 2011 January 29.31, we can conservativelyconstrain the outburst onset to fall within the window from ∼ χ = 8 . t , the true power-law index will be smaller thanwhat we have inferred, and more consistent with most previously observed GRB optical af-terglows. If the outburst actually occurred earlier, the decay index would steepen somewhat.But temporal indices α & . Here we wish to estimate P chance , the a posteriori likelihood that the coincident quiescentcounterpart detected at late times in our Keck/LRIS imaging is unrelated to PTF11agg(i.e., the transient source). We have measured the areal surface density of objects of thisbrightness in our imaging of field 100033, finding σ ( R ≤ .
2) = 0 .
03 galaxies arcsec − ; wenote that this is consistent with the results from Hogg et al. 1997 using entirely differentfields. Using 150 mas, or three times the uncertainty in the astrometric tie between the P48and Keck/LRIS images, as our search radius, and following Bloom et al. (2002), we find that P chance = 2 × − . We therefore consider it highly likely that this source is the quiescentcounterpart of PTF11agg; however, we consider alternative possibilities below as well. 9 – We began radio observations of the field of PTF11agg with the National Radio Astron-omy Observatory’s (NRAO ) Karl G. Jansky Very Large Array (VLA; Perley et al. 2011) on2011 March 11 (∆ t = 40 d). The array was in the “B” configuration until 2011 May 6, thenthe “BnA” configuration until 2011 June 1, and the “A” configuration thereafter. Over thecourse of our monitoring, the angular resolution ranged from 0 . ′′ . ′′
2. The VLA data werereduced with the Astronomical Image Processing System (AIPS) . For flux calibration, weused the source 3C 147, while phase calibration was performed using the objects J0823+2223and J0832+1832. As a check of our flux calibration, we have verified that flux measurementsof our phase calibration sources remain stable throughout the course of our observations.We observed PTF11agg at high frequencies (mm wavelengths) with the Combined Arrayfor Research in Millimeter-wave Astronomy (CARMA) beginning on 2011 March 14 (∆ t =43 d) and continuing for approximately one month. For our CARMA observations, thearray was in the “D” configuration, and the beam had an angular diameter of 10 ′′ . Thetotal bandwidth (lower sideband and upper sideband) was 8 GHz, and the local oscillatorfrequency was 93.6 GHz. The optical depth at high (230 GHz) frequency ranged from fair( τ ≈ .
4; phase noise ≈ ◦ ) on March 14 and April 11, to good ( τ ≈ .
1; phase noise ≈ ◦ )on April 7. Data were reduced using standard techniques within the MIRIAD environment(Sault et al. 1995).A transient radio counterpart was detected with both facilities. The radio counterpartwas unresolved (smallest beam size of 190 mas) and consistent with zero circular polarization( q .
10 %) at all epochs. The results of our EVLA and CARMA monitoring are displayedin Table 2, while the 8 GHz light curve is plotted in Figure 3.
To calculate the radio spectral energy distribution (SED), we must interpolate the var-ious observing frequencies to a common epoch. To provide the longest lever arm, we per-form this analysis at the two epochs of our 93 GHz CARMA detections: 2011 March 14.05 The National Radio Astronomy Observatory is a facility of the National Science Foundation (NSF)operated under cooperative agreement by Associated Universities, Inc.
10 – Time Since Explosion (d) G H z Sp e c t r a l L u m i n o s i t y ( e r g s − H z − ) Mean GRB Radio AfterglowPTF11agg at z =1.5
Fig. 3.—
The 8 GHz radio light curve of PTF11agg, at an assumed redshift of 1.5 (in the middle of ourallowed range: 0 . . z . . § z = 0 .
5, the 8 GHz spectral luminositywould be a factor of 15 smaller, while at z = 3 . σ upper limit. (∆ t ≈
43 d) and 2011 April 7.03 (∆ t ≈
67 d). We have linearly interpolated flux-densitymeasurements made immediately before and after these epochs at frequencies of 5 and 8 GHz.Due to the relatively sparse coverage at 22 GHz, we have simply adopted the flux densityat the closest epoch in time (note that for 2011 March 18 we averaged the two 22 GHzmeasurements obtained on this day). The resulting SEDs are plotted in Figure 4.We fit a power law of the form f ν = f ν β to the data, where f ν is the flux density (in µ Jy), ν is the observing frequency (in GHz), β is the power-law spectral index, and f is 11 – Observed Frequency (GHz) F l u x D e n s i t y ( µ J y ) f ν ∝ ν Fig. 4.—
PTF11agg spectral energy distribution at radio frequencies. The observations at lower frequen-cies have been interpolated to common epochs (∆ t ≈
43 and 67 d) to match the times of our CARMAobservations. the flux density at a fiducial frequency of 1 GHz. For the first epoch (∆ t ≈
43 d), we find β = 0 . ± .
08. On the second epoch (∆ t ≈
67 d), we measure β = 0 . ± .
07. Giventhe relatively large degree of variability (see below), together with the sparse coverage athigh frequencies, we adopt β = 1 / The presence of nonthermal radio emission provides two powerful and independentmeans to constrain the angular size of the emitting region. To begin with, the brightness 12 –temperature ( T B ) of an incoherent radio emitter cannot exceed its equipartition value of T B, eq ≈ K (Readhead 1994; Kulkarni et al. 1998). The brightness temperature is givenby T B = c k B ν f ν π Θ , (1)where c is the speed of light, k B is Boltzmann’s constant, ν is the observing frequency, f ν isthe observed flux density, and Θ is the angular diameter of the emitting region. Adopting T B . K in Equation 1 thus implies a lower limit on the angular diameter of the source:Θ & . (cid:18) f ν µ Jy (cid:19) / (cid:16) ν GHz (cid:17) − µ as . (2)As can be seen from Equation 2, the strictest lower limits on the size of the emittingregion are derived from observations at the lowest frequencies (assuming a power-law spectralindex β < > µ as. Most ofour early observations at 5 and 8 GHz yield comparable (though slightly less strict) limits.Separately, we can constrain the angular size of the source from the detection of inter-stellar scattering and scintillation (ISS; Rickett 1990). To quantify the degree of variationinduced by the scattering electrons, we calculate the modulation index, m p ( ν ) = p V ( f ν ) − h σ ih f ν i , (3)where V ( f ν ) is the variance of the flux density (with respect to an assumed model), h σ i is the average of the square of the individual measurement uncertainties, and h f ν i is theaverage of the flux density.We calculated the modulation indices at 5 and 8 GHz, neglecting higher frequencies dueto the relative lack of observations. We fit the light curves at both frequencies to a power-lawmodel of the form f ν = f ( t − t ) − α , finding best-fit temporal indices of α = − . ± . α = 0 . ± .
06. This power-law modelthen forms the reference which we use to calculate the variance at each frequency. In thismanner, we find m p (5 GHz) = 0 .
42 and m p (8 GHz) = 0 . ν , the transition frequency between the strongand weak scattering regimes. For the line of sight to PTF11agg (Galactic coordinates l =202 . ◦ , b = 29 . ◦ ), we find ν = 11 GHz. For a point source, the maximum degree ofmodulation ( m p = 1) will occur at this transition frequency. It is therefore not unreasonableto expect our observations at 5, 8, and (possibly) 22 GHz to suffer from some degree of ISS. 13 –For ν = 11 GHz, our observations at 5 and 8 GHz will be in the strong scatteringregime ( ν < ν ). Furthermore, given the relatively broad bandwidth of our observations(∆ ν/ν ≈ . m p ( ν ) = (cid:18) νν (cid:19) / . (4)For the line of sight to PTF11agg, we therefore expect a significant degree of modulation fora point source at our observing frequencies: m p (5 GHz) = 0 . m p (8 GHz) = 0 . r / Θ) / ,where Θ r is the size of the Fresnel scattering disk (Walker 1998),Θ r = 8 √ Dν (cid:16) ν ν (cid:17) / µ as , (5)where D is the effective distance to the scattering screen ( D = 0 .
78 kpc for the line of sightto PTF11agg). If we solve for the angular diameter corresponding to the observed degreeof modulation at each frequency, we find Θ(8 GHz) = 10 µ as and Θ(5 GHz) = 34 µ as. Wetherefore conclude that the angular size of the emitting region at ∆ t ≈
100 d is Θ ≈ µ as. γ -ray Limits At the time of discovery, three primary high-energy facilities were monitoring the skyto search for the prompt emission from GRBs. The Third InterPlanetary Network (IPN;Hurley et al. 2010) is a group of nine satellites sensitive to high-energy emission. Whenmultiple satellites detect a GRB, the sky localization can be reconstructed from light traveltime constraints. The IPN provides essentially continuous all-sky coverage (i.e., 100% dutycycle), with a sensitivity to fluences (10 keV – 5 MeV) of S γ & × − erg cm − (at 50%efficiency; i.e., half of the GRBs with this fluence are too faint to trigger the IPN detectors).In addition to the IPN, the Gamma-Ray Burst Monitor (GBM; Meegan et al. 2009) on the Fermi satellite, and the Burst Alert Telescope (BAT; Barthelmy et al. 2005) on the
Swift satellite (Gehrels et al. 2004), also regularly discover a large number of GRBs. The GBMdetects bursts down to a 8 keV – 1 MeV fluence of S γ & × − erg cm − , but has a fieldof view of 8.8 sr (the area of the sky unocculted by the Earth in the Fermi orbit) and aduty cycle of & Swift
BAT has detected events with 15–150 keV fluencesas low as 6 × − erg cm − , but only observes a field of view of 2 sr with a duty cycle of 14 – ∼ § ∼
12 hr window. We further conducted a search for untriggered eventsin the GBM data in the energy range 10–300 keV on several different time scales (0.256 s,0.512 s, 1.024 s, 2.048 s, 4.096 s, and 8.192 s) . No potential high-energy counterparts toPTF11agg were found.Given the field of view and duty cycle of the GBM and BAT, there is a significant likeli-hood that events below the IPN sensitivity threshold would be missed by both instruments.For example, for a GRB with fluence above the GBM sensitivity level (but below the IPNthreshold), the probability of a nondetection from both instruments is as high as ∼ Swift and
Fermi orbits), we consider a fluence of S γ . − erg cm − (i.e., twice the all-sky IPN sensitivity) areasonable limit on any high-energy prompt emission associated with PTF11agg. Given theextremely weak correlation between prompt γ -ray fluence and optical afterglow brightness(Nysewander et al. 2009), this limit is consistent with the known properties of GRBs andtheir afterglows. To search for an X-ray counterpart, we obtained observations of the location of PTF11aggwith the X-ray Telescope (XRT; Burrows et al. 2005) on board the
Swift satellite on 2011March 13 (∆ t = 42 d). Data were reduced using the automated pipeline described byButler & Kocevski (2007). No X-ray source is detected at the location of PTF11agg at thistime. Assuming a power-law spectrum with a photon index of Γ = 2, we derive a 3 σ upperlimit on the 0.3–10 keV flux of f X < × − erg cm − s − .Finally, we note that no historical X-ray emission has been reported at this location,neither in the ROSAT
All-Sky Survey (0.1–2.4 keV; Voges et al. 1999) nor in any compilations Two individual GBM detectors with a significance of 4.0 σ and 3.8 σ above background were required fora trigger to register in this search.
15 –of known Galactic X-ray sources (accessed via the HEASARC and SIMBAD databases).
3. Comparison with Known Galactic Transients
The combination of (1) a rapidly fading optical transient (∆ R & t = 2 d) and(2) a faint, blue ( g ′ − R = 0 . ± .
29 mag), quiescent optical counterpart makes PTF11aggunique amongst the thousands of discoveries by PTF to date. Together with (3) the long-lived (∆ t ≈
300 d) radio emission, here we attempt to simultaneously account for these threedistinguishing characteristics.In order to understand the nature of the emission from PTF11agg, we must constrainits distance. In this section, we first consider a Galactic origin by comparing PTF11agg withknown classes of Galactic transients.Assuming the faint optical source is associated with PTF11agg (i.e., the quiescent coun-terpart), the measured color, g ′ − R = 0 . ± .
29 mag, implies a spectral type of ∼ F2( T eff ≈ σ limit on thecolor ( g ′ − R < .
04 mag), we can rule out single main-sequence stars with T eff . d &
90 kpc. This firmly rules out an association withthe Praesepe cluster ( d ≈
175 pc); in fact, only 4 globular clusters are known to exist atsuch large distances in the extreme outer halo of the Milky Way (e.g., AM 1 at d ≈
120 kpc;Madore & Arp 1979). In addition to the extremely small source densities this far in the halo,the inferred lower limit on the radio luminosity at such a distance ( νL ν & erg s − ) isthree orders of magnitude larger than the most luminous known stellar radio sources (e.g.,RS CVn binaries, FK Com class stars, and Algol-class stars; G¨udel 2002).While a posteriori unlikely, it is nonetheless important to consider that the quiescentoptical source may be unrelated to PTF11agg. Absent color information, an optical nonde-tection, even at the depth of our late-time imaging, is not sufficient to rule out a Galacticorigin. With their smaller effective temperatures, low-mass stars and brown dwarfs (in partic-ular ultracool stars, with spectral type later than M7) emit little flux in the optical bandpass.Furthermore, ultracool stars are known to exhibit high-amplitude, short timescale (minutesto hours) optical and radio outbursts that have in the past been mistaken for extragalac-tic transients (Becker et al. 2004; Kulkarni & Rau 2006; Mahabal et al. 2012; Berger et al. See http://heasarc.gsfc.nasa.gov . See http://simbad.u-strasbg.fr/simbad .
16 –2012a).We can use our NIR limits on the quiescent emission at the location of PTF11aggto calculate the minimum distance to an ultracool star as a function of spectral type; inother words, for each spectral type, any object closer than this “detectability” distancewould be identified in our NIR imaging. For spectral types later than ∼ M4, the strongestconstraint is provided by our deepest epoch of K s -band imaging: K s > . K s -band magnitudes and distance (parallax) measurements for ultracool stars fromDahn et al. (2002) and Patten et al. (2006), we fit a low-order polynomial to calculate theabsolute K s -band magnitude as a function of spectral type, M K s ( ST ), where ST = 5 for M5, ST = 12 for L2, etc. We find the scatter about our derived absolute K s -band magnitude fitis ∼ .
30 mag (i.e., 30%) over the range M5–T8. We then convert the observed peak radioflux density ( f ν, peak ≈ µ Jy) to a lower limit on the radio luminosity ( νL ν ) using thesedistance constraints.The resulting luminosity limits, as a function of spectral type, are plotted in Figure 5.For comparison, we have also plotted all radio observations of ultracool stars from the lit-erature (see figure caption for references). Our luminosity limits are typically at least twoorders of magnitude larger than the most luminous known ultracool stellar flares. Evencomparing with the recently detected flare from the T6.5 dwarf 2MASS J1047+21, by farthe coolest brown dwarf detected at radio frequencies (Route & Wolszczan 2012), our limitsrequire a radio luminosity a factor of >
20 times larger. We furthermore see no evidence fora high degree of circular polarization (common to many, though not all, flares; Berger 2006;Hallinan et al. 2007), and the radio emission from PTF11agg is much more long-lived thanthese low-mass stellar outbursts (durations typically of only hours).In addition to stellar flares, binary systems where one member is a compact object(white dwarf, neutron star, or black hole) are known sources of optical and radio outburstsin the Milky Way. Such a system could circumvent two issues with Galactic transientswe previously identified. First, the energy release during the accretion process is more thansufficient to power the observed radio flux; Cyg X-3 (Geldzahler et al. 1983), for example, hasreached peak radio luminosities in excess of 10 erg s − . Second, the presence of an accretiondisc can alter the optical color of such systems. Accordingly, our previous inference that thequiescent counterpart must lie at d &
90 kpc would no longer be valid.We consider first X-ray binaries, where the degenerate primary is a neutron star or blackhole. Of particular interest are the subclass of microquasars (Mirabel & Rodr´ıguez 1999),whose powerful radio jets exhibit apparent superluminal motion (and thus imply a relativisticoutflow). Due to the lack of a bright quiescent optical counterpart, we consider only low-mass systems where the accretion occurs via Roche-lobe overflow from the nondegenerate 17 –
M4 M6 M8 L0 L2 L4 L6 L8 T0 T2 T4 T6 T8 Y0
Spectral Type R a d i o L u m i n o s i t y ( e r g s − ) P T F a gg Upper LimitsQuiescent EmissionFlaring Emission
Fig. 5.—
Lower limits on the radio luminosity of PTF11agg (solid black line). For each spectral type, wecalculate a minimum “detectability” distance using limits from our NIR imaging (i.e., any source more nearbywould have been detected). We then convert this distance to a lower limit on the radio luminosity basedon the observed peak flux from PTF11agg. Shown for comparison are radio observations of ultracool starsfrom the literature (Berger et al. 2010; Berger 2006; Antonova et al. 2007; Audard et al. 2007; Berger 2002;Berger et al. 2001, 2009, 2008b,a, 2005; Burgasser & Putman 2005; Hallinan et al. 2007; Route & Wolszczan2012; McLean et al. 2012). The inferred luminosity is several orders of magnitude larger than that of anypreviously observed low-mass star or brown dwarf, either in a quiescent or flaring state. secondary (low-mass X-ray binaries, or LMXBs).Black hole LMXBs are typically characterized by well-defined “states”: correlationsbetween X-ray spectra, X-ray flux, and radio emission (Remillard & McClintock 2006).Radio emission is observed in a well-defined region of this hardness-intensity phase space(Fender et al. 2004; Falcke et al. 2004). In the so-called low-hard state, thought to cor-respond to low ( . . L Edd ), radiatively inefficient accretion (Esin et al. 1997), relatively 18 –steady radio emission from a jet is observed in most black hole X-ray binary systems. LikePTF11agg, the radio spectrum is flat or inverted ( f ν ∝ ν β , with β ≈ f ν (radio) ∝ f ν (X − ray) . . Using the derived formulation from Gallo et al. (2003) and theobserved radio flux, we would expect an X-ray flux of f X ≈ × − erg cm − s − (note thatthis estimate is entirely independent of the distance to the source). This is more than anorder of magnitude above our derived X-ray limits. We further note that while neutron starX-ray binaries do not obey the same radio–X-ray correlation in the hard state, the ratio ofX-ray to radio luminosity is even larger in these sources (Muno et al. 2005).Alternatively, the most luminous radio flares from LMXBs arise as the system transitionsthrough the intermediate state into a bright, quasi-thermal outburst (jet emission at thehighest X-ray fluxes appears to be largely suppressed; Fender et al. 2004). Unlike the steadyradio jets in the low-hard phase, this state transition in the accretion flow (from radiativelyinefficient, advection-dominated to geometrically thin, optically thick; Esin et al. 1997) cansometimes cause the ejection of relativistic material (Mirabel & Rodr´ıguez 1999). WhileLMXBs in this state do not always follow the same radio–X-ray correlation (Gallo et al.2003), the radio spectrum from this extended emission becomes optically thin. The X-rayand optical fluxes can rise by several orders of magnitude on a time scale of only a few daysduring these “X-ray novae,” but typically both take many months to return to quiescence(Tanaka & Shibazaki 1996; Charles & Coe 2006).PTF11agg differs from these X-ray novae in several major respects. Most importantly,to reach the intermediate (and, ultimately, high) state where the radio flare is launched, thecompact primary must be accreting material at a substantial fraction of the Eddington limit( & γ -ray satellites. But for any reasonable Galactic distance scale ( d .
10 kpc), our X-ray limits rule out emission at the level of 10 − L Edd (for a 1 M ⊙ black holeor neutron star). While our X-ray observations were obtained 42 d after the initial opticaloutburst, this is comparable to the e -folding time of these systems. As it requires ∼ t . τ ≈ l = 202 ◦ , b = +29 ◦ ), is inconsistent with the known population ofLMXBs (van Paradijs & White 1995; White & van Paradijs 1996), which have a scale heightof d z . ∼ new type of Galactic outburst, characterized by (1) bright, rapidly fading opticalemission; (2) a long-lived radio transient; and (3) an extremely subluminous ( M R ≈
11 magfor d = 10 kpc) quiescent optical counterpart. Given the broad agreement between ourobservations and the properties of long-duration GRB afterglows ( §
4. An Extragalactic Origin: Implications and Comparisons
Having rejected a Galactic origin for PTF11agg, we now consider the possibility thatit resides instead at a cosmological distance (i.e., well beyond the Local Group and into theHubble flow). Assuming the quiescent counterpart is indeed the host galaxy of PTF11agg,we constrain possible redshifts in § § § § Assuming the quiescent optical source is related to PTF11agg ( § I absorption along the line ofsight (i.e., the Lyman break). Our g ′ detection implies that redshifted Ly α ( λ rest = 1216 ˚A)falls at an observed wavelength of λ Ly α . g ′ filter bandpass).This results in an upper limit on the host-galaxy redshift of z . M UV . −
16 mag, or L & . L ∗ ; Reddy et al. 2008), we place a lower limit on the host redshift of z & .
5. A similar lower limit is derived if we compare the observed R -band brightness withthat of known host galaxies of long-duration GRBs (Jakobsson et al. 2012).We therefore conclude that the redshift of PTF11agg should fall somewhere in the range0 . . z . . In § > µ as at ∆ t obs ≈
42 d, and Θ ≈ µ as at∆ t obs ≈
100 d. To convert these to constraints on the outflow velocity, we use the redshiftlimits derived above: 0 . . z . . βctd A (1 + z ) , (6)where Γ is the outflow Lorentz factor (Γ ≡ (1 − β ) − / ), c is the speed of light, t is thetime since outburst (in the observer frame), d A is the angular-diameter distance, and z isthe source redshift.At z = 0 .
5, where our limits on the outflow velocity are the weakest, we find Γ > . t obs ≈
42 d, and Γ ≈ . t obs ≈
100 d. These limits vary little over our redshift rangeof interest, due primarily to the limited evolution of the angular-diameter distance over thisrange: at z = 3 .
0, we find Γ > . t obs ≈
42 d, and Γ ≈ . t obs ≈
100 d. 21 –We therefore conclude that, even at this late time, the ejecta powering the transientemission from PTF11agg are at least transrelativistic. For any more realistic form for theejecta deceleration (e.g., Blandford & McKee 1976), we infer that PTF11agg was initially atleast a modestly relativistic explosion.
Only a handful of extragalactic sources are known to produce relativistic ejecta: GRBs,with initial Lorentz factors as least as large as several hundred (Lithwick & Sari 2001), andpossibly greater than 1000 (Abdo et al. 2009); AGNs, in particular the subclass of blazars,with Lorentz factors as large as 50 (Lister et al. 2009); and the recently discovered relativistictidal disruption flares (TDFs; Levan et al. 2011; Bloom et al. 2011; Zauderer et al. 2011;Burrows et al. 2011; Cenko et al. 2012), with initial Lorentz factors ∼
10 (Metzger et al.2012; Berger et al. 2012b; van Velzen et al. 2011). Though with only one or two examplesto date, the known relativistic TDFs do not appear to vary in the optical on time scales asshort as those of PTF11agg, where δt ≪ ∼ L X ≈ erg s − . Even at z = 3, our X-ray limits ( § L X < × erg s − .Blazars, however, are known to vary in the optical on short time scales ( δt < m p & .
1% of the population; Lovell et al. 2008). More importantly,the degree of optical variability observed from PTF11agg, in particular the amplitude frompeak to quiescence (∆ R & t ≈ § z & .
5. PTF11agg as a GRB: Untriggered, Orphan, or Dirty Fireball?
Broadly speaking, there are three reasons why a distant, relativistic outburst may lackdetected prompt high-energy emission. The null hypothesis is a lack of sky coverage (i.e.,an “untriggered” GRB), as the more sensitive high-energy satellites (
Swift and
Fermi ) haveonly a ∼
60% combined likelihood of detecting any given event ( § γ -ray fluence from the only all-sky satellite available (the IPN) corresponds to an isotropic γ -ray energy release of E γ, iso = (2–200) × erg from z = 0 . γ -ray emitting material (i.e., an on-axis orphan afterglow; Nakar & Piran 2003), or if, asexpected, the outflow spreads laterally at late times and illuminates an increasing fractionof the sky (i.e., an off-axis orphan afterglow; Rhoads 1997; Perna & Loeb 1998; Nakar et al.2002). The discovery of a bona fide orphan afterglow would provide robust constraints on theGRB beaming fraction, still a large source of uncertainty in calculations of the true energyrelease and the all-sky rate of GRBs.Finally, a source may lack detectable high-energy emission altogether, either becauseno high-energy photons were produced, or such emission may be unable to escape to distantobservers due to some internal suppression mechanism. It has long been noted (e.g., Piran2004, and references therein) that the baryon composition of the relativistic jet in the fireballmodel must be very finely tuned in order to generate any detectable prompt high-energyemission (the so-called “baryon loading problem”). Without any baryons in the ejecta,the internal shocks thought to power the prompt emission will not form . But with toolarge a baryon fraction, the jet will not accelerate to a sufficiently high initial Lorentz factor(Γ & e − – e + pair production (Huang et al. 2002;Ghirlanda et al. 2012). Such explosions, dubbed “dirty fireballs,” have long been predicted An alternative possibility is that the prompt emission is generated by magnetic dissipation in a Poynting-flux-dominated outflow (see, e.g., Lyutikov & Blandford 2003).
23 –(Dermer et al. 2000; Huang et al. 2002; Rhoads 2003) to occur as a result of a modest baryonloading of the jet; a proton content as small as M & − M ⊙ will lower the initial Lorentzfactor sufficiently (Γ ≈ E KE /M c ), yet can still produce the observed broad-band afterglow.Distinguishing between a source that produces no high-energy emission whatsoever and onein which these photons are unable to escape is clearly challenging – for the remainder of thiswork we shall refer to such objects generically as dirty fireballs or afterglows lacking prompthigh-energy emission.Here we attempt to discriminate between these competing hypotheses through two dif-ferent means. First, we distinguish between on-axis and off-axis models by comparing theobserved optical and radio emission with analytic and numerical predictions for GRB after-glow light curves in the fireball model ( § § In the standard GRB afterglow fireball model (see, e.g., Piran 2004 for a review), rel-ativistic ejecta with (kinetic) energy E KE sweep up material in the circumburst medium,forming a collisionless shock and accelerating electrons to a power-law distribution of en-ergies with exponent p and minimum Lorentz factor γ m . It is assumed that a constantfraction of the total post-shock energy density is partitioned to the electrons ( ǫ e ) and themagnetic field ( ǫ B ). These accelerated electrons then emit synchrotron radiation, poweringthe long-lived X-ray, optical, and radio afterglow.The observed afterglow spectrum depends on the relative ordering of three critical fre-quencies: the frequency where self-absorption becomes important ( ν a ), the characteristicfrequency of the emission ( ν m ), and the frequency above which electrons are able to cool effi-ciently through radiation ( ν c ). We shall assume that all our observations occur in the “slow”cooling regime ( ν m < ν c ), and that the self-absorption frequency falls below the frequencyrange probed by our observations ( ν a < Hz).The light curve produced by such emission depends on the radial profile of the circum-burst medium into which the shock is expanding. The simplest circumburst medium toconsider is one in which the density is constant ( ρ ∝ r ). This scenario is also referred to asan interstellar medium (ISM; Sari et al. 1998), and is parametrized in terms of the particlenumber density n , where ρ = m p n g cm − .Long-duration GRBs, however, have been conclusively linked to the deaths of massivestars (e.g., Woosley & Bloom 2006). In the late stages of evolution, massive Wolf-Rayet 24 –stars are stripped of their outer H and (possibly) He envelopes in a wind, leaving behinda signature ρ ∝ r − density profile that should be discernible in the afterglow light curve.Wind-like environments (Chevalier & Li 2000) are parametrized in terms of A ∗ , where ρ =5 × A ∗ r − g cm − .Finally, we note that the hydrodynamical evolution also depends on the geometry of theoutflow. GRBs are now widely believed to be aspherical explosions (Rhoads 1999; Sari et al.1999), biconical jets with half-opening angle θ j . At early times, the jet emission is collimatedinto a narrow cone ( θ eff ≈ Γ − ≪ θ j , where Γ is the Lorentz factor of the expanding shock)due to relativistic beaming. As the shock slows, however, simple analytic solutions suggestthat lateral spreading of the jet becomes important, and on-axis observers eventually “miss”emission from wider angles. This hydrodynamic transition manifests itself as an achromaticsteepening in the afterglow light curve (the “jet break”), with an expected post-break decayproportional to t − p . While more recent numerical simulations have suggested a more complexpicture of the jet-break phenomenon (Zhang & MacFadyen 2009; van Eerten et al. 2010;Granot & Piran 2012), the assumption of a general light-curve steepening around the timewhen Γ = 1 /θ j remains largely valid.At ∆ t ≈ . ν opt < ν c ). If we also assume that these earlyoptical data occur before any jet break (see below), the observed temporal decay index( α opt = 1 . ± .
35) can be translated directly into the electron spectral index p . For aconstant-density medium, we find p = 3 . ± .
47, while for a wind-like environment, weinfer p = 2 . ± .
46. Electron spectral indices derived from previous observations of long-duration GRBs (Shen et al. 2006; Starling et al. 2008; Curran et al. 2010) fall in the range ∼ t &
40 d, the approximate radiospectral index, β radio ≈ .
3, implies that the peak synchrotron frequency ν m is not wellbelow the radio at this time (or else we would expect β radio ≈ − ν m (∆ t = 40 d) &
10 GHz, and f ν m (∆ t = 40 d) & µ Jy. For a wind-like medium, ν m ∝ t − / and f ν m ∝ t − / . Extrapolating back to the time of optical discovery, we conclude ν m (∆ t = 0 . & × Hz, and f ν m (∆ t = 0 . & × µ Jy. For ν m < ν < ν c , f ν ∝ ν (1 − p ) / ≈ ν − . . Thus, we find that the inferred optical ( R -band) flux at discovery, f ν & µ Jy, is a factor of ∼ , largelyconsistent with values derived from previous GRB afterglow modeling (Panaitescu & Kumar2001a,b; Yost et al. 2003), although with a somewhat smaller circumburst density ( n . . − ). The best-fit model, assuming the observer is oriented directly along the jet axis (i.e., θ obs = 0), is plotted in Figure 6 ( E KE = 3 × erg, θ j = 0 .
50 rad, n = 1 × − cm − , ǫ e = 0 . ǫ B = 0 .
1, and p = 2 . z = 1 models is the implied angular size of the source (Θ).In the best-fit on-axis model, we find that the outflow has an angular size Θ ≈
40 mas at∆ t = 40 d. This is somewhat larger than the value inferred from our scintillation analysis,but within a factor of two. However, many of the other z = 1 models that provide areasonable fit to the data are likely to be too spatially extended to scintillate strongly at∆ t ≈
100 d. One potential solution may be a more distant origin; due to cosmological timedilation, an observer-frame time of ∆ t = 100 d would correspond to only 25 rest-frame dayspost-explosion at z = 3, half the expansion time as inferred at z = 1. Given the large spreadin acceptable models, however, we do not explore this possibility further here.Next we consider limits on the opening angle and observer orientation from the observedoptical and radio emission. After the jet break, the peak synchrotron flux declines linearlywith time (in a constant-density environment). Thus, if the jet break occurred well beforethe first radio observations, the large radio flux would be difficult to reconcile with our earlyoptical observations. We consider it likely, then, that t j &
40 d. These conclusions are largelyconfirmed by our numerical models, where we find that the opening angle is only weaklyconstrained to be θ j & .
15 rad.In addition to cases where the observer is oriented directly along the jet axis (i.e., θ obs = 0), we also have considered more general geometries, where the observer may beoriented off-axis, either within (i.e., θ obs < θ j ) or outside ( θ obs > θ j ) the jet opening angle.For simplicity, we consider only “top-hat” jet geometries, where the jet Lorentz factor isgiven by a step function. For θ obs > θ j , observers will see rising emission until approximately In all cases the predicted X-ray flux is well below the XRT limit ( §
26 – -1 Time Since Explosion (d) -1 F l u x D e n s i t y ( µ J y ) Model 1 (θ obs =0; θ jet =0.50)Model 2 (θ obs =0.19; θ jet =0.20)Optical (R-band)Radio (8 GHz)
Fig. 6.—
The solid line shows the best-fit afterglow model (van Eerten et al. 2012) when the relativisticjet is oriented directly along the line of sight to the observer (i.e., θ obs = 0). The dashed curve displaysthe best-fit model when the viewing angle is allowed to vary freely. Given the relatively sparse dataset (inparticular the lack of X-ray observations), a wide variety of models are able to reproduce the observed opticaland radio emission. However, we find it impossible to reproduce the observed emission when the viewingangle is outside the cone of the jet (i.e., θ obs > θ j ). We therefore consider it unlikely that viewing angle alonecan account for the lack of high-energy emission from PTF11agg. the time of the jet break, after which the decay will resemble the on-axis case. For observersoff-axis but within the jet opening angle, the modifications to the on-axis afterglow lightcurves will be more subtle (Granot et al. 2002; Zhang & MacFadyen 2009; van Eerten et al.2010).The arguments used above to infer t j &
40 d necessarily require that the observer cannotbe well outside the jet opening angle (or else we would expect to see post jet-break decay). 27 –This result is borne out by our numerical modeling, where geometries with θ obs > θ j areunable to accurately reproduce the observed light curves. Allowing the observer orientationto vary as a free parameter, the best-fit afterglow model is plotted as a dashed line in Figure 6( E KE = 9 × erg, θ j = 0 .
20 rad, θ obs = 0 .
19 rad, n = 1 × − cm − , ǫ e = 0 . ǫ B = 0 . p = 3 . t = 67 d, f ν ∝ ν β , with β ≈ .
3, is in-consistent with Sedov-Taylor blast-wave evolution. In other words, the outgoing shock wavehas not transitioned to nonrelativistic expansion at this point in time. The nonrelativistictransition will occur at a time (Wygoda et al. 2011) t nr = 1100 (cid:18) E KE erg (cid:19) (cid:16) n − (cid:17) d . (7)Even neglecting our previous finding of a low circumburst density, we infer a sizeable lowerlimit on the blast-wave kinetic energy: E KE & erg. Integrating over the observed 8 GHzradio light curve, we measure a fluence of S rad = 2 × − erg cm − . For a typical cosmologicaldistance ( z = 1, or d L = 2 × cm), this corresponds to a radiated energy of E rad =3 × erg, a typical radiative efficiency for a GRB. But for a Galactic outburst, the radiatedenergy would be many orders of magnitude smaller ( E rad = 6 × erg at d = 10 kpc). Unlessthe radiative efficiency was incredibly small, we once again conclude that PTF11agg mustlie at a cosmological distance. Our objective in this section is to estimate the number of GRB optical afterglows dis-covered by chance (i.e., not as a result of deliberate follow-up observations of a high-energytrigger) by PTF. If the likelihood of chance detection of an untriggered afterglow with PTFis significant, we will conclude that the rate of PTF11agg-like events is consistent with therate of normal (i.e., on-axis) long-duration GRBs. If this probability is small, then we canuse these calculations to place lower limits on the observed frequency of PTF11agg-like out-bursts (in units of the GRB rate). Given the relatively complex nature of PTF scheduling(Law et al. 2009), we have conducted a series of Monte Carlo simulations to this end.PTF began full operations on about 2009 April 1. We have retrieved a listing of allimages obtained beginning at this time through 2012 December 31, or over a period of 45months. We removed fields at Galactic latitude | b | < ◦ (due to the large foreground 28 –extinction) . The resulting sample includes 129,206 pointings, each covering an area of7.2 deg . The sample comprises 1940 unique fields, each imaged an average of 67 times.Since launch, the Swift
BAT detects GRBs at a rate of ≈
90 yr − . The field-of-viewof the BAT is ∼ ∼ ∼
630 yr − . Over the 3.75 yr period of interest, the totalnumber of all-sky GRBs is ∼ t (uniformly distributed between 2009 April 1 and 2012 December 31) andspatial coordinates α , δ (isotropically distributed on the sky). To estimate the duration overwhich the optical afterglow would be detectable by PTF, we utilize the sample of 29 long-duration afterglows from the Palomar 60 inch (P60) Swift afterglow catalog (Cenko et al.2009). These events were selected solely on the basis of visibility to Palomar Observatory,so they should represent an unbiased sample of the
Swift afterglow brightness distribution.For each event in the P60-
Swift sample, we have calculated the amount of time followingthe high-energy trigger that the afterglow is brighter than R = 20 mag. These values rangefrom <
204 s (GRB 050721) to 1.2 d (GRB 050820A). Each mock GRB is randomly assignedone of the 29 actual “visibility windows” from this sample .For each mock GRB, we then determine if the event occurred within the 7.2 deg foot-print of any individual PTF image, and, if so, if the time of observation occurred withinthe necessary window during which the afterglow was brighter than 20 mag. The number ofafterglows detected in each trial ( N GRB ), together with the number of individual frames onwhich each detected afterglow was brighter than the P48 sensitivity limit ( N Det ), were thenrecorded. The results of 1000 individual runs (i.e., different randomly selected groups of2360 GRBs) constitute a sufficiently large sample to evaluate the likelihood of serendipitousdetection of long-duration GRB afterglows with PTF.In the 1000 trials conducted, at least one GRB afterglow was detected (i.e., N GRB ≥ P ( N GRB ≥
1) = 97%. The expectation Given that the primary objective of PTF is the discovery of extragalactic transients, this represents lessthan 10% of the total number of observations. See http://swift.gsfc.nasa.gov/docs/swift/archive/grb table . For GRBs without any detected optical afterglow (e.g., “dark” bursts), we use the earliest non-detectionbelow our sensitivity threshold for the visibility window. If anything, this would bias us to over-estimate theexpected number of untriggered GRB afterglow detections by PTF.
29 –value for the number of afterglows detected is λ = 3 .
3. The distribution of the numberof afterglows detected in our 1000 trials is reasonably well described by Poisson statistics(Figure 7). In this respect, then, PTF11agg appears to be consistent with a normal on-axisGRB.However, the field in which PTF11agg was identified (the Beehive cluster) is atypicalamongst PTF pointings. Most fields are only observed 2 or 3 times per night (multiple imagesare used to identify Solar-System objects). But the Beehive cluster is a “high-cadence” field,observed many times ( &
10) per night during its observing season. Instead of calculatingthe rate of afterglow detections over the entire survey (i.e., N GRB ), a more appropriatecomparison would limit the scope to similar high-cadence fields.We therefore consider on how many individual images each of the 3340 “detected” GRBafterglows (in our 1000 trials) were above the P48 limiting magnitude (i.e., N Det ). This isillustrated in Table 3. The vast majority of the afterglows are detected on only one or twoimages (87%). In fact, in our 1000 trials, an optical afterglow was detected on at least tenindividual images only 11 times (i.e., P ( N Det ≥
10) = 2 . R <
20 mag.We can understand this result analytically in the following manner. In the case wherethe integration time ( δt ) is much smaller than the period over which a transient is visible( τ ), the number of detectable events at any given time will be q = Ω N τ π, (8)where Ω is the field of view (in steradian) and N is the all-sky event rate. For long-durationGRB optical afterglows, serendipitous detection by PTF will be dominated by the ∼ R <
20 mag for τ ≈ N Det > / π = 1 . × − sr (7.2 deg ), andadopting N ≈ ×
630 yr − and τ ≈ . × − (1 d), we find q ≈ . × − events perfield.The expected number of detected events, λ , will then be qN Obs , where N Obs is the numberof (independent) measurement epochs. Over the two-year period of interest, the number ofindividual P48 images obtained is N Obs (all) = 1 . × . Thus, we predict λ ≈ .
7, in goodagreement with the results of our Monte Carlo simulations.Conversely, we can calculate the relative frequency of high-cadence ( N Obs [ > N Obs ( >
10) (i.e., more than 10 observations 30 –
Number of On-Axis GBs Detected N o r m a li z e d F r e q u e n c y Fig. 7.—
Normalized histogram of the number of serendipitous detections of normal on-axis GRB afterglowsby PTF in our 1000 Monte Carlo trials. The distribution is reasonably well described by a Poisson functionwith λ = 3 . of a field obtained in a single night) occur with a frequency of 1% when compared withregular-cadence fields ( N Obs [1] + N Obs [2]).From this analysis, we conclude that the rate of PTF11agg-like events is inconsistentwith the rate of long-duration GRBs with 97.4% confidence. Admittedly a number of as-sumptions went into this analysis, and one should always be careful with results drawn fromsuch an a posteriori analysis. But independent of the exact likelihood, we conclude that theprobability of untriggered afterglow detection in a high-cadence PTF field is small. Eitherwe have been quite lucky, or we may have uncovered a new, more common class of distant,relativistic outbursts lacking entirely in high-energy emission. 31 –It is crucial to verify, however, that our inferred rate does not violate any other limits onshort-timescale transients, either from PTF itself, or from previous optical and radio surveys.As highlighted above, low-cadence fields are observed significantly more frequently with PTFthan high-cadence fields like the Beehive. Thus, any short-timescale (∆ t . N Obs (1) and N Obs (2) fields than high-cadence fields. In thecase of PTF11agg, repeating the above Monte Carlo simulations for a transient populationwith five times the GRB event rate (but the same optical brightness distribution), we findan expected number of detected sources of λ = 16 . bona fide transient(e.g., to “discover” the source). Thus, any PTF11agg-like outburst with only a single de-tection ( N Det = 1) will never be discovered by our survey. Likewise, there may be subtlebiases limiting our capability to identify and/or conduct follow-up observations of similarshort-timescale transients with only a few detections.Whether these discovery biases are sufficient to account for the lack of similar sources inour low-cadence fields with PTF remains to be seen. We have attempted to search through allPTF discoveries that were detected only on a single night (independent of N Det ) , but haveyet to uncover any additional viable candidates. Ultimately, future wide-field, high-cadenceoptical surveys may be required to resolve this issue.Finally, we compare our derived rate of PTF11agg-like events with previous searchesfor orphan optical (Vanden Berk et al. 2002; Becker et al. 2004; Rykoff et al. 2005; Rau et al.2006, 2008) and radio (Levinson et al. 2002; Berger et al. 2003; Gal-Yam et al. 2006; Soderberg et al.2006a) afterglows, to verify that our results are consistent with these limits. The tightest con-straints on the rate of relativistic outbursts come from radio surveys, where Gal-Yam et al.(2006) derive a limit on the all-sky volumetric rate of GRB-like explosions of ˙ N < eventsGpc − yr − . Even assuming an all-sky GRB rate as large (Guetta & Della Valle 2007) as100 Gpc − yr − (more recent estimates suggest a significantly smaller value; Butler et al.2010), a population of PTF11agg-like events occurring at a rate of ∼ We cannot avoid the requirement of at least two detections, however. Otherwise we would be completelyswamped with asteroids, which are detected at a rate of thousands per night.
32 –GRBs is consistent with these results. Our derived rate is therefore orders of magnitude lowerthan the all-sky rate of Type Ibc supernovae ( ˙ N = 2 . × Gpc − yr − ; Li et al. 2011). Itmay approach the rate of low-luminosity GRBs (Soderberg et al. 2006b; Cobb et al. 2006;Guetta & Della Valle 2007), although this depends both on the assumed beaming correctionand the true GRB rate.
6. Summary and Conclusions
To summarize our results, we report here the discovery of PTF11agg, a rapidly fadingoptical transient with a long-lived, scintillating radio counterpart. Together with the ob-served optical and radio light curves, the detection of a faint, blue quiescent counterpart atthe location of PTF11agg indicates that the transient likely originated in the distant uni-verse. Using our measurements of the source size derived from the radio observations, weinfer that PTF11agg must be powered by a relativistic outflow. These properties are allconsistent with the population of long-duration GRB afterglows, marking the first time suchan event has been discovered at cosmological distances without a high-energy trigger.Searching various high-energy satellites, we find no potential γ -ray counterpart forPTF11agg. We therefore consider three possible explanations that can simultaneously ac-count for a GRB-like explosion without any associated prompt high-energy emission: anuntriggered GRB, an orphan afterglow, and a dirty fireball.Using the all-sky rate of GRBs discovered by the Swift satellite, together with a mea-surement of their observed optical brightness distribution, we have calculated the likelihoodof serendipitous untriggered GRB afterglow detection by PTF (April 2009 – December 2012).Surprisingly, we found that the a posteriori probability of untriggered GRB afterglow de-tection in a high-cadence field like the one where PTF11agg was found (11 observations ona single night) is only 2.6%. While we cannot rule out entirely our null hypothesis thatPTF11agg is an untriggered GRB, this probability is sufficiently low that we consider alter-native interpretations as well.The afterglow emission from an orphan GRB will rise in flux at early times, as moreand more of the jet becomes visible due to relativistic beaming effects. Using both analyticand numerical formulations, we are unable to reproduce the observed PTF11agg light curvesunless the observer viewing angle is within the opening angle of the jet. While these modelsassume a relatively simple jet structure, the requirement of rising afterglow emission at earlytimes is a robust prediction for all off-axis models.A more intriguing possibility is that PTF11agg may represent a new class of relativistic 33 –outbursts with little or no corresponding high-energy emission. In much the same way thatSN 2009bb (Soderberg et al. 2010) demonstrated that the more nearby, subluminous class ofGRBs may generate relativistic ejecta yet lack high-energy emission, PTF11agg may play ananalogous role for the more energetic, cosmologically distant sample of long-duration GRBs.Dirty fireballs (i.e., a baryon-loaded jet) are one possible explanation, though alternativepossibilities surely exist as well.In this picture, the inferred rate of PTF11agg-like events must be four times higher(90% confidence) than the rate of on-axis long-duration GRBs. When combined with tradi-tional core-collapse supernovae and long-duration GRBs, these objects would enable a morecomplete census of the deaths of massive stars, and also provide a probe of the locationof massive-star formation in distant galaxies without the need for a high-energy satellitetrigger.Regardless of its ultimate origin, we expect such sources to be discovered in large num-bers by ongoing and future wide-field, high-cadence optical surveys such as the CatalinaReal-Time Transient Survey (Drake et al. 2009), PTF, Pan-STARRS (Kaiser et al. 2010),and the Large Synoptic Survey Telescope (Ivezic et al. 2008). Furthermore, the discovery ofPTF11agg bodes well for optical surveys in the future era of gravitational wave astronomy, asthe electromagnetic counterparts of gravitational wave sources should exhibit largely similarobservational signatures (though they are also expected to be associated with more nearbygalaxies; Nakar & Piran 2011; Metzger & Berger 2012).We wish to thank David Levitan and Kunal Mooley for obtaining observations used inthis work, and John Tomsick for valuable comments on the manuscript. We are gratefulto the following IPN team members for sharing their data: K. Hurley, I. G. Mitrofanov,D. Golovin, M. L. Litvak, A. B. Sanin, W. Boynton, C. Fellows, K. Harshman, R. Starr,S. Golenetskii, R. Aptekar, E. Mazets, V. Pal’shin, D. Frederiks, D. Svinkin, A. von Kienlin,X. Zhang, K. Yamaoka, T. Takahashi, M. Ohno, Y. Hanabata, Y. Fukazawa, M. Tashiro,Y. Terada, T. Murakami, K. Makishima, T. Cline , S. Barthelmy, J. Cummings, N. Gehrels,H. Krimm, D. Palmer, J. Goldsten, V. Connaughton, M. S. Briggs, and C. Meegan.A.V.F. and his group acknowledge generous financial assistance from Gary & Cyn-thia Bengier, the Richard & Rhoda Goldman Fund, the Christopher R. Redlich Fund,NASA/
Swift grants NNX10AI21G and NNX12AD73G, the TABASGO Foundation, andNSF grant AST-1211916. A.C. acknowledges support from LIGO, which was constructed bythe California Institute of Technology and the Massachusetts Institute of Technology withfunding from the NSF and operates under cooperative agreement PHY-0757058. D.A.P. issupported by NASA through Hubble Fellowship grant HST-HF-51296.01-A awarded by the 34 –Space Telescope Science Institute, which is operated by the Association of Universities for Re-search in Astronomy, Inc., for NASA, under contract NAS 5-26555. Research by A.G.Y. andhis team is supported by grants from the ISF, BSF, GIF, the EU/FP7 via an ERC grant,and a Kimmel Award. P.J.G. acknowledges support from Caltech during his 2011 sabbaticalstay. E.O.O. is incumbent of the Arye Dissentshik career development chair and is gratefulto support by a grant from the Israeli Ministry of Science. A.A.M. is supported by theNSF Graduate Research Fellowship Program. J.S.B. acknowledges NSF grant CDI-0941742.M.M.K acknowledges generous support from the Hubble Fellowship and Carnegie-PrincetonFellowship. D.P. is grateful for the AXA research fund. A.S. is supported by a MinervaFellowship.Observations were obtained with the Samuel Oschin telescope and the Hale telescopeat Palomar Observatory as part of the Palomar Transient Factory project, a scientic collab-oration between the California Institute of Technology, Columbia University, Las CumbresObservatory, the Lawrence Berkeley National Laboratory, the National Energy Research Sci-entic Computing Center, the University of Oxford, and the Weizmann Institute of Science.The National Energy Research Scientific Computing Center, supported by the Office of Sci-ence of the U.S. Department of Energy, provided staff, computational resources, and datastorage for this project. Support for CARMA construction was derived from the Gordon andBetty Moore Foundation, the Kenneth T. and Eileen L. Norris Foundation, the James S. Mc-Donnell Foundation, the Associates of the California Institute of Technology, the Universityof Chicago, the states of California, Illinois, and Maryland, and the NSF. Ongoing CARMAdevelopment and operations are supported by the NSF under a cooperative agreement, andby the CARMA partner universities. Some of the data presented herein were obtained at theW. M. Keck Observatory, which is operated as a scientific partnership among the CaliforniaInstitute of Technology, the University of California and NASA; the Observatory was madepossible by the generous financial support of the W. M. Keck Foundation. PAIRITEL isoperated by the Smithsonian Astrophysical Observatory (SAO) and was made possible bya grant from the Harvard University Milton Fund, a camera loan from the University ofVirginia, and continued support of the SAO and UC Berkeley. The PAIRITEL project isfurther supported by NASA/
Swift
Guest Investigator grant NNX08AN84G. This work madeuse of data supplied by the UK
Swift
Science Data Centre at the University of Leicester. Italso made use of the NASA/IPAC Extragalactic Database (NED), which is operated by theJet Propulsion Laboratory, California Institute of Technology, under contract with NASA.In addition, we have utilized the SIMBAD database, operated at CDS, Strasbourg, France.
Facilities:
PO: 1.2m (PTF), Hale (WIRC), Keck:I (LRIS), Magellan: Baade (IMACS),FLWO: 2MASS (PAIRITEL), VLA, CARMA, Fermi (GBM), Swift (BAT, XRT) 35 –
REFERENCES
Abdo, A. A., Ackermann, M., Ajello, M., et al. 2009, ApJ, 706, L138Ag¨ueros, M. A., Covey, K. R., Lemonias, J. J., et al. 2011, ApJ, 740, 110Aihara, H., Allende Prieto, C., An, D., et al. 2011, ApJS, 193, 29Antonova, A., Doyle, J. G., Hallinan, G., Golden, A., & Koen, C. 2007, A&A, 472, 257Audard, M., Osten, R. A., Brown, A., et al. 2007, A&A, 471, L63Barthelmy, S. D., Barbier, L. M., Cummings, J. R., et al. 2005, Space Sci. Rev., 120, 143Becker, A. C., Wittman, D. M., Boeshaar, P. C., et al. 2004, ApJ, 611, 418Benz, A. O., Fuerst, E., & Kiplinger, A. L. 1983, Nature, 302, 45Benz, A. O., Gudel, M., & Mattei, J. A. 1996, in Astronomical Society of the Pacific Con-ference Series, Vol. 93, Radio Emission from the Stars and the Sun, ed. A. R. Taylor& J. M. Paredes, 188Berger, E. 2002, ApJ, 572, 503—. 2006, ApJ, 648, 629Berger, E., Fong, W., Sanders, N., & Chornock, R. 2012a, The Astronomer’s Telegram, 4619,1Berger, E., Kulkarni, S. R., Frail, D. A., & Soderberg, A. M. 2003, ApJ, 599, 408Berger, E., Zauderer, A., Pooley, G. G., et al. 2012b, ApJ, 748, 36Berger, E., Ball, S., Becker, K. M., et al. 2001, Nature, 410, 338Berger, E., Rutledge, R. E., Reid, I. N., et al. 2005, ApJ, 627, 960Berger, E., Gizis, J. E., Giampapa, M. S., et al. 2008a, ApJ, 673, 1080Berger, E., Basri, G., Gizis, J. E., et al. 2008b, ApJ, 676, 1307Berger, E., Rutledge, R. E., Phan-Bao, N., et al. 2009, ApJ, 695, 310Berger, E., Basri, G., Fleming, T. A., et al. 2010, ApJ, 709, 332 36 –Bertin, E., Mellier, Y., Radovich, M., et al. 2002, in Astronomical Society of the PacificConference Series, Vol. 281, Astronomical Data Analysis Software and Systems XI,ed. D. A. Bohlender, D. Durand, & T. H. Handley, 228Blandford, R. D., & McKee, C. F. 1976, Physics of Fluids, 19, 1130Blanton, M. R., & Roweis, S. 2007, AJ, 133, 734Bloom, J. S., Kulkarni, S. R., & Djorgovski, S. G. 2002, AJ, 123, 1111Bloom, J. S., Starr, D. L., Blake, C. H., Skrutskie, M. F., & Falco, E. E. 2006, in Astronom-ical Society of the Pacific Conference Series, Vol. 351, Astronomical Data AnalysisSoftware and Systems XV, ed. C. Gabriel, C. Arviset, D. Ponz, & S. Enrique, 751Bloom, J. S., Giannios, D., Metzger, B. D., et al. 2011, Science, 333, 203Bloom, J. S., Richards, J. W., Nugent, P. E., et al. 2012, PASP, 124, 1175Burgasser, A. J., & Putman, M. E. 2005, ApJ, 626, 486Burrows, D. N., Hill, J. E., Nousek, J. A., et al. 2005, Space Sci. Rev., 120, 165Burrows, D. N., Kennea, J. A., Ghisellini, G., et al. 2011, Nature, 476, 421Butler, N. R., Bloom, J. S., & Poznanski, D. 2010, ApJ, 711, 495Butler, N. R., & Kocevski, D. 2007, ApJ, 663, 407Butler, N. R., Kocevski, D., Bloom, J. S., & Curtis, J. L. 2007, ApJ, 671, 656Cannizzo, J. K. 1993, The Limit Cycle Instability in Dwarf Nova Accretion Disks, ed.Wheeler, J. C., 6Cardelli, J. A., Clayton, G. C., & Mathis, J. S. 1989, ApJ, 345, 245Cenko, S. B., Kelemen, J., Harrison, F. A., et al. 2009, ApJ, 693, 1484Cenko, S. B., Krimm, H. A., Horesh, A., et al. 2012, ApJ, 753, 77Chandra, P., & Frail, D. A. 2012, ApJ, 746, 156Charles, P. A., & Coe, M. J. 2006, Optical, ultraviolet and infrared observations of X-ray binaries, ed. Lewin, W. H. G. & van der Klis, M. (Cambridge, UK: CambridgeUniversity Press), 215 37 –Chevalier, R. A., & Li, Z.-Y. 2000, ApJ, 536, 195Cobb, B. E., Bailyn, C. D., van Dokkum, P. G., & Natarajan, P. 2006, ApJ, 645, L113Corbel, S., Nowak, M. A., Fender, R. P., Tzioumis, A. K., & Markoff, S. 2003, A&A, 400,1007Cordes, J. M., & Lazio, T. J. W. 2002, arXiv:astro-ph/0207156Curran, P. A., Evans, P. A., de Pasquale, M., Page, M. J., & van der Horst, A. J. 2010, ApJ,716, L135Dahn, C. C., Harris, H. C., Vrba, F. J., et al. 2002, AJ, 124, 1170Dermer, C. D., Chiang, J., & Mitman, K. E. 2000, ApJ, 537, 785Drake, A. J., Djorgovski, S. G., Mahabal, A., et al. 2009, ApJ, 696, 870Dressler, A., Bigelow, B., Hare, T., et al. 2011, PASP, 123, 288Esin, A. A., McClintock, J. E., & Narayan, R. 1997, ApJ, 489, 865Falcke, H., K¨ording, E., & Markoff, S. 2004, A&A, 414, 895Fender, R. P., Belloni, T. M., & Gallo, E. 2004, MNRAS, 355, 1105Gal-Yam, A., Ofek, E. O., Filippenko, A. V., Chornock, R., & Li, W. 2002, PASP, 114, 587Gal-Yam, A., Ofek, E. O., Poznanski, D., et al. 2006, ApJ, 639, 331Gallo, E., Fender, R. P., & Pooley, G. G. 2003, MNRAS, 344, 60Gehrels, N., Chincarini, G., Giommi, P., et al. 2004, ApJ, 611, 1005Geldzahler, B. J., Johnston, K. J., Spencer, J. H., et al. 1983, ApJ, 273, L65Ghirlanda, G., Nava, L., Ghisellini, G., et al. 2012, MNRAS, 420, 483Granot, J., Panaitescu, A., Kumar, P., & Woosley, S. E. 2002, ApJ, 570, L61Granot, J., & Piran, T. 2012, MNRAS, 421, 570G¨udel, M. 2002, ARA&A, 40, 217Guetta, D., & Della Valle, M. 2007, ApJ, 657, L73Hallinan, G., Bourke, S., Lane, C., et al. 2007, ApJ, 663, L25 38 –Hogg, D. W., Pahre, M. A., McCarthy, J. K., et al. 1997, MNRAS, 288, 404Huang, Y. F., Dai, Z. G., & Lu, T. 2002, MNRAS, 332, 735Hurley, K., Golenetskii, S., Aptekar, R., et al. 2010, in Deciphering the Ancient Universewith Gamma-Ray Bursts, Vol. 1279, American Institute of Physics Conference Series,ed. N. Kawai & S. Nagataki, 330–333Hynes, R. I., Mauche, C. W., Haswell, C. A., et al. 2000, ApJ, 539, L37Ivezic, Z., Tyson, J. A., Acosta, E., et al. 2008, arXiv:astro-ph/0805.2366Jakobsson, P., Hjorth, J., Malesani, D., et al. 2012, ApJ, 752, 62Kaiser, N., Burgett, W., Chambers, K., et al. 2010, in Ground-based and Airborne TelescopesIII, Vol. 7733, Society of Photo-Optical Instrumentation Engineers (SPIE) ConferenceSeries, 77330Kann, D. A., Klose, S., Zhang, B., et al. 2010, ApJ, 720, 1513King, A. R., & Ritter, H. 1998, MNRAS, 293, L42K¨ording, E., Rupen, M., Knigge, C., et al. 2008, Science, 320, 1318Kulkarni, S. R., & Rau, A. 2006, ApJ, 644, L63Kulkarni, S. R., Frail, D. A., Wieringa, M. H., et al. 1998, Nature, 395, 663Law, N. M., Kulkarni, S. R., Dekany, R. G., et al. 2009, PASP, 121, 1395Levan, A. J., Tanvir, N. R., Cenko, S. B., et al. 2011, Science, 333, 199Levinson, A., Ofek, E. O., Waxman, E., & Gal-Yam, A. 2002, ApJ, 576, 923Levitan, D., Fulton, B. J., Groot, P. J., et al. 2011, ApJ, 739, 68Li, W., Chornock, R., Leaman, J., et al. 2011, MNRAS, 412, 1473Lister, M. L., Cohen, M. H., Homan, D. C., et al. 2009, AJ, 138, 1874Lithwick, Y., & Sari, R. 2001, ApJ, 555, 540Lovell, J. E. J., Rickett, B. J., Macquart, J.-P., et al. 2008, ApJ, 689, 108Lyutikov, M., & Blandford, R. 2003, arXiv:astro-ph/0312347 39 –MacLeod, C. L., Ivezi´c, ˇZ., Sesar, B., et al. 2012, ApJ, 753, 106Madore, B. F., & Arp, H. C. 1979, ApJ, 227, L103Mahabal, A. A., et al. 2012, The Astronomer’s Telegram, 4586, 1McLean, M., Berger, E., & Reiners, A. 2012, ApJ, 746, 23Meegan, C., Lichti, G., Bhat, P. N., et al. 2009, ApJ, 702, 791Metzger, B. D., & Berger, E. 2012, ApJ, 746, 48Metzger, B. D., Giannios, D., & Mimica, P. 2012, MNRAS, 420, 3528Mirabel, I. F., Dhawan, V., Mignani, R. P., Rodrigues, I., & Guglielmetti, F. 2001, Nature,413, 139Mirabel, I. F., & Rodr´ıguez, L. F. 1999, ARA&A, 37, 409Muno, M. P., Belloni, T., Dhawan, V., et al. 2005, ApJ, 626, 1020Nakar, E., & Piran, T. 2003, New Astronomy, 8, 141—. 2011, Nature, 478, 82Nakar, E., Piran, T., & Granot, J. 2002, ApJ, 579, 699Nysewander, M., Fruchter, A. S., & Pe’er, A. 2009, ApJ, 701, 824Oates, S. R., Page, M. J., Schady, P., et al. 2009, MNRAS, 395, 490Ofek, E. O., Laher, R., Law, N., et al. 2012, PASP, 124, 62Oke, J. B., & Gunn, J. E. 1983, ApJ, 266, 713Oke, J. B., Cohen, J. G., Carr, M., et al. 1995, PASP, 107, 375Panaitescu, A., & Kumar, P. 2001a, ApJ, 560, L49—. 2001b, ApJ, 554, 667Panaitescu, A., & Vestrand, W. T. 2008, MNRAS, 387, 497Patten, B. M., Stauffer, J. R., Burrows, A., et al. 2006, ApJ, 651, 502Perley, R. A., Chandler, C. J., Butler, B. J., & Wrobel, J. M. 2011, ApJ, 739, L1 40 –Perna, R., & Loeb, A. 1998, ApJ, 509, L85Piran, T. 2004, Reviews of Modern Physics, 76, 1143Rahmer, G., Smith, R., Velur, V., et al. 2008, in Ground-based and Airborne Instrumentationfor Astronomy II, Vol. 7014, Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series, 70144Ramsay, G., Barclay, T., Steeghs, D., et al. 2012, MNRAS, 419, 2836Rau, A., Greiner, J., & Schwarz, R. 2006, A&A, 449, 79Rau, A., Ofek, E. O., Kulkarni, S. R., et al. 2008, ApJ, 682, 1205Rau, A., Schwarz, R., Kulkarni, S. R., et al. 2007, ApJ, 664, 474Rau, A., Kulkarni, S. R., Law, N. M., et al. 2009, PASP, 121, 1334Readhead, A. C. S. 1994, ApJ, 426, 51Reddy, N. A., Steidel, C. C., Pettini, M., et al. 2008, ApJS, 175, 48Remillard, R. A., & McClintock, J. E. 2006, ARA&A, 44, 49Rhoads, J. E. 1997, ApJ, 487, L1—. 1999, ApJ, 525, 737—. 2003, ApJ, 591, 1097Rickett, B. J. 1990, ARA&A, 28, 561Route, M., & Wolszczan, A. 2012, ApJ, 747, L22Rykoff, E. S., Aharonian, F., Akerlof, C. W., et al. 2005, ApJ, 631, 1032—. 2009, ApJ, 702, 489Sari, R., Piran, T., & Halpern, J. P. 1999, ApJ, 519, L17Sari, R., Piran, T., & Narayan, R. 1998, ApJ, 497, L17Sault, R. J., Teuben, P. J., & Wright, M. C. H. 1995, in Astronomical Society of the PacificConference Series, Vol. 77, Astronomical Data Analysis Software and Systems IV, ed.R. A. Shaw, H. E. Payne, & J. J. E. Hayes, 433 41 –Schlegel, D. J., Finkbeiner, D. P., & Davis, M. 1998, ApJ, 500, 525Shen, R., Kumar, P., & Robinson, E. L. 2006, MNRAS, 371, 1441Skrutskie, M. F., Cutri, R. M., Stiening, R., et al. 2006, AJ, 131, 1163Soderberg, A. M., Nakar, E., Berger, E., & Kulkarni, S. R. 2006a, ApJ, 638, 930Soderberg, A. M., Kulkarni, S. R., Nakar, E., et al. 2006b, Nature, 442, 1014Soderberg, A. M., Chakraborti, S., Pignata, G., et al. 2010, Nature, 463, 513Spergel, D. N., Bean, R., Dor´e, O., et al. 2007, ApJS, 170, 377Starling, R. L. C., van der Horst, A. J., Rol, E., et al. 2008, ApJ, 672, 433Tanaka, Y., & Shibazaki, N. 1996, ARA&A, 34, 607Tody, D. 1986, in Instrumentation in astronomy VI, Vol. 627, Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series, ed. D. L. Crawford, 733Tomsick, J. A., Halpern, J. P., Kemp, J., & Kaaret, P. 1999, ApJ, 521, 341Uemura, M., Kato, T., Matsumoto, K., et al. 2000, PASJ, 52, L15van Eerten, H., van der Horst, A., & MacFadyen, A. 2012, ApJ, 749, 44van Eerten, H., Zhang, W., & MacFadyen, A. 2010, ApJ, 722, 235van Paradijs, J., & White, N. 1995, ApJ, 447, L33van Velzen, S., K¨ording, E., & Falcke, H. 2011, MNRAS, 417, L51Vanden Berk, D. E., Lee, B. C., Wilhite, B. C., et al. 2002, ApJ, 576, 673Voges, W., Aschenbach, B., Boller, T., et al. 1999, A&A, 349, 389Wagner, R. M., Foltz, C. B., Shahbaz, T., et al. 2001, ApJ, 556, 42Walker, M. A. 1998, MNRAS, 294, 307White, N. E., & van Paradijs, J. 1996, ApJ, 473, L25Wilson, J. C., Eikenberry, S. S., Henderson, C. P., et al. 2003, in Instrument Design andPerformance for Optical/Infrared Ground-based Telescopes, Vol. 4841, Society ofPhoto-Optical Instrumentation Engineers (SPIE) Conference Series, ed. M. Iye &A. F. M. Moorwood, 451 42 –Woosley, S. E., & Bloom, J. S. 2006, ARA&A, 44, 507Wygoda, N., Waxman, E., & Frail, D. A. 2011, ApJ, 738, L23Yost, S. A., Harrison, F. A., Sari, R., & Frail, D. A. 2003, ApJ, 597, 459Zauderer, B. A., Berger, E., Soderberg, A. M., et al. 2011, Nature, 476, 425Zhang, W., & MacFadyen, A. 2009, ApJ, 698, 1261Zurita, C., Casares, J., Shahbaz, T., et al. 2000, MNRAS, 316, 137
This preprint was prepared with the AAS L A TEX macros v5.2.
43 –Table 1. Optical/Near-Infrared Observations of PTF11agg
Date Telescope/Instrument Filter Exposure Time Magnitude(MJD) (s)55590.30519 P48/CFHT12k R > . R
60 18 . ± . R
60 18 . ± . R
60 18 . ± . R
60 18 . ± . R
60 18 . ± . R
60 18 . ± . R
60 18 . ± . R
60 18 . ± . R
60 19 . ± . R
60 19 . ± . R
60 19 . ± . R
420 22 . ± . R > . H > . J > . K s > . g ′ . ± . R . ± . g ′ . ± . R . ± . I > . K s > .
44 –Table 2. Radio Observations of PTF11agg
Date Observatory Frequency Integration Time Flux Density(2011 UT) (GHz) (min) ( µ Jy)Mar. 11.27 VLA 8.46 15.5 300 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± − ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± < ±
45 –Table 3. PTF GRB Simulation Results
N N
Det N Obs >>