Mon.Not.R.Astron.Soc.000,000–000(0000) Printed26February2015 (MNLATEXstylefilev2.2) Bright but slow – Type II supernovae from OGLE-IV – Implications for magnitude limited surveys D. Poznanski1(cid:63), Z. Kostrzewa-Rutkowska,2, L. Wyrzykowski2,3, & N. Blagorodnova3 1School of Physics and Astronomy, Tel-Aviv University, Tel Aviv 69978, Israel. 2Warsaw University Observatory Al. Ujazdowskie 4, 00-478 Warszawa, Poland 3Institute of Astronomy, University of Cambridge, Madingley Road, Cambridge CB3 0HA, UK 5 26February2015 1 0 2 ABSTRACT b We study a sample of 11 Type II supernovae (SNe) discovered by the OGLE- e IV survey. All objects have well sampled I-band light curves, and at least one F spectrum. We find that 2 or 3 of the 11 SNe have a declining light curve, and 5 spectra consistent with other SNe II-L, while the rest have plateaus that can be 2 as short as 70d, unlike the 100d typically found in nearby galaxies. The OGLE SNe are also brighter, and show that magnitude limited surveys find SNe that ] are different than usually found in nearby galaxies. We discuss this sample in E the context of understanding Type II SNe as a class and their suggested use as H standard candles. . h Key words: Supernovae: general p - o r t s 1 INTRODUCTION ple,Arcavietal.(2012)findacleargapinthedistribution a oflight-curvedeclinerates,betweenstandardplateau-like [ Type II supernovae (SNe II), are perhaps the simplest SNeII-PanddecliningSNeII-L,aswellasratheruniform and in some ways best understood stellar explosions. 2 plateau durations for SNe II-P (near 100d). In contrast, We know they result from the core collapse of massive v Andersonetal.(2014)whouseabluerphotometricband, stars,thosewithmassesnearthe8–20M range,mostse- 2 (cid:12) findacontinuumofdeclinerates,removingtheabilityto 5 curely through archival progenitor detections (see review separatetheSNeII-PfromtheII-L(seesimilarresultby 4 by Smartt2009),weknowthattheirejectaarecomposed Sandersetal.2015).Faranetal.(2014a)andFaranetal. 3 ofmostlyhydrogen(seereviewofSNspectroscopictypes (2014b)finduniformplateausbutshowthatthegapinde- 0 byFilippenko1997),andfromextensiveobservationsand clinesemergeslargelyfromtheanalysismethod.Sanders . modeling we seem to have a fair understanding of most 1 et al. (2015) find rather uniform durations, if somewhat 0 of their photometric and spectroscopic evolution. shorter and with outliers (90±10d). While some of the 5 Nevertheless,manyquestionsremainunsatisfactorily differences are semantic, and others at least partly arise 1 answered,fromthemechanismleadingtotheirsuccessful from methodological differences (different bands, or defi- : explosion which is still largely mysterious (e.g., Bruenn v nitionoftheplateauduration),thesamplesdooftenseem et al. 2014, and references therein), to details regard- i different, which is puzzling. X ing their light curve shapes, and distribution of shapes, r through the fate of the more massive of these stars, near a 20M(cid:12)(Smartt 2009). In this short paper we attempt to address some of With advances in detector technology and comput- these discrepancies using yet another independent sam- ing,SNe,onceascarcecommodity,arenowobservablein ple. The SNe presented here have all been detected by largenumbers,andthefieldisevolvingfromdetaileddis- the Transient Detection System of the OGLE-IV survey cussions of single objects, to samples that are analyzed (Kozl(cid:32)owski et al. 2013), and are a subset of the SNe pre- in bulk. Several such samples of SNe II have been re- sented in Wyrzykowski et al. (2014). Briefly, in OGLE- centlyanalyzedviavariousmeans(e.g.,Arcavietal.2012; IV about 650 deg2 are observed with an average 5d ca- Maguire et al. 2012; Faran et al. 2014a; Anderson et al. dence. An automated pipeline finds transients down to 2014; Faran et al. 2014b; Sanders et al. 2015; Spiro et al. ∼ 20mag using image subtraction (see more details in 2014). However, with the availability of greater datasets Wyrzykowski et al. 2014). Using 11 SNe II, we examine somequestionshaveactuallybecomemuddier.Forexam- theirbasicobservationalparameters–suchaslightcurve shapes, ejecta velocities, luminosities – in the context of otherrecentlypublishedsamples–andattempttorecon- (cid:63) [email protected] cile often-conflicting findings. (cid:13)c 0000RAS 2 Poznanski et al. ⊕ 13-005 (63d) 22 ⊕ 13-011 (15d) 20 ⊕ 13-045 (36d) 18 ⊕ 13-046 (30d) 16 t ns ⊕ 13-047 (13d) o c14 + ⊕ 13-048 (30d) λ f ed12 ⊕ 13-135 (39d) z mali10 ⊕ 13-144 (15d) r o N 8 ⊕ 14-004 (48d) ⊕ 14-009 ( 7d) 6 ⊕ 14-009 (20d) 4 2 ⊕ 14-018 ( 8d) Hδ Hγ Hβ Hα all @ 8000 km s−1 4000 5000 6000 7000 8000 9000 10000 ˚ Rest Wavelength [A] Figure 1.Rest-framespectraofthesample(pink),overlaidwithsmoothedcurves(black),withtheirphase(asderivedfromthe photometry)inparentheses.Wemarkthetelluricfeatureat7614˚A,andtheBalmerseriesoffsetby8000kms−1. On top of the often-noisy spectra we overplot the result Table 1.SNSample ofsmoothingthemwithaSavitzky-Golayfilter(Savitzky &Golay1964).RedshiftswerecompiledbyWyrzykowski SNname zhost µ(mag)aE(B-V)MWb Explosion(MJD) et al. (2014), based on the spectroscopy. OGLE13-005 0.07 37.47 0.028 2456241.7±2.9 OGLE13-011 0.05 36.71 0.047 2456298.25±6.45 Qualitative inspection of the spectra reveals that OGLE13-045 0.10 38.29 0.030 2456483.4±2.5 OGLE13-046 0.07 37.47 0.028 2456489.9±2.0 OGLE-13-011, OGLE13-045, OGLE13-046, OGLE13- OGLE13-047 0.06 37.12 0.029 2456505.3±4.5 OGLE13-048 0.06 37.12 0.038 2456489.4±1.5 048, OGLE13-135, OGLE13-144, OGLE14-004, and OGLE13-135 0.057 37.00 0.157 2456620.65±2.05 OGLE14-018,allhavebroadhydrogenlines,andlooklike OGLE13-144 0.04 36.21 0.113 2456635.7±6.0 OGLE14-004 0.03 35.56 0.067 2456660.3±2.5 typical SNe II at their respective photometric phases. A OGLE14-009 0.056 36.96 0.068 2456688.2±1.5 OGLE14-018 0.03 35.56 0.087 2456702.2±1.5 classification with the SN Identification Program (SNID; Blondin & Tonry 2007) with the default templates and aDistance modulus assuming concordance cosmology (Planck parametersfindsthesame.OGLE13-144hastheweakest Collaborationetal.2014). ratio of Hα absorption to emission, as often seen in SNe bMilkyWayextinctionfromSchlafly&Finkbeiner(2011). II-L (Schlegel 1996; Guti´errez et al. 2014; Faran et al. 2014b). OGLE14-009 is featureless. The narrow hydro- 2 SPECTROSCOPY AND SAMPLE gen emission in one of the spectra is consistent with be- SELECTION ing from the host galaxy (full width at half maximum of about1000kms−1).Therearenoobviousspectralindica- OutoftheSNeinWyrzykowskietal.(2014),wefocuson tions it is securely a type II, though SNe II-L sometimes objects that can be spectroscopically classified as Type develop lines only later in their evolution (e.g., Faran II, either II-P or II-L, avoiding SNe of Type IIn and IIb. etal.2014b).Wetentativelykeepitinthesampleforfur- We further excluded objects with a light curve that was ther discussion, with the possibility it is not a type II-P clearlyIIb-like(i.e.,similartoSN1993J,Richmondetal. or II-L. Due to its lack of features (that is intrinsic, and 1994).Sincemostofourobjectsonlyhaveonespectrum, notduetoasignal-to-noiseissue,thecontinuumisclearly anditwasoftentakenearly,wemayhavesomeinterlop- detected) SNID is unhelpful in this case. OGLE13-047 is ing SN IIb in the sample. Most of the spectra were ac- best fit by SNID to the historical SNe II-L, SNe 1979C quired by the PESSTO project (Smartt et al. 2014), and and1980K,hasastrongHβ absorption,butnoHα.This downloaded from WISEREP (Yaron & Gal-Yam 2012), isreminiscentoftheearlyspectraoftheSNeII-L,2001fa except for OGLE13-0051 which was observed and clas- and 2005dq, and the superluminous II-L SN2008es who sified by Prieto & Morrell (2013). The SN list can be developed lines late in their evolution, starting from Hβ seen in Table 1, and the spectra in Figure 1, with the as well (Faran et al. 2014b). phases as determined from the photometry (see below). Based on spectral properties, we therefore find that thesampleisindeedcomposedofSNeII,asconstructed, 1 For compactness, we somewhat shorten the names. OGLE- most of them ‘regular’ SNe II-P, while OGLE13-047, SN-20XX-YYYisshortenedtoOGLEXX-YYY. OGLE13-144,andperhapsOGLE14-009arethemostII- (cid:13)c 0000RAS,MNRAS000,000–000 Bright but slow 3 Llike.Asweshowbelow,thesearealsothemostdeclining its explosion date. Shifting it back by ∼ 6 days would objects. make it consistent with the SNe II-L in our sample and We measure the ejecta velocities of all the SNe (ex- withthetemplatefromFaranetal.(2014b),thusclearing cept for OGLE14-009 which has no features to fit), and the gap further. use them in section 4. Traditionally, the FeII λ5169 ab- This gap is reminiscent of the results found by Ar- sorption line velocity, as measured in mid-plateau, is cavi et al. (2012), and contrasts with the findings of An- considered a good proxy for the velocity of the photo- derson et al. (2014), Faran et al. (2014b), and Sanders sphere(e.g.,Schmutzetal.1990;Dessart&Hillier2005). et al. (2015). The existence of such a gap, or its absence, Sincethespectrahavetypicallylowsignal-to-noiseratios could indicate whether there is a continuum in the prop- (S/N), and were often taken early, when the line has not erties of SNe II progenitors. However, often these differ- yet developed much, we measure the FeII velocity indi- ent works are difficult to compare because their samples rectly.UsingthesamemethodasPoznanskietal.(2009), wereobtainedindifferentbands(e.g.,V-bandforAnder- Poznanski, Nugent & Filippenko (2010) and Poznanski son et al. 2014, R-band for Arcavi et al. 2012). When we (2013), we cross correlate the spectra, focusing on the compare the OGLE light curves to the I-band templates areabluerofHα–dominatedinearlyspectrabyHβ,and of Faran et al. (2014b), it appears that the samples are laterspectrabytheFeIIline–withalibraryofhighS/N different. While the II-L SNe in our sample match the spectra for which the velocity of the λ5169 line has been template reasonably well, the SNe II-P have markedly measureddirectly.AsshownbyPoznanski,Nugent&Fil- shorter-duration plateau. We measure the plateau dura- ippenko (2010) and Faran et al. (2014a), the Hβ velocity tions following the definition from Faran et al. (2014a), is linearly related to the FeII velocity. The velocity from from the date of explosion to the phase at which there the cross-correlation and its uncertainty are then prop- isa0.5magdeclinefromaverageplateaumagnitude(the agated to day 50 past explosion, following Nugent et al. average is calculated between days 25 and 75). (2006),whoshowedthatphotosphericvelocitiesofSNeII- Most SNe II-P, including most of our sample, reach Pfollowatightpowerlawrelation.Weusetheimproved their peak luminosity quickly, within a week or so from determinationofthephasedependanceofthevelocityby explosion(seeFaranetal.2014aforarecentcompilation Faran et al. (2014a). We note that Faran et al. (2014b) ofrisetimes,aswellasValentietal.2014forarecentslow foundthatSNeII-Lhaveadifferent,slower,velocityevo- riser).However,OGLE13-048reachespeakbrightnesslate lution, with some scatter. As a result, our calculations in its evolution, about 20d from explosion. This is remi- probably underestimate the velocity for such SNe. niscentoftherareexplosionsofbluesupergiants,suchas thecanonicalSN1987AorSN2000cb(Kleiseretal.2011), butSN1987AreachedpeakI-bandmagnitudeabout80d 3 PHOTOMETRIC PROPERTIES afterexplosion,andSN2000cbtookroughly60d.In87A- likeexplosionsthelackofluminosityearlyonisattributed In Figure 2 we show the rest-frame I-band light curves to the small radius of the progenitor – the energy is ex- of the sample (Table 2 includes the photometry). The pended on expansion – and the later luminosity is dom- photometry derived by Wyrzykowski et al. (2014) is cor- inated by the decay of 56Ni. This explanation cannot be rected for galactic extinction using the maps of Schlafly summonedhere,since20dpastexplosionistooearlyfor & Finkbeiner (2011), is offset in time to match at the the 56Ni to peak (or would require the Ni to be located SNexplosionsdays,andK-correctedusingthespectraof mostly far in the outer ejecta, which is unreasonable). SN1999em to determine the correction at every phase. OGLE14-009, with its intrinsically featureless spec- Theexplosiondaysweredeterminedasthemidpointbe- tra, has a declining light curve. Out of caution, we com- tweenthefirstdetectionandthelastnon-detection.Since pareitslightcurvetoaSNIa,usingsyntheticphotometry thetargetcadenceofOGLEis5d,thetypicaluncertainty ontheIatemplatesofNugent,Kim&Perlmutter(2002) is 2d. We also fit a spline curve to each SN. One can see as in, e.g., Poznanski, Maoz & Gal-Yam (2007). We find in Figure 2 that the splines capture well the variability that while the brightness is broadly consistent, the light timescales of the various light curves. curve of OGLE14-009 is significantly broader than that While the SNe have a broad range of luminosities, of a typical SN Ia, requiring a stretch of about 60 per- spanningabouttwomagnitudes,thesampleismorenar- cent (which then does not fit the rising part of the light rowly distributed and the SNe are brighter than samples curve), or a ∆M ∼ 0.2. This is broader than the un- I,15 found in nearby galaxies, such as the samples recently usuallyslowlydecliningSN2001ay(Baronetal.2012).It discussedbyArcavietal.(2012),Andersonetal.(2014), isthereforeunlikelythatOGLE14-009isaSNIa(orany or Faran et al. (2014a). OGLE12-047, which has a single other SN type with a nickel driven light curve), given its earlyspectrumsimilartothesuperluminousSN2008esas spectra and photometry. It could still possibly be a SN discussedabove,isalsothebrightestobjectinoursample, IIn, as these can have wildly different light curves (e.g., withapeakmagnitudeofaboutM∼−19mag,indicating Miller et al. 2010), and featureless spectra at early times that these two SNe are somewhat similar. (e.g. SN 1998S, Fassia et al. 2001). If we also normalize the SNe to have the same peak magnitude, using the splines to determine the peak (a slight but essential modification to the recipe of Arcavi 4 CORRELATIONS et al. 2012 where SNe II-P and II-L were treated differ- ently a-priori), as seen in Figure 3, the range of decline WhetherornotthereisagapbetweenSNeII-LandSNe rates become apparent, and a minor gap may be seen to II-P, recent studies indicate that Type II SNe (barring emerge between the more or less declining SNe. Clearly, SNe IIb and IIn) form a one-parameter family. Brighter thethreedecliningSNe,OGLE13-047,OGLE13-144,and SNe have higher ejecta velocities, and more declining OGLE14-009,arealsothemostII-Llikespectroscopically, light-curves(e.g., Hamuy&Pinto2002;Poznanskietal. as discussed above. Furthermore, OGLE-13-011, which 2009;Andersonetal.2014;Sandersetal.2015).Further- somewhat fills that gap, has the largest uncertainty on more,Poznanski(2013)findsthatthebrighterSNecome (cid:13)c 0000RAS,MNRAS000,000–000 4 Poznanski et al. -19 -18.5 -18 -17.5 g] a m M [I -17 -16.5 OGLE13-005 OGLE13-011 -16 OGLE13-045 OGLE13-046 OGLE13-047 OGLE13-048 OGLE13-135 -15.5 OGLE13-144 OGLE14-004 OGLE14-009 OGLE14-018 -15 0 20 40 60 80 100 120 Days since explosion (rest-frame) Figure 2.AbsolutemagnitudelightcurvesoftheOGLEsample,aswellasourbestfittingsplinefits.Wecropthefigureat120d, the period most relevant to this study, though some SNe have light curves that extend beyond these limits. We use the complete lightcurvesforthesplinefitting. these findings with the numerical simulations of Dessart, Table 2.K-CorrectedRest-FramePhotometry Livne & Waldman (2010), it appears that brighter SNe should have shorter plateaus, down to about 80d for SNnamea Phase I ∆I stars above 20M . Therefore mass determines the en- (cid:12) OGLE14-018 1.47270 19.438 0.083 ergy, luminosity, velocity, plateau duration, and decline OGLE14-018 5.37956 18.653 0.022 rate.NotehoweverthatPoznanskietal.(2009)findthat OGLE14-018 8.25974 18.496 0.016 declining SNe do not obey the luminosity-velocity rela- OGLE14-018 12.12427 18.496 0.025 tionfoundbyHamuy&Pinto(2002),(andasmentioned OGLE14-018 25.70538 18.457 0.018 above, their velocity evolution might also differ) which OGLE14-018 35.29636 18.371 0.017 may indicate that they do not follow other scalings ei- OGLE14-018 39.20328 18.338 0.018 ther. OGLE14-018 41.15068 18.345 0.018 OGLE14-018 42.08987 18.339 0.024 InthetoppanelofFigure4thatincludesthesample OGLE14-018 43.06376 18.455 0.029 ofSNeII-PfromFaranetal.(2014a),aswellasourOGLE OGLE14-018 44.01478 18.384 0.022 sample, one can see that indeed there is a weak correla- OGLE14-018 45.95610 18.341 0.019 tion between plateau duration and magnitude, such that OGLE14-018 88.60024 18.540 0.041 only brighter events can have short plateaus, but there OGLE14-018 100.24717 19.107 0.065 doesnotseemtobeapopulationoffaintSNewithshort OGLE14-018 182.25189 20.740 0.167 plateaus. However, in the lower panel one can see that a OGLE14-009 1.44567 19.641 0.100 significant fraction on the OGLE SNe have low velocity, OGLE14-009 3.36477 19.482 0.079 andshortplateaus,contrarytotheexpectationsfromthe OGLE14-009 4.30878 19.425 0.065 one dimensional picture above. OGLE14-009 6.19501 19.202 0.049 In figure 5 we examine the correlation between peak OGLE14-009 9.03278 19.169 0.057 magnitudeanddeclinerate,comparingtothefindingsof OGLE14-009 13.74580 19.185 0.073 Anderson et al. (2014). There are two difficulties when OGLE14-009 17.52362 19.335 0.081 comparingthesesamples.First,sinceourlightcurvesare OGLE14-009 19.41935 19.295 0.061 not very well sampled we cannot differentiate between OGLE14-009 22.25030 19.471 0.069 OGLE14-009 24.97752 19.470 0.096 various phases these authors define. Instead we find the ... decline rate by asking at what phase td the spline curve that was fit to every object crosses 0.5mag. The decline is then 50/t in units of mag/100d, where the uncer- aFulltableinonlineversion. d tain explosion date dominates the uncertainty. Secondly, our light curves are in I-band, while Anderson et al. frommoremassiveprogenitors(seealsoSmartt2009)that (2014) only studied V-band data. Using the templates have had a much greater energy deposited by the explo- from Faran et al. (2014b), we find that SNe II have a sion in the envelope, so that the energy E scales as M3, peak color of V −I = 0.4−0.7, and SNe II-L decline whereMistheinitialmassoftheprogenitor.Combining about twice as faster in V than in I. For this qualita- (cid:13)c 0000RAS,MNRAS000,000–000 Bright but slow 5 tive comparison we therefore apply an offset of 0.5mag akintothefallingoftheplateaureinforcessuchapicture toourpeakbrightness,andscaleourdeclinesbyafactor where the early light curve is dominated by the bulk of of 2. Our objects seem in agreement with the sample of the envelope: its profile, and its composition. See similar Anderson et al. (2014). recent observations by Valenti et al. (2015). In figure 6 we study the ‘Hamuy & Pinto’ velocity- OGLE13-047 reaches −19mag, about 40 times luminosityrelation.Wecompareoursampletothesample brighterthanunder-luminousSNeII-PsuchasSN2005cs. compiledbyPoznanskietal.(2009)–whichincludesdata Assumingthephysicsdrivingthelight-curvesofSNeII-L from Hamuy & Pinto (2002) and Nugent et al. (2006) as and II-P are similar, This large span is hard to recon- well as a subset of the SNe in Faran et al. (2014a) and cilewiththeanalyticalfindingsofPopov(1993),ormore Faran et al. (2014b)2, and the SDSSII sample from Poz- recentlyGoldfriend,Nakar&Sari(2014),wherethelumi- nanski, Nugent & Filippenko (2010), which is a reanal- nosity during the photospheric phase is found to be only ysis of the sample of D’Andrea et al. (2010). The line weakly dependent on the various parameters. The range showsthebestfitluminosity-velocityrelation,asderived observed would require a deposited energy 100 times by Poznanski et al. (2009). Clearly the OGLE SNe are larger in luminous events, radii 1000 times larger, wildly allover-luminous,orhaveslowejecta,whencomparedto different opacities, or any combination thereof. However, most of these samples, as all of the objects lie beneath iftheenergystronglydependsontheprogenitormass,as the line, similarly to the SDSSII sample. Also, it seems foundbyPoznanski(2013),oneobtainsaluminositythat that the decline rate is a weak indicator of fit quality as depends quadratically on the mass, alleviating some of declining objects (from either samples) do not seem par- thedifficultytoexplainthisrangeofluminosities,though ticularlyscatteredorbiased,thoughiftakenindividually, not all. theydonotseemtofollowtherelationatall.Oneshould We find 2 or 3 SNe II-L, out of a sample of 11. A bear in mind that this plot does not take into account II-L fraction of ∼ 30 percent could be surprising con- any color information – a tracer of dust extinction and sidering their scarcity in nearby searches (Li et al. 2011; intrinsic variance – that is typically used to reduce the Faranetal.2014b),butthedifferencebetweentheOGLE scatter in such diagrams. All luminosities here are under sample and nearby samples can be explained as stem- the assumption of no significant extinction in the host ming from the different selection biases influencing dif- galaxy. ferent SN searches. Nearby searches are more complete Surprisinglythough,excludingthedecliningobjects, in luminosity, sensitive to much fainter SNe. The effec- theOGLEandSDSSIISNedoseemtohavealuminosity- tivesurveyvolumeforSNenear−15mag(likeSN2005cs velocitycorrelation,albeititisoffsetfromtheonederived for example), is about 250 times smaller than for SNe from nearby samples. While this could be a dust bias – reaching−19mag.Magnitudelimitedsurveysareseverely nearbysampleshavemoredustySNewhichappearfainter Malmquistbiased,findingthebrightestobjectsofagiven onthisdiagram–aswasshowntobepossibleforSDSSII distribution, even when intrinsically rare. Furthermore, sample(Poznanski,Nugent&Filippenko2010),wedonot nearby searches are typically focused on massive, lumi- have color information for the OGLE sample to test this nous, star-forming, spiral galaxies, preferring SNe that hypothesisfully.WedohoweversearchforNaIDabsorp- would occur in such metal rich environments. This could tioninallofthespectraandfindnone(thoughthisisonly also change the ratio of SN types (Arcavi et al. 2010). a weak indicator of extinction, as shown by Poznanski The OGLE sample, like the SDSSII sample from et al. 2011, and our S/N is typically low). As mentioned D’Andrea et al. (2010) before it, is therefore biased to- before, the velocity derivation, using the power-law be- wards brighter SNe that have shorter plateaus (or an ac- havior found by Nugent et al. (2006), was shown not to tualdecline),buttheirvelocitiesdonotmatchtheexpec- be applicable to SNe II-L by Faran et al. (2014b). These tations from nearby searches. The sample from Sanders SNe appear to evolve more slowly, so that their velocity etal.(2015)seemssimilarlybiased.IntheirFigure14one here might be underestimated. Correcting for this possi- canseethattheirSNepeakbrighter,around–18mag,de- ble bias would somewhat increase their velocity. pendingontheband,anddeclinerapidly,asseenintheir Figure11.Infact,themajorityoftheirSNewouldbeSNe II-L by the Li et al. (2011) criterion of 0.5 mag/50d in R.ThebrightestSNeinthesamplefromAndersonetal. 5 DISCUSSION (2014)alsohaveshortplateaus,drivingtheirmeancloser Examining the light curve of OGLE13-047, our brightest to ours. andbestobservedSNII-L,onecanseethatafteraperiod ofdeclineofabout80d,itgoesthroughasecond,sharper, drop,akintothefalling-offthephotosphericphaseofSNe 6 CONCLUSIONS II-P. Since it it also brighter than typical SNe II-P, it Having analyzed a sample of 11 SNe from the OGLE-IV couldbeperhapsexplainedwithasimilarmodelbutwith survey, we find that two or three of them are SNe II- a additional energy source that supplies an extra hump L. They are distinct both in their decline and in their on top of the plateau. Alternatively, a different profile of spectroscopic properties as previously suggested. the stellar envelope before the explosion could perhaps The 8 SNe II-P while rather standard in most re- account for it. Recently Goldfriend, Nakar & Sari (2014) spects,aremoreluminousthantypicallyfoundbynearby showed that the shape of a Type II light curve depends searches, have shorter-duration plateaus, and rise times strongly on the mass profile, with the plateau resulting thatcanbeaslongas20d,morethandoublethetypical from a somewhat fortuitous coincidence with hydrogen timescale. recombination of a narrow range of profiles. A late drop Wethereforefindthatasingleparametercannotex- plain the range of outcomes from massive hydrogen-rich 2 SN2002hh was omitted due to its abnormally high extinc- core collapse events. While the diversity for a subset of tion. thesecanbeshowntofollowmass,searchesthataremore (cid:13)c 0000RAS,MNRAS000,000–000 6 Poznanski et al. 0 OGLE SNe IIP (I-band adjusted) OGLE SNe IIL (I-band adjusted) de 5 Anderson et al. (2014) u nit 0.5 mag OGLE13-005 ys] 4 d OGLE13-011 da Shifted I-ban 1 .15 OOOOOOOGGGGGGGLLLLLLLEEEEEEE11111113333334-------000011044443405678544 ate [mag/100 23 OGLE14-009 R OGLE14-018 e 2 0 20 40 60 80 100 120 eclin 1 D Days since explosion (rest-frame) 0 Figure 3. Light curves normalized at peak luminosity com- paredtotheI-bandtemplatesofFaranetal.(2014b).Wefind 3SNethatdeclineabout0.5magin50d,whiletherestofthe -1 samplehasamarkedplateau,albeitashortone. -13 -14 -15 -16 -17 -18 -19 M (V-band) [mag] max −19 −18.5 Figure 5. Correlation of peak magnitude with decline rate. ag) −18 Grey points from Anderson et al. (2014), blue (green) crosses m are the SNe II-P (II-L) from OGLE. In order to compare our e(−17.5 datatotheV-bandmagnitudesinAndersonetal.(2014),we d gnitu −17 pmeualktipmlyagonuirtuId-ebabnyd0d.5ecmliange,rbaatseesdbyontwthoe, wtehmilpelaotffessetotfinFgartahne a−16.5 M etal.(2014b). nd −16 a B−15.5 I −15 N. Sanders for comments on this manuscript. We further thank the referee for their review of this work. −14.5 7 D.P. acknowledges support from the Alon fellowship foroutstandingyoungresearchers,andtheRaymondand 6 Beverly Sackler Chair for young scientists. This work −1)5 waspartiallysupportedbythePolishMinistryofScience s andHigherEducationthroughtheprogram“IdeasPlus” m k award No. IdP2012 000162. 3104 This research made use of the Weiz- (,h503 mann interactive supernova data repository p v (www.weizmann.ac.il/astrophysics/wiserep), as 2 well as the NASA/IPAC Extragalactic Database (NED) which is operated by the Jet Propulsion Laboratory, 1 California Institute of Technology, under contract with 50 60 70 80 90 100 110 120 130 140 NASA. This work is based on observations collected at PlateauLength(days) theEuropeanOrganisationforAstronomicalResearchin Figure 4.Correlationofvelocity(bottompanel)andI-band the Southern hemisphere, Chile as part of PESSTO (the magnitude (top) with plateau duration. Black points from Public ESO Spectroscopic Survey for Transient Objects Faran et al. (2014a), red points from the numerical models Survey) ESO programme ID 188.D-3003. The research ofDessart,Livne&Waldman(2010)assumingE∝M3 (Poz- leading to these results has received funding from nanski2013),blue(green)crossesaretheSNeII-P(II-L)from the European Union Seventh Framework Programme OGLE. While there is a weak correlation between brightness (FP7/2007- 2013) under grant agreement no. 264895. andplateauduration,thesamedoesnotapplytothevelocity. sensitive to bright SNe find objects that behave differ- ently.Acompletecensusofcorecollapseshouldtherefore REFERENCES rely on a careful combination of samples from both mag- Anderson J. P. et al., 2014, ArXiv:1403.7091 nitude and volume limited surveys, or a very wide un- Arcavi I. et al., 2012, ApJ, 756, L30 targeted search, such as the Palomar Transient Factory Arcavi I. et al., 2010, ApJ, 721, 777 (PTF; Law et al. 2009) that can also find a substantial Baron E., Ho¨flich P., Krisciunas K., Dominguez I., number of faint events and account for them properly. Khokhlov A. M., Phillips M. M., Suntzeff N., Wang L., 2012, ApJ, 753, 105 Blondin S., Tonry J. L., 2007, ApJ, 666, 1024 ACKNOWLEDGMENTS Bruenn S. W. et al., 2014, ArXiv:1409.5779 WethankthePESSTOcollaborationandJ.Prietoforthe D’Andrea C. B. et al., 2010, ApJ, 708, 661 spectrausedinthispaper,andJ.Anderson,I.Arcavi,and Dessart L., Hillier D. J., 2005, A&A, 439, 671 (cid:13)c 0000RAS,MNRAS000,000–000 Bright but slow 7 4 P09 II-P 3.9 P09 II-L 3.8 SDSSII OGLE II-P 3.7 OGLE II-L L∝v2 3.6 0 5 v g o l3.5 3.4 3.3 3.2 3.1 −13 −14 −15 −16 −17 −18 −19 MI Figure6.Luminosity-velocityrelation.SamplesfromPoznanskietal.(2009)inred(SNeII-Linblack),fromSDSSII(Poznanski, Nugent&Filippenko2010)ingrey,andOGLEinblue(SNeII-Lingreen).TheSNefrommagnitudelimitedsurveys(SDSSIIand OGLE)areclearlyoffset,thoughtheymayfollowtheirownluminosity-velocityrelation. Dessart L., Livne E., Waldman R., 2010, MNRAS, 408, Smartt S. J. et al., 2014, ArXiv:1411.0299 827 Spiro S. et al., 2014, MNRAS, 439, 2873 Faran T. et al., 2014a, MNRAS, 442, 844 Valenti S. et al., 2014, MNRAS, 438, L101 Faran T. et al., 2014b, MNRAS, 445, 554 Valenti S. et al., 2015, ArXiv e-prints Fassia A. et al., 2001, MNRAS, 325, 907 WyrzykowskiL(cid:32).etal.,2014,ActaAstronomica,64,197 Filippenko A. V., 1997, ARA&A, 35, 309 Yaron O., Gal-Yam A., 2012, PASP, 124, 668 GoldfriendT.,NakarE.,SariR.,2014,ArXiv:1404.6313 Guti´errez C. P. et al., 2014, ArXiv:1403.7089 Hamuy M., Pinto P. A., 2002, ApJ, 566, L63 Kleiser I. K. W. et al., 2011, MNRAS, 415, 372 Kozl(cid:32)owski S. et al., 2013, Acta Astronomica, 63, 1 Law N. M. et al., 2009, PASP, 121, 1395 Li W. et al., 2011, MNRAS, 412, 1441 Maguire K. et al., 2012, MNRAS, 420, 3451 Miller A. A. et al., 2010, MNRAS, 404, 305 NugentP.,KimA.,PerlmutterS.,2002,PASP,114,803 Nugent P. et al., 2006, ApJ, 645, 841 Planck Collaboration et al., 2014, A&A, 571, A16 Popov D. V., 1993, ApJ, 414, 712 Poznanski D., 2013, MNRAS, 436, 3224 Poznanski D. et al., 2009, ApJ, 694, 1067 Poznanski D., Ganeshalingam M., Silverman J. M., Fil- ippenko A. V., 2011, MNRAS, 415, L81 PoznanskiD.,MaozD.,Gal-YamA.,2007,AJ,134,1285 PoznanskiD.,NugentP.E.,FilippenkoA.V.,2010,ApJ, 721, 956 Prieto J. L., Morrell N., 2013, The Astronomer’s Tele- gram, 4734, 1 RichmondM.W.,TreffersR.R.,FilippenkoA.V.,Paik Y., Leibundgut B., Schulman E., Cox C. V., 1994, AJ, 107, 1022 Sanders N. E. et al., 2015, ApJ, 799, 208 SavitzkyA.,GolayM.J.E.,1964,AnalyticalChemistry, 36, 1627 Schlafly E. F., Finkbeiner D. P., 2011, ApJ, 737, 103 Schlegel E. M., 1996, AJ, 111, 1660 Schmutz W., Abbott D. C., Russell R. S., Hamann W.- R., Wessolowski U., 1990, ApJ, 355, 255 Smartt S. J., 2009, ARA&A, 47, 63 (cid:13)c 0000RAS,MNRAS000,000–000