The death ofm assive stars -I.O bservationalconstraints on the progenitors oftype II-P supernovae

A B ST R A C T W epresenttheresultsofa 10.5yr,volum elim ited (28M pc)search forsupernova (SN)progenitorstars.In doing so we com pile allSNe discovered within thisvolum e (132,ofwhich 27% aretypeIa)and determ inetherelativeratesofeach sub-typefrom literature studies.The core-collapse SNe break down into 59% II-P and 29% Ib/c, with the rem ainderbeing IIb (5% ),IIn(4% )and II-L (3% ).There havebeen 20 II-P SNewith high quality opticalornear-IR pre-explosion im agesthatallow am eaningful search for the progenitor stars.Inve cases they are clearly red supergiants,one case isunconstrained,two fallon com pactcoevalstarclustersand the othertwelve haveno progenitordetected.W ereview and updatealltheavailabledata forthehost galaxies and SN environm ents (distance,m etallicity and extinction) and determ ine m assesand upperm assestim atesforthese20progenitorstarsusingtheSTARS stellar evolutionarycodeandasingleconsistenthom ogeneousm ethod.A m axim um likelihood calculation suggeststhatthem inim um stellarm assfora typeII-P to form is m m in = 8:5 + 1:5 Mand the m axim um m ass for II-P progenitors is m m ax = 16:5 � 1:5M � , assum ing a Salpeterinitialm assfunction holdsforthe progenitorpopulation (in the range = 1:35 + 0:3 0:7 ).Them inim um m assisconsistentwith currentestim atesforthe upperlim ittowhitedwarfprogenitorm asses,butthem axim um m assdoesnotappear consistent with m assive star populations in LocalGroup galaxies.Red supergiants in the LocalGroup have m asses up to 25Mand the m inim um m ass to produce a W olf-Rayet star in single star evolution (between solar and LM C m etallicity) is sim ilarly 25-30M � .The reason we have not detected any high m ass red supergiant progenitorsabove17Misunclear,butweestim ate thatitisstatistically signicant at2.4� condence.Two sim ple reasonsforthiscould be thatwehavesystem atically underestim ated the progenitor m asses due to dust extinction or that stars between 17-25MproduceotherkindsofSNewhich arenotII-P.W ediscussthesepossibilities andnd thatneitherprovidesa satisfactory solution.W e term thisdiscrepancy the \red supergiantproblem " and speculatethatthesestarscould havecorem asseshigh enough to form black holesand SNe which are too faintto have been detected.W e com pare the 56 Nim asses ejected in the SNe to the progenitor m ass estim ates and �nd thatlow lum inosity SNe with low 56 Niproduction are m ostlikely to arise from