The host galaxy of the super-luminous SN 2010gx and limits on explosive nickel-56 production

0.24 MB·English

by  Ting-Wan Chen

Draftversion October 16,2012
PreprinttypesetusingLATEXstyleemulateapjv.5/2/11
THE HOST GALAXY OF THE SUPER-LUMINOUS SN 2010GX AND LIMITS ON EXPLOSIVE 56NI
PRODUCTION
Ting-Wan Chen1; Stephen J. Smartt1; Fabio Bresolin 2; Andrea Pastorello3; Rolf-Peter Kudritzki 2; Rubina
Kotak 1; Matt McCrum 1; Morgan Fraser 1; Stefano Valenti 4,5
Draft versionOctober 16, 2012
ABSTRACT
2
Super-luminous supernovae have a tendency to occur in faint host galaxieswhich are likely to have
1
low mass and low metallicity. While these extremely luminous explosions have been observed from
0
z = 0.1 to 1.55, the closest explosions allow more detailed investigations of their host galaxies. We
2
present a detailed analysis of the host galaxy of SN 2010gx(z =0.23), one of the best studied super-
ct luminous supernovae. The host is a dwarf galaxy (Mg = −17.42±0.17) with a high specific star
O formation rate. It has a remarkably low metallicity of 12+log(O/H)= 7.5±0.1 dex as determined
from the detection of the [OIII] λ4363˚A line. This is the first reliable metallicity determination of

5 a super-luminous supernova host. We collected deep multi-epoch imaging with Gemini + GMOS
1 between200−550daysafterexplosiontosearchforanysignofradioactive56Ni, whichmightprovide
furtherinsightsontheexplosionmechanismandtheprogenitor’snature. Wereachgriz magnitudesof
] m ∼26, but do not detect SN 2010gxat these epochs. The limit implies that any 56Ni production
O AB
wasbelowthatofSN1998bw(aluminoustypeIcSNthatproducedaround0.4M⊙ of56Ni). The low
C volumetric rates of these supernovae (∼ 10−4 of the core-collapse population) could be qualitatively
h. matched if the explosion mechanism requires a combination of low-metallicity (below 0.2Z⊙), high
p progenitor mass (>60M⊙) and high rotation rate (fastest 10% of rotators).
- Subject headings: Supernovae: general — Supernovae: individual (SN 2010gx)
o
r
t
s 1. INTRODUCTION shock-interaction of the SN ejecta with massive C-O
a
shells (Blinnikov & Sorokina 2010).
[ Recently, the new generation of wide-field sur-
vey projects, such as the Panoramic Survey Tele- Neill et al. (2011) showed that most SLSNe tend to
1 scope And Rapid Response System (Pan-STARRS- occur in star-forming dwarf galaxies, which have high
v specific star formation rates. Their intrinsic rate is
1) (Kaiser et al. 2010) and the Palomar Transient
7 10−8Mpc−3year−1; or 10−4 of the normal core-collapse
Factory (PTF) (Rau et al. 2009), have carried out
2 SN rate (Quimby et al. 2011; Gal-Yam 2012). However
searchesforsupernovaewithoutaninherentgalaxybias.
0 beyond this we lack information on their explosion sites
Quimby et al. (2011) identified a class of hydrogen-
4 andwhattheprogenitorstarpopulationmaybe. Anim-
poor “super-luminous supernovae (SLSNe)” with typ-
0. ical g−band absolute magnitudes of around −21.5 , portant question is whether or not the apparent prefer-
1 which is 10 to 100 times brighter than normal core- encetheyhavefordwarfhostgalaxiesisrelatedtorequir-
2 collapse SNe. Their intrinsic brightness has allowed ing low metallicity progenitor stars. As low metallicity
1 them to be discovered in the Pan-STARRS1 survey be- affects stellar structure and evolution through changes
v: tween z = 1−1.55 (Chomiuk et al. 2011; Berger et al. in stellar winds, opacity and rotation, quantitative mea-
surements are required. In this letter we study the host
i 2012). Pastorello et al. (2010) studied one of the clos-
X est examples, SN 2010gx (z = 0.23), in detail show- galaxy of one of the nearest SLSNe, SN 2010gx, and
searchforevidenceoflateemissionfromradioactive56Ni
r ing it to transition to a normal type Ic SN at 40
a days after peak, while multi-colour photometry rules decay.
out 56Ni as powering the peak magnitude. Two phys-
2. OBSERVATIONS
ical mechanisms for SN 2010gx(and the supernovae like
it from Quimby et al.; Chomiuk et al.) have been pro- We used Gemini South and North Observatories with
posed. One is that the type Ic SN is boosted in lumi- Gemini Multi-Object Spectrograph (GMOS) to collect
nosity by energy deposition from magnetar spin down griz photometry between 200−550 days after the max-
(Kasen & Bildsten2010). Thesecondistheshockbreak- imum of SN 2010gx (Table1). To check for the pres-
out through a dense wind (Chevalier & Irwin 2011), or ence of residual supernova flux at 200−350 days after
peak, we aligned and rescaled all images to the same
1Astrophysics Research Centre, School of Mathematics and pixelgridusingIRAF/geomapandgeotran packages
Physics,Queen’sUniversityBelfast,BelfastBT71NN,UK and made the assumption that the final image taken in
2Institute for Astronomy, 2680 Woodlawn Drive, Honolulu,
January 2012 only contains flux from the host galaxy.
HI96822, USA
3INAF-Osservatorio Astronomico di Padova, vicolo We performed imaging subtraction using the High Or-
dell’Osservatorio5,35122Padova,Italy derTransformofPSFANdTemplateSubtraction(hot-
4Las Cumbres Observatory Global Telescope Network, Inc. pants) software 6, to subtract the January 2012 images
SantaBarbara,CA93117,USA
5Department of Physics, University of California Santa
Barbara,SantaBarbara,CA93106-9530, USA 6http://www.astro.washington.edu/users/becker/hotpants.html
2 Chen et al.
from the earlier epochs. The host galaxy subtracts off lines) of ∼ 6˚A in the red, and ∼ 4.5˚A in the blue. The
almost perfectly and no SN signalis detected. To deter- combinedblueandredspectraprovidespectralcoverage
mine a limiting magnitude for the SN flux in the three between4300and8500˚Aandtheoverlapregionwasused
epochs, we added 6 fake stars, one at the supernova po- to ensure a uniform flux calibration. Three strong lines
sitionandfive within a surroundingradiusof0.6 arcmin ([OIII] λ5007,[OIII] λ4959,andHβ) lie within the over-
toeachofthe 200−350daysimages. We ranthis proce- lap region. We measured the line fluxes in both spectra
duremultiple timeswitharangeoffakestarmagnitudes and applied a linear scaling to bring the blue spectrum
andrepeatedimagesubtractionwithhotpants. Wede- into agreement with the red. The line fluxes of these
termined5σand3σdetectionlimitsbyrequiringthatwe three lines agreed to within 3% after this rescaling.
couldvisuallydetectafakeSNexactlyatthe positionof Linefluxmeasurementsweremadeafterfittingapoly-
SN 2010gxand that the measured standarddeviation of nomial function to subtract the continuum, and Fig.3
the set of fake stars added was 0.2 and 0.3 magnitudes shows the final combined spectrum of the host galaxy
respectively. The resulting upper limits are in Table1. with this continuum subtracted. We fitted Gaussian
The apparent magnitudes (SDSS AB system) of the lineprofileswithinacustombuiltIDLenvironment,fix-
host galaxy,SDSS J112546.72-084942.0,are g =23.71± ing the full-width-half-maximum (FWHM) of the mean
0.14, r = 22.98 ± 0.10, i = 22.90 ± 0.13 and z = of the three strong lines; 6.03˚A and 4.45˚A for the red
22.98±0.16. These were obtained from aperture pho- and blue spectra respectively (see Table2). We further
tometry (within IRAF/daophot) on the best seeing normalized the spectrum and determined the equivalent
images, using a zero-point calibration achieved with 10 widthofeachfeature. Theuncertaintyestimationis cal-
SDSS reference stars. As deep, pre-explosion images of culatedusingtheexpressionfromP´erez-Montero& D´ıaz
the host galaxy were not available, we compared our (2003),derivedfromtheequivalentwidthofthelineand
photometry results to both SDSS catalog magnitudes, the rms of the continuum (also listed in Table2).
and our own photometry performed on the SDSS im-
ages. In the g−band, the SDSS catalog Petrosian mag- 3. ANALYSIS
nitude for the object is 22.8 ± 0.5, while we measure 3.1. 56Ni mass estimation for SN 2010gx
22.6±0.6magintheSDSSimage. Thelargeuncertainty
The fact that SN 2010gx was not detected after the
is due to the marginal detection. Our deep Gemini im-
host galaxy image subtraction at epochs between +226
age shows the host resolved from a fainter neighbouring
and+334daysallowsus todetermine anupper limitfor
galaxy (which is unresolved from the host in SDSS). If
we use a large aperture on our Gemini g−band image, the mass of 56Ni ejected. This is somewhat uncertain
as it relies on assuming that the gamma ray trapping in
toincludefluxfromtheneighbouringgalaxythenweob-
taing =23.29±0.16mag. Within the uncertainties,the SN 2010gx behaves like in other Ic SNe in the nebular
phase,andthelateluminosityisdirectlyproportionalto
SDSS and Gemini magnitudes are similar and there is
the mass of 56Co present at any epoch.
no hint that the +550 days Gemini images are brighter
We measured SN 2010gx fluxes at the effective rest
than the SDSS images and contain SN flux.
wavelength of the ri− filters and used a luminosity dis-
Long slit spectroscopy of SN 2010gx was carried out
tance of 1114.7Mpc (m = 40.24). We determined
on 5 June 2010 (Pastorello et al. 2010) with the Gemini u
the upper limit of the luminosities and compared with
SouthTelescope+GMOS.TheB600grating(IDG5323)
SN 1998bw(Clocchiatti et al. 2011), a Ic SN with a high
was used (1800s exposure) to obtain a spectral coverage
in the observer frame of 4300−6400˚A (henceforth re- 56Nimass,inJohnsonV andCousinsRintherestframe
(see Fig.2). The most interesting measurement is the
ferred to as the “blue spectrum”). Although this spec- i−band limit for SN 2010gx flux at +226 days, which is
trum contains flux from SN 2010gx, we fitted a high or-
significantly below what we would expect if SN 1998bw
derpolynomial(order15)throughtheSNcontinuumand
subtracted off this contribution leaving a flat spectrum wereat this distance. If we take 0.4M⊙ as the represen-
tative 56Ni mass for SN 1998bw (McKenzie & Schaefer
for which emissionlines between 3500−6000˚A could be
1999;Sollerman et al.2000;Maeda et al.2006),thenthe
measured. WecouldnotmeasuretheHαlineinthisspec-
flux limit of SN 2010gx (0.8 that of SN 1998bw; Fig.2),
trumasitwascontaminatedbyastrongskylineandslit
nodding was not employed. A second spectrum was ob- suggests an upper limit of about 0.3M⊙ of 56Ni ejected
by SN 2010gx.
tainedon23Dec. 2011withGMOSattheGeminiNorth
Telescope, using the R400 grating (ID G5305; 3×1800s 3.2. Host galaxy metallicity and other properties
exposures), to cover 5070−8500˚A (henceforth referred
We carried out an abundance analysis of the host
to as the red spectrum). The use of an order-blocking
galaxy based on the determination of the electron tem-
filter (GG455) truncated the wavelengthcoveragein the
perature, and detection of the [OIII] λ4363 auroral line
blue. Detrending of the data, such as bias-subtraction,
(see Bresolin et al. 2009 for further details). Firstly, the
wasestablishedusingstandardtechniqueswithinIRAF.
emission line fluxes were reddening corrected using the
For the red spectra, individual exposures were nodded Seaton (1979) Milky Way law of A =3.17×E(B−V).
along the slit, and we used two-dimensional image sub- V
We assumed case B recombinationwhich requires an in-
tractiontoremovetheskybackground. Thespectrawere
trinsic line ratio of Hα/Hβ = 2.86 and Hγ/Hβ = 0.47
wavelength calibrated using daytime arcs (CuAr lamp)
at T = 104K (Osterbrock 1989). The observed ra-
andfluxcalibratedwiththespectrophotometricstandard e
tio of Hα/Hβ = 3.88 yields an extinction coefficient
Feige 34.
For both red and blue spectra, a 1.5′′ slit was used, c(Hβ)= 0.48, which we used to calculate reddening-
correctedline fluxes. The observedline fluxesanderrors
yieldingaresolution(asmeasuredfromnarrownightsky
are listed in Table2.
SN 2010gx host 3
We confidently detected the [OIII] λ4363 auroral line ing a Salpeter initial mass function with red horizontal
(S/N∼ 6) in the GMOS spectrum (see Fig.3), allowing branch morphology. The measured colours give reason-
ustoestimatetheelectrontemperatureandcalculatethe able agreement with a population model of age between
oxygen abundance directly. As in Bresolin et al. (2009), 20 to 30 Myr, and a stellar mass around 6.17×107M⊙.
we used a custom-written pyraf script based on the We also employed the MAGPHYS galaxy SED models
IRAF/nebula package. The direct method of estimat- ofda Cunha et al.(2008). Fig.1showstheobservedSED
ingoxygenabundancesusestheratiooftheintensitiesof andthe redshiftedMAGPHYSbest-fit model spectrum
[OIII]λ4959,λ5007/[OIII]λ4363linestodeterminatethe : totalstellarmassof1.58×108M⊙ andhence aspecific
electron temperature (T ) in the ionized gas region. We star formation rate (sSFR) of 2.59×10−9year−1.
e
findanoxygenabundanceof12+log(O/H)=7.46±0.10
4. DISCUSSION ANDCONCLUSION
dex, remarkably low even for dwarf galaxies (see Fig.3).
Assumingasolarvalueof8.69±0.05dex(Asplund et al. It has already been demonstrated that the peak lumi-
2009), this implies that the host galaxy is 1.2 dex below nosity of super-luminous SNe such as SN 2010gx can-
solar abundance or just 0.06Z⊙. notbe due to 56Ni (Pastorello et al. 2010;Quimby et al.
The reddening-corrected, and z−corrected flux of Hα 2011; Chomiuk et al. 2011). Chomiuk et al. illustrated
is 3.49 × 10−16 (ergs−1cm−2), from which we deter- that applying Arnett’s rule (Arnett 1982), results in an
mined the star formation rate (SFR) of the host to be unphysicalsolution, with the massof 56Ni exceeding the
0.41M⊙year−1fromthecalibrationofKennicutt(1998): total C+O ejecta mass. Our search for a luminous tail
SFR (M⊙year−1)= 7.9× 10−42 L(Hα) (ergss−1). We phase in SN 2010gx was not motivated by what powers
also estimated the SFR from the [OII] λ3727 line flux, the peak luminosity, rather it is to determine if there is
whichgives 0.36±0.10M⊙year−1 (Kennicutt 1998) and any sign of radioactive 56Co at this stage and any simi-
larity to 56Ni-rich Ic SNe such as SN 1998bw. The only
is consistentwith Hα measurement. Berger et al.(2012)
robustconclusionwecandrawisthatthereisunlikelyto
used [OII] λ3727 to estimate the SFR of the ultra-
be any excessive mass of 56Co above and beyond about
luminous PS1-12bam host, which at z = 1.566 has Hα
shiftedtotheNIR.TheagreementbetweentheSFResti- 0.3M⊙. At a similar phase, SN 2010gx is fainter than
SN 1998bw as shown in Fig.2. This limit on the late
matesfrom[OII]λ3727andHαforthehostofSN2010gx
luminosity will also allow restrictions on the magnetar
supports [OII] λ3727 being reliable for these types of
models of Kasen & Bildsten (2010) when put in context
galaxies at higher redshift. We detected the galaxy ISM
absorption of the Mg II λλ2796/2803˚A doublet in the with other super-luminous supernovae of similar types
(Inserra et al. in prep).
Geminispectrumof1Apr. 2010(Pastorello et al.2010).
Our detection of the [OIII] λ4363 auroral line gives
Theequivalentwidthsare1.6±0.2and2.6±0.3˚Arespec-
confidence to our measurement of the remarkably low
tively, giving a ratio W2796/W2803 = 0.6. The strength metallicityof12+log(O/H)=7.46±0.10dex(0.06Z⊙).
of the line λ2803 component is comparable with that
In Fig.3 we show a compilation of other dwarf galax-
seentowardGRBlinesofsiteandsomewhathigherthan
ies with the [OIII] λ4363 direct method measurements,
observed for PS1-12bam (Berger et al. 2012). The fact
including four GRB hosts from Kewley et al. (2007).
that the line ratio is significantly smaller than 2 (the
The host of SN 2010gx is on the lower end and is
ratio of their oscillator strengths) suggests we are not
amongst the lowest metallicity dwarf galaxies known. It
on the linear regime, nor the square root regime on the
has been clear from the early discoveries of these SNe
curve of growth which prevents further use of the lines
at z < 1 (Pastorello et al. 2010; Quimby et al. 2011;
as physical diagnostics. Nevertheless the absorption line
Chomiuk et al. 2011), that the hosts are faint dwarf
strengths are similar to PS1-12bamand GRB sightlines.
galaxies and Neill et al. (2011) showed that they have
The line centroids give identical redshifts (z =0.231) to
high sSFRs. If the extremely low value we determine is
the[OIII]emissionlines(z =0.230)detectedinthesame
a common factor amongst these SNe then it will be a
spectrum, giving reassurance that the Mg II absorption
major constraint on the progenitor channel.
does arise in the host.
ThemagnetarscenariosuggestedbyKasen & Bildsten
The g−band absolute magnitude of the host galaxy is
(2010)toexplainthese SNemaybenefitfromlowmetal-
M =−17.42±0.17(adoptingaHubbleconstantofH =
g 0 licity progenitors. Massive stars rotate more rapidly at
72kms−1Mpc−1, Ω = 0.3, Ω =0.7, to obtain the lu-
M λ SMC metallicity (0.2Z⊙) than solar (Hunter et al. 2008
minosity distance 1114.7 Mpc; Wright 2006), applying
; Martayan et al. 2007), although 0.06Z⊙ is unexplored
the MilkyWay extinctionof0.13(Schlafly & Finkbeiner
territory. Evolutionary models of rotating stars have fo-
2011)), a k-correction 0.27 ± 0.08 (Chilingarian et al.
cusedattentiononlowmetallicityenvironmentsasmass-
2010) and a correction for internal dust extinction of
loss is reduced (Mokiem et al. 2007) hence angular mo-
∼0.5mag. Thelattercomesfromthemeasuredintrinsic
mentum loss is lower(Maeder & Meynet 2000). The en-
Hα/Hβ ratioandassumingR =3.16(applicableforan
V ergy input required from a magnetar requires the neu-
LMC environment ; Pei 1992)7 tron star to be both highly magnetic (B ∼ 1014G) and
The observed photometric g − r, r − i colour rotating rapidly at formation (P ∼ 2−20ms). It must
i
terms were corrected for Milky Way extinction also occur in a Wolf-Rayet or carbon-oxygen star, since
(Schlafly & Finkbeiner 2011) and used to fit the SED there is no sign of hydrogenin the spectra of SN 2010gx
of a stellar population model (Maraston 2005). We or any similar SNe. Gal-Yam (2012) gives an estimated
used the low-metallicity (Z∼ 1/20Z⊙) models, assum- relativerateofthe number of2010gx-likeSNe compared
to the total core-collapse SN rate of 10−4. If the pro-
7 Thiswouldbereducedby0.05magifweusedavaluesuitable genitorchannelrequireslowmetallicity(aswemeasure),
foranSMCenvironment;RV =2.93(Pei1992). high mass (to produce a WC star) and fast rotation (to
4 Chen et al.
produce a rapidly rotating magnetar) then this very low low metallicity (e.g. in IZw13 at 0.02Z⊙) (Brown et al.
rate may not be unexpected. Young et al. (2008) esti- 2002) could suggest that pulsational mass-loss is more
mate that ∼1% of the star formation in the Local Uni- prevalent at low metallicity. This one measurement is
verse (z < 0.04) is in galaxies with Z < 0.2Z⊙; around intriguingnewinformationontheseextremeSNebutfur-
13% of massive stars above the SN threshold of 8M⊙ thermeasurementsofthemetallicity,starformationrate
are>60M⊙ (assumingaSalpeterIMF);andabout10% and volumetric rates of these supernovae are requiredto
of O-stars in the SMC rotate faster than ∼300kms−1 constrain their explosion channels further.
(Mokiem et al.2007). Onecanspeculatethatallofthese
three conditions are required to produce the rate of the ACKNOWLEDGMENTS
2010gx-like SNe and the estimated rate (which is itself S.J.S acknowledges support from the European Re-
uncertain)wouldbe qualitativelyreproduced. Thehigh- searchCouncilintheformofanAdvancedGrant. T.-W.
mass requirement could also be substituted for an in- ChenappreciatestoMeng-ChungTsai,ZhengZhengand
teractingbinarysystemwhichcausesspin-upintheCO- Kai-Lung Sun for valuable comments and Prof. Christy
corethatcollapses,similartotheproposedGRBsystems Tremonti, Dr. Elisabete da Cunha and Dr. Stephane
(Cantiello et al. 2007). The Chevalier & Irwin (2011) Charlot for useful advice; thanks to all QUB SN group
scenario of interaction with a dense wind would prob- members for discussions.
ably require extreme progenitor mass and pulsational Facilities: Gemini North and South (GMOS)
mass-loss. The fact thatwe see Wolf-Rayetstarsatvery
REFERENCES
Amor´ın,R.,P´erez-Montero,E.,V´ılchez,J.M.,&Papaderos,P. Kudritzki,R.-P.,Urbaneja,M.A.,Gazak, Z.,etal.2012,ApJ,
2012, ApJ,749,185 747,15
Arnett,W.D.1982, ApJ,253,785 Lee,H.,Skillman,E.D.,Cannon,J.M.,etal.2006,ApJ,647,970
Asplund,M.,Grevesse,N.,Sauval,A.J.,&Scott, P.2009, Lee,J.C.,Salzer,J.J.,&Melbourne,J.2004,ApJ,616,752
ARA&A,47,481 Maeda,K.,Nomoto,K.,Mazzali,P.A.,&Deng,J.2006, ApJ,
Berg,D.A.,Skillman,E.D.,Marble,A.R.,etal.2012,ApJ,754, 640,854
98 Maeder,A.,&Meynet,G.2000,ARA&A,38,143
Berger,E.,Chornock,R.,Lunnan, R.,etal.2012,ApJ,755,L29 Maraston,C.2005, MNRAS,362,799
Blinnikov,S.I.,&Sorokina,E.I.2010, ArXive-prints Martayan,C.,Fr´emat,Y.,Hubert,A.-M.,etal.2007,A&A,462,
Bresolin,F.,Gieren,W.,Kudritzki,R.-P.,etal.2009,ApJ,700, 683
309 McKenzie,E.H.,&Schaefer,B.E.1999,PASP,111,964
Brown,T.M.,Heap,S.R.,Hubeny,I.,Lanz,T.,&Lindler,D. Modjaz,M.,Kewley,L.,Kirshner,R.P.,etal.2008,AJ,135,
2002, ApJ,579,L75 1136
Cantiello,M.,Yoon,S.-C.,Langer,N.,&Livio,M.2007,A&A, Mokiem,M.R.,deKoter,A.,Vink,J.S.,etal.2007,A&A,473,
465,L29 603
Chevalier,R.A.,&Irwin,C.M.2011,ApJ,729,L6 Neill,J.D.,Sullivan,M.,Gal-Yam,A.,etal.2011,ApJ,727,15
Chilingarian,I.V.,Melchior,A.-L.,&Zolotukhin, I.Y.2010, Osterbrock,D.E.1989, Astrophysicsofgaseous nebulaeand
MNRAS,405,1409 activegalacticnuclei
Chomiuk,L.,Chornock,R.,Soderberg,A.M.,etal.2011,ApJ, Pagel,B.E.J.,Edmunds,M.G.,Blackwell,D.E.,Chun,M.S.,
743,114 &Smith,G.1979, MNRAS,189,95
Clocchiatti,A.,Suntzeff, N.B.,Covarrubias,R.,&Candia,P. Papaderos,P.,Guseva,N.G.,Izotov, Y.I.,&Fricke,K.J.2008,
2011, AJ,141,163 A&A,491,113
daCunha,E.,Charlot,S.,&Elbaz,D.2008,MNRAS,388,1595 Pastorello,A.,Smartt,S.J.,Botticella,M.T.,etal.2010, ApJ,
Gal-Yam,A.2012,Science, 337,927 724,L16
Gal-Yam,A.,Mazzali,P.,Ofek,E.O.,etal.2009,Nature,462, Pei,Y.C.1992,ApJ,395,130
624 P´erez-Montero,E.,&D´ıaz,A.I.2003, MNRAS,346,105
Guseva,N.G.,Papaderos, P.,Meyer,H.T.,Izotov, Y.I.,& Quimby,R.M.,Kulkarni,S.R.,Kasliwal,M.M.,etal.2011,
Fricke,K.J.2009,A&A,505,63 Nature,474,487
Hoyos,C.,Koo,D.C.,Phillips,A.C.,Willmer,C.N.A.,& Rau,A.,Kulkarni,S.R.,Law,N.M.,etal.2009,PASP,121,1334
Guhathakurta, P.2005, ApJ,635,L21 Richer,M.G.,&McCall,M.L.1995, ApJ,445,642
Kaiser,N.,Burgett,W.,Chambers,K.,etal.2010,inSocietyof Schlafly,E.F.,&Finkbeiner,D.P.2011,ApJ,737,103
Photo-Optical Instrumentation Engineers(SPIE)Conference Seaton, M.J.1979,MNRAS,187,73P
Series,Vol.7733, SocietyofPhoto-Optical Instrumentation Skillman,E.D.,Kennicutt, R.C.,&Hodge,P.W.1989,ApJ,
Engineers(SPIE)ConferenceSeries 347,875
Kasen,D.,&Bildsten,L.2010,ApJ,717,245 Sollerman,J.,Kozma,C.,Fransson,C.,etal.2000, ApJ,537,
Kennicutt, Jr.,R.C.1998,ARA&A,36,189 L127
Kennicutt, Jr.,R.C.,Bresolin,F.,&Garnett, D.R.2003,ApJ, Wright,E.L.2006,PASP,118,1711
591,801 Young,D.R.,Smartt,S.J.,Mattila,S.,etal.2008,A&A,489,
Kewley,L.J.,Brown,W.R.,Geller,M.J.,Kenyon, S.J.,& 359
Kurtz,M.J.2007,AJ,133,882 Young,D.R.,Smartt,S.J.,Valenti,S.,etal.2010, A&A,512,
A70
SN 2010gx host 5
Fig.1.—Left: upper-Colourcombined(gri)GeminiGMOSimagesofthehostgalaxyofSN2010gx;middle-AfakeSN5σdetectionin
r-bandaftermagesubtraction;bottom-Afake3σdetection. Right: Photometryofthehostgalaxy(redcircle)withthebest-fit(χ2=0.58)
modelgalaxySED(blackline)fromMAGPHYS(daCunhaetal.2008),thebluelineshowsthesamemodelwithoutattenuationbydust.
6 Chen et al.
100
SN 2010gx- i
SN 1998bw- Rc
SN 2010gx- r
s] SN 1998bw- V
g
n
A
/
s
/ 10
g
r
e

7
3
0
1

g ]
s 2.2
o g
[l n
A
y
sit 1 g/s/ 1.8
o r
e
in 7 1.4
m 3
0
u 1
L [ 1
222 226 230
0.1
100 150 200 250 300
Phase [Days]
Fig. 2.—ThelimitingluminositiesofSN 2010gx compared withSN 1998bw (Clocchiatti etal.2011). At226 days wehave afluxlimit
of0.8thatofSN1998bw.
SN 2010gx host 7
6
5
1 0.3
s] Hγ [NII] [SII]
g
n 4 [OIII]
A
2/
m 0.0
c 3 0
s/ [OIII]
g/ 4300 4400 6500 6600 6700 6800
r
e
7 2 [OII]
1
-0
x [1 1 Hγ
u [NII]
Fl
0
Hδ Hβ Hα
[NeIII] [OIII] [OIII] [SII]
-1
3500 4000 4500 5000 5500 6000 6500 7000
Wavelength [Angs]
1100
SN 2010gx Host, this work
Green Pea, Amorin et al. 2012
99..55 GRB Host, Modjaz et al. 2008
Skillman et al. 1989
Richer and McCall, 1995
99
))
HH 88..55
O/O/
((
gg
oo
+l+l 88
22
11
77..55
Lee et al. 2006
Kewley et al. 2007
77 Guseva et al. 2009
Lee et al. 2004
Hoyos et al. 2005
Berg et al. 2012
66..55
--1100 --1122 --1144 --1166 --1188 --2200 --2222
MM
BB
Fig. 3.—Upper: SpectrumofthehostgalaxyofSN2010gx. The[OIII]λ4363˚Aaurorallineisshownintheinset. Bottom: Luminosity-
metallicity relationship for dwarf galaxies in the local Universe, showing the metallicity as measured using the direct (Te) method, and
usingdatafromtheliteraturelistedinthefigurekey. Greylinesaretheboundaryofnormalgalaxymass-metallicityrelationadaptedfrom
Kudritzkietal.(2012),wheremetallicitieshavebeencalculatedusingbluesupergiants.
Date MJD Phase g r i z Telescope Instrument
2010Jun4 55351.97 56.5 21.13(0.19) GS GMOS-S
2010Jun8 55354.93 58.9 23.55(0.28) 21.52(0.09) 21.31(0.12) WHT ACAM
2010Jun8 55355.48 59.3 21.58(0.13) 21.31(0.16) FTS EM03
2010Jun13 55360.47 63.4 23.91(0.30) 21.81(0.20) 21.56(0.13) FTS EM03
2010Jun13 55360.92 63.8 23.95(0.49) 21.83(0.12) 21.58(0.29) LT RATcam
2010Jun16 55363.48 65.8 >23.73 21.91(0.16) FTS EM03
2010Jun29 55376.38 76.3 22.47(0.21) 22.38(0.33) FTS EM03
2010Dec30 55560.63 226.1 H22.90(0.13) H22.98(0.16) GN GMOS-N
>25.94(0.17) >24.58(0.29)
al. 2011Jan14 55575.32 238.1 H23.71(0.14) H22.98(0.10) GS GMOS-S
t >26.42(0.18) >25.68(0.11)
e
n 2011Feb28 55620.12 274.5 H23.20(0.13) H23.00(0.13) GS GMOS-S
he >24.78(0.13) >23.35(0.32)
C 2011May20 55701.28 340.5 H23.72(0.13) GN GMOS-N
>26.00(0.15)
2011May25 55706.08 344.4 H23.08(0.08) H23.02(0.11) H23.06(0.16) GS GMOS-S
>25.35(0.15) >24.60(0.20)
2012Jan18 55944.22 538.0 H23.03(0.10) GS GMOS-S
2012Jan20 55946.23 539.6 H23.17(0.11) GS GMOS-S
2012Jan21 55947.21 540.4 H23.62(0.11) GS GMOS-S
TABLE 1
Observed photometry ofSN2010gxand itshost galaxy. Datauntil2010Dec30 are fromPastorello et al. (2010). The symbol”H” indicateshost galaxy
photometry,andthe”>”expressesthelimitingmagnitudeoftheSN(5σ). Phasehasbeencorrectedfortimedilation,basedontheSNexplosionatMJD55282.51.
8
SN 2010gx host 9
SDSSname SDSSJ112546.72-084942.0
RA 11:25:46.72
Dec -08:49:42.0
Redshift 0.23
Apparentg (mag) 23.71±0.14
Apparentr (mag) 22.98±0.10
Apparenti(mag) 22.90±0.13
Apparentz (mag) 22.98±0.16
Galacticextinctiong (mag) 0.13
k-correctiong (mag) 0.27±0.08
Internal extinctiong (mag) ∼0.5(RV =3.16)
Luminositydistance(Mpc) 1114.7
Absoluteg (mag) -17.42±0.17
PhysicalDiameter(kpc) 3.1
HαFlux(ergs−1 cm−2) 3.49e-16
SFR(M⊙ year−1) 0.41
Stellarmass(M⊙) 108.17
sSFR(year−1) 2.76×10−9
12+log(O/H) 7.46±0.1
Line MeasuredFlux±Error
(ergs−1 cm−2)
[OII]λ3727 1.297e-16±4.33e-18
[NeIII]λ3868 6.120e-17±2.65e-18
Hδ λ4102 4.254e-17±2.97e-18
Hγ λ4340 5.267e-17±3.59e-18
[OIII]λ4363 2.269e-17±2.33e-18
Hβ λ4861 1.285e-16±3.25e-18
[OIII]λ4959 2.129e-16±3.37e-18
[OIII]λ5007 6.569e-16±8.13e-18
Hαλ6563 4.996e-16±5.07e-18
[NII]λ6584 9.807e-18±8.21e-19
[SII]λ6717 1.918e-17±7.96e-19
[SII]λ6731 1.820e-17±9.29e-19
c(Hβ) 0.48±0.04
-W(Hβ)(˚A) 52
TABLE 2
Mainproperties of the hostgalaxyof SN2010gx.