SN 2013aa and SN 2017cbv: Two Sibling Type Ia Supernovae in the spiral galaxy NGC 5643
Christopher R Burns, Chris Ashall, Carlos Contreras, Peter Brown, Maximilian Stritzinger, M M Phillips, Ricardo Flores, Nicholas B Suntzeff, Eric Y Hsiao, Syed Uddin, Joshua D Simon, Kevin Krisciunas, Abdo Campillay, Ryan J Foley, Wendy L Freedman, Lluís Galbany, Consuelo González, Peter Hoeflich, S Holmbo, Charles D Kilpatrick, Robert P Kirshner, Nidia Morrell, Nahir Muñoz-Elgueta, Anthony L Piro, César Rojas-Bravo, David Sand, Jaime Vargas-González, Natalie Ulloa, Jorge Anais Vilchez
DDraft version April 29, 2020
Typeset using L A TEX twocolumn style in AASTeX62
SN 2013aa and SN 2017cbv: Two Sibling Type Ia Supernovae in the spiral galaxy NGC 5643 ∗ Christopher R. Burns, Chris Ashall, Carlos Contreras, Peter Brown, Maximilian Stritzinger, M. M. Phillips, Ricardo Flores, Nicholas B. Suntzeff, Eric Y. Hsiao, Syed Uddin, Joshua D. Simon, Kevin Krisciunas, Abdo Campillay, Ryan J. Foley, Wendy L. Freedman, Llu´ıs Galbany, Consuelo Gonz´alez, Peter Hoeflich, S. Holmbo, Charles D. Kilpatrick, Robert P. Kirshner,
11, 12
Nidia Morrell, Nahir Mu˜noz-Elgueta, Anthony L. Piro, C´esar Rojas-Bravo, David Sand, Jaime Vargas-Gonz´alez, Natalie Ulloa, and Jorge Anais Vilchez Observatories of the Carnegie Institution for Science, 813 Santa Barbara St, Pasadena, CA, 91101, USA Department of Physics, Florida State University, Tallahassee, FL 32306, USA Carnegie Observatories, Las Campanas Observatory, Casilla 601, Chile George P. and Cynthia Woods Mitchell Institute for Fundamental Physics and Astronomy, Texas A&M University, Department ofPhysics and Astronomy, College Station, TX, 77843, USA Department of Physics and Astronomy, Aarhus University, Ny Munkegade 120, DK-8000 Aarhus C, Denmark Department of Physics & Astronomy, San Francisco State University, 1600 Holloway Avenue, San Francisco, CA, 94132, USA Departamento de F´ısica y Astronom´ıa, Universidad de La Serena, Av. Cisternas 1200 Norte, La Serena, Chile Department of Astronomy and Astrophysics, University of California, Santa Cruz, CA 95064, USA Department of Astronomy & Astrophysics, University of Chicago, 5640 South Ellis Avenue, Chicago, IL 60637, USA Departamento de F´ısica Te´orica y del Cosmos, Universidad de Granada, E-18071 Granada, Spain Gordon and Betty Moore Foundation, 1661 Page Mill Road, Palo Alto, CA 94304, USA Harvard-Smithsonian Center for Astrophysics, 60 Garden Street, Cambridge, MA 02138, USA Carnegie Observatories, Las Campanas Observatory, Casilla 601, La Serena, Chile Department of Astronomy/Steward Observatory, 933 North Cherry Avenue, Rm. N204, Tucson, AZ 85721-0065, USA Centre for Astrophysics Research, School of Physics, Astronomy and Mathematics, University of Hertfordshire, College Lane, HatfieldAL10 9AB, UK Centro de Astronom´ıa (CITEVA), Universidad the Antofagasta, Avenida Angamos 601, Antofagasta, Chile (Accepted April 26, 2020)
Submitted to ApJABSTRACTWe present photometric and spectroscopic observations of SN 2013aa and SN 2017cbv, two nearlyidentical type Ia supernovae (SNe Ia) in the host galaxy NGC 5643. The optical photometry has beenobtained using the same telescope and instruments used by the
Carnegie Supernova Project . Thiseliminates most instrumental systematics and provides light curves in a stable and well-understoodphotometric system. Having the same host galaxy also eliminates systematics due to distance andpeculiar velocity, providing an opportunity to directly test the relative precision of SNe Ia as standardcandles. The two SNe have nearly identical decline rates, negligible reddening, and remarkably similarspectra and, at a distance of ∼
20 Mpc, are ideal as potential calibrators for the absolute distanceusing primary indicators such as Cepheid variables. We discuss to what extent these two SNe can beconsidered twins and compare them with other supernova “siblings” in the literature and their likelyprogenitor scenarios. Using 12 galaxies that hosted 2 or more SNe Ia, we find that when using SNe Ia,and after accounting for all sources of observational error, one gets consistency in distance to 3%.
Keywords: supernovae: general, cosmology: cosmological parameters, ISM:dust, extinction
Corresponding author: Christopher [email protected] ∗ This paper includes data gathered with the 6.5-meter Magel-lan Telescopes located at Las Campanas Observatory, Chile. a r X i v : . [ a s t r o - ph . S R ] A p r Burns, Ashall, Contreras, Brown, et al. INTRODUCTIONThrough the application of empirical-based calibra-tion techniques (Phillips 1993; Hamuy et al. 1996;Phillips et al. 1999), type Ia supernovae (SNe Ia) haveserved as robust extragalactic distance indicators. Earlywork focused on the intrinsic scatter with respect toa first or second order fit of absolute magnitudes atmaximum light vs. decline rate at optical wavelengths(Hamuy et al. 1996; Riess et al. 1999). The scatterin this luminosity-decline rate relation ranged betweennearly 0.2 mag in the U band, to 0.15 mag in the I band. Later work leveraging the near-infrared (NIR),where dimming due to dust is an order of magnitudelower (Fitzpatrick 1999), showed similar dispersions(Krisciunas et al. 2004; Wood-Vasey et al. 2008; Fo-latelli et al. 2010; Kattner et al. 2012) with possibly thelowest scatter at H band (Mandel et al. 2009).Working in the spectral domain, Fakhouri et al. (2015)introduced the notion of SN Ia twins, which are SNe Iathat have similar spectral features and therefore are ex-pected to have similar progenitor systems and explo-sion scenarios. They showed that sub-dividing the sam-ple into bins of like “twinness” results in dispersionsin distances to the SNe of ∼ .
08 mag. Then again,Foley et al. (2018a) demonstrated that two such twins(SN 2011by and SN 2011fe) appear to differ in intrinsicluminosity by ∆ M V = 0 . ± .
069 mag.Most of these analyses, however, are based on lowred-shift ( z (cid:46) .
1) samples, which are prone to extravariance due to the peculiar velocities and bulk flows oftheir host galaxies relative to cosmic expansion. Study-ing two (or more) SNe Ia hosted by the same galaxy,which we shall call “siblings” (Brown 2015), offers thepossibility of comparing their inferred distances withoutthis extra uncertainty, allowing a better estimate of theerrors involved. Studying supernova siblings also miti-gates any extra systematics that may be correlated withhost-galaxy properties (Sullivan et al. 2010; Kelly et al.2010; Lampeitl et al. 2010).The first study of SN Ia siblings was by Hamuy et al.(1991), who considered NGC 1316 (Fornax A), whichhosted two normal SNe Ia: SN 1980N and SN 1981D.Using un-corrected peak magnitudes of these SNe, theyfound the inferred distances differed by ∼ . −
8% in distance, however much of that was likely due to differences in photometric sys-tems, some of which were difficult to characterize. Theyalso found a larger discrepancy ( ∼ m ( B ) being a poor measure of thedecline rate for the fastest-declining SNe Ia. Using thecolor-stretch parameter s BV (Burns et al. 2018) insteadreduced the discrepancy to less than the measurementerrors.Gall et al. (2018) compared the distances from two“transitional” SNe Ia (Pastorello et al. 2007; Hsiao et al.2015; Ashall et al. 2016): SN 2007on and SN 2011ivhosted by NGC 1404, another Fornax cluster member.In this case, the photometric systems were identical, yetthe discrepancy in distances was ∼
14% in the opticaland ∼
9% in the NIR. It was argued that the observeddiscrepancy must be due to physical differences in theprogenitors of both systems. More specifically, the cen-tral densities of the progenitor white dwarfs (WDs) werehypothesized to differ (Gall et al. 2018; Hoeflich et al.2017; Ashall et al. 2018).In this paper, we consider two SNe Ia hosted by thespiral galaxy NGC 5643: SN 2013aa and SN 2017cbv.Both have extensive optical photometry obtained withthe 1-meter Swope telescope at Las Campanas Observa-tory (LCO) using essentially the same instrument andfilters. SN 2013aa was also observed in the optical byGraham et al. (2017) and SN 2017cbv was observed inthe optical by Sand et al. (2018), both using the LasCumbres Global Telescope (LCOGT) facilities. High-quality, NIR photometry is available for both objects,though from different telescopes. Both SNe Ia appearto be normal with respect to decline-rate and have col-ors consistent with minimal to no reddening. Spectraof the SNe taken at similar epochs indicate they arenot only siblings, but also spectroscopically very similar.SN 2017cbv is also unusual in having a very conspicu-ous ”blue bump” early in its light curve (Hosseinzadehet al. 2017a), which has only been seen in one other case(Marion et al. 2016). Lastly, NGC 5643 is close enoughto have its distance determined by primary methods,such as Cepheid variables and the Tip of the Red Gi-ant Branch (TGRB), making it an important anchor formeasuring the Hubble constant. OBSERVATIONS Between observations of SN 2013a and SN 2017cbv, the directcamera CCD was upgraded from the original SITe3 to e2V. Thisintroduces a change in the zero-points of the filters, but leavestheir relative shapes nearly identical. ibling Type Ia Supernovae RA (J2000) D e c ( J ) W1 12 34 5 678 910 11 12 1314 15 1617 18 1920 212223 24 2526272829 3031 32333435 36 37383940 414244 103104105 108 109110111113114
Figure 1.
The field of SN 2013aa and SN 2017cbv. The im-age was taken on the du Pont telescope using the direct CCDcamera through the B filter. The locations of the two SNeare labeled in red. The local sequence stars are plotted asyellow circles and labeled with their identification numbers.North is up, East is to the left. In this section we briefly describe the observations ofboth SNe Ia, the data reduction methods, and the pho-tometric systems involved.2.1.
Photometry
SN 2013aa was observed as part of the
Carnegie Su-pernova Project II (Phillips et al. 2019; Hsiao et al.2019, hereafter CSP-II). Optical imaging was obtainedwith the Swope telescope equipped with the direct SITe3CCD imager and a set of ugriBV filters. NIR photome-try was obtained using the 2.5-meter du Pont telescopewith a set of
Y JH filters. The observing procedures,data reduction, and photometric systems are outlinedin other CSP papers (Krisciunas et al. 2017; Phillipset al. 2019).SN 2017cbv was observed as part of the Swope Super-nova Survey (Coulter et al. 2017, Rojas-Bravo et al., inprep). Similar to SN 2013aa, optical photometry wasalso obtained on the Swope telescope with the directcamera, but with an upgraded e2V CCD, which has beenfully characterized by the CSP-II (Phillips et al. 2019).While the surveys employed slightly different strategiesfor target selection, this does not affect the observationspresented here. The exact pointings for SN 2013aa andSN 2017cbv observations differ slightly to position eachSN near the center of a chip. However because of thelarge Swope field of view, there are typically 25 localstandard stars in common between the images, whichwe use to calibrate all SN photometry. We emphasize that the optical photometry of bothSNe and the NIR photometry of SN 2013aa are given inthe CSP natural system. Using a long temporal base-line of observations at LCO, the CSP has determinedthe color terms that transform the instrumental mag-nitudes obtained on the Swope and du Pont telescopesinto the standard magnitudes of Landolt (1992) ( BV ),Smith et al. (2002) ( ugri ), Persson et al. (1998) ( JH )and Krisciunas et al. (2017) (Y). Using these color termsin reverse, we produce natural system magnitudes ofour standards and local sequence stars, which are thenused to differentially calibrate the photometry of eachSN. This greatly simplifies the procedure of transform-ing to another photometric system so long as the filterfunctions used to obtain science images are accuratelymeasured.The NIR photometry of SN 2017cbv was obtained byWee et al. (2018) using ANDICAM on the SMARTS 1.3-meter telescope at the Cerro Tololo Inter-American Ob-servatory (CTIO). They used standard observing pro-cedures, and tied their optical and NIR SN photome-try to a single local sequence star. Comparing our B -and V -band photometry after computing S-corrections(Stritzinger et al. 2002), we note a systematic differenceof order 0.1 – 0.2 mag, our photometry being dimmer.This appears to stem from the photometry of their sin-gle local sequence star (star 1 in Wee et al. 2018, starW1 in Figure 1), which is very red ( B − V = 1 . V ∼
13 mag). The red colorleads to a relatively large color-term correction to trans-form the instrumental magnitude to standard and mayintroduce a large systematic error. In fact, this star isbright enough to be saturated in all but our shallowestexposures, so was not used to calibrate the Swope pho-tometry. SN2017cbv was also observed independentlywith ANDICAM by a separate group and we find theirdata to be in agreement with the CSP data (Wang etal, in preparation).The local sequence star used by Wee et al. (2018) tocalibrate the NIR photometry, in contrast, has colorsmore consistent with the Persson et al. (1998) standards( J − H ∼ . Y -band calibration from Krisciunas et al. (2017),with no color term applied, making their Y -band pho-tometry nearly identical to the CSP natural system, dif-fering only in the shapes of the Y bandpasses. Unfortu-nately, there is no overlap between our RetroCam andthe ANDICAM fields, making a direct comparison oflocal sequence stars impossible.Lastly, the Neil Gehrels Swift Observatory observedboth SN 2013aa and SN 2017cbv as part of the
The SwiftOptical/Ultraviolet Supernova Archive (SOUSA; Brown
Burns, Ashall, Contreras, Brown, et al. u Days after B maximum B O b s e r v e d M a g n i t u d e V g r i
SN2013aaSN2017cbv
Days after B maximum Y O b s e r v e d M a g n i t u d e J H SN2013aaSN2017cbv
Days after B maximum
UVM2 O b s e r v e d M a g n i t u d e UVW1
UVW2
SN2013aaSN2017cbv
Figure 2.
Optical, NIR and UV light curves of SN 2013aa and SN 2017cbv. The blue points correspond to SN 2013aa and theorange points correspond to SN 2017cbv. The NIR photometry of SN 2017cbv was obtained from images taken with ANDICAM.The UV points are from SWIFT. The solid black curves are
SNooPy fits to SN 2013aa. et al. 2014). Details of the photometric data reductionand calibration are given in Brown et al. (2009).
Table 1 . A log of the visual-wavelength and NIR spectra presented inthis work.SN T spec a Phase b Instrument/TelescopeJD − Optical
SN 2013aa 56339.4 − NIR
SN 2013aa 56357.90 +14.7 FIRE/BaadeSN 2017cbv 57857.64 +17.1 FIRE/Baade a Time of spectral observation. b Phase of spectra in rest frame relative to the epoch of B -band max-imum. Spectroscopy
Visual-wavelength spectra of SN 2017cbv were ob-tained from Hosseinzadeh et al. (2017a), while thoseof SN 2013aa come from WISeREP (Yaron & Gal-Yam2012), with the exception of two spectra that were ob-tained with WiFeS (Dopita et al. 2007). WiFeS is an In- tegral Field Unit (IFU) on the 2.3-meter Australian Na-tional University telescope located at the Siding SpringObservatory. The WiFeS IFU has a 25 ×
38 arcsec field ofview. Two gratings were used, for the blue and red cam-eras respectively, with a resolution of R = 3000 alongwith a dichroic at 5600 ˚A. The data were reduced us-ing pyWiFeS (Childress et al. 2014). The spectra werecolor-matched to the multi-band photometry to ensureaccurate flux calibration.We also present NIR spectra of both SN 2013aa andSN 2017cbv obtained with the FIRE spectrograph on the6.5-meter Magellan Baade telescope at LCO. The spec-tra were reduced and corrected for telluric features fol-lowing the procedures described by Hsiao et al. (2019).Additionally, high-resolution spectra of SN 2013aa andSN 2017cbv were obtained using the Magellan InamoriKyocera Echelle (MIKE; Bernstein et al. 2003). Thedata were reduced using the MIKE pipeline (Kelson2003) and procedures outlined in Simon et al. (2010).A journal of spectroscopic observations is provided inTable 1. RESULTSWe present here the analysis of the data from the pre-vious section, including photometric classification andthe distances derived from common template light curvefitters. The spectra are used to delve into the physicalcharacteristics of the explosions and progenitors.3.1.
Decline Rate
Figure 2 presents a comparison of the photometry ofSN 2013aa and SN 2017cbv obtained with the Swope, du ibling Type Ia Supernovae Table 2.
Comparison of LC parametersParameter SN 2013aa SN 2017cbvSpline Fits t Bmax (MJD) 56343.20(07) 57840.54(15)∆ m ( B ) (mag) 0.95(01) 0.96(02) s DBV B max V max B max − V max -0.048(005) -0.056(015)SNooPy a t Bmax (MJD) 56343.42(34) 57840.39(34) s BV E ( B − V ) (mag) -0.03(06) 0.03(06) A V (mag) -0.06(12) 0.06(12) µ (mag) 30.47(08) 30.46(08)MLCS2k2 t Bmax (MJD) 56343.40(11) days 57839.79(06)∆ (mag) -0.09(02) -0.26(02) A V (mag) -0.04(05) 0.23(05) µ (mag) b t max (MJD) 56343.95(03) 57840.66(03) x x c (mag) -0.20(02) -0.02(03) B max (mag) 11.01(02) 11.17(03) µ (mag) c Note —All magnitudes and colors are corrected forMilky-Way extinction E ( B − V ) = 0 .
15 mag basedon Schlafly & Finkbeiner (2011). a Fit using the
EBV2 model with R V = 2. b Re-scaled from H = 65 km s − Mpc − to H =72 km s − Mpc − . c Using “JLA” calibration of Betoule et al. (2014) with M B re-scaled from H = 70 km s − Mpc − to H =72 km s − Mpc − Pont, SMARTS, and Swift telescopes. The light curvesof the two objects are remarkably similar, suggestingtheir progenitors could also be very similar. Table 2lists pertinent photometric parameters estimated via themethods described below.The most straightforward photometric diagnosticto compare is the light curve decline-rate parameter∆ m ( B ) (Phillips 1993). Using the light curve analysis Days relative to B-band maximum B - V ( m a g ) Figure 3.
The B − V color-curve evolution of SN 2013aaand SN 2017cbv. The time of B − V maximum relative to B -band maximum divided by 30 days is defined as the colorstretch, s DBV , which is nearly identical for these two SNe. package
SNooPy (Burns et al. 2011), the photometry isinterpolated at day 15 in the rest frame of the SNe afterapplying K-corrections (Oke & Sandage 1968) and timedilation corrections. This results in nearly the same de-cline rate ∆ m ( B ) (cid:39) .
95 mag, indicating both objectsdecline slightly slower than the typical ∆ m ( B ) valueof 1.1 mag (Phillips et al. 2019, Fig. 10).Another way to characterize the decline-rate of theSNe directly from photometry is to measure the epochthat the B − V color-curves reaches their maximum point(i.e., when they are reddest) in their evolution relativeto the epoch of B -band maximum. Dividing by 30 daysgives the observed color-stretch parameter s DBV . Burnset al. (2014) showed that the color-stretch parameterwas a more robust way to classify SNe Ia in terms oflight curve shape, intrinsic colors and distance (Burnset al. 2018). Figure 3 shows the B − V color-curves forSN 2013aa and SN 2017cbv. Fitting the B − V color-curves with cubic splines, we find identical color-stretchvalues of s DBV = 1 .
11 for both objects, again makingthem slightly slower than typical SNe Ia.For the purposes of determining distances for cosmol-ogy, it is more common to fit multi-band photometrysimultaneously, deriving a joint estimate of the declinerate, color/extinction, and distance. To this end, thelight curves of both SNe Ia are fit with 3 light curve fit- Analysis for this paper was done with
SNooPy version2.5.3, available at https://csp.obs.carnegiescience.edu/data/snpyor https://github.com/obscode/snpy We shall use a super-script D to distinguish between thedirectly-measured color-stretch s DBV and the value inferred bymulti-band template fits, s BV . Burns, Ashall, Contreras, Brown, et al. ( m a g ) s BV X SN2013aaSN2017cbvsub-luminous A V ( m a g ) E ( B V ) C o l o r SN2013aaSN2017cbvsub-luminous
Figure 4.
Comparison of parameters from three different SN Ia light curve fitters for the CSP-I sample. (Left) Comparingthe shape parameters, the
MLCS2k2 parameter ∆ (top panel) and
SALT2 parameter X (bottom panel) are plotted versus the SNooPy color-stretch parameter s BV . (Right) Comparing color/reddening parameters, the MLCS2k2 parameter A V (top panel)and SALT2 color parameter (bottom panel) are plotted versus the
SNooPy E ( B − V ) parameter. SN 2013aa and SN 2017cbv arelabeled with blue and orange circles, respectively. ting methods: 1) SNooPy , 2)
SALT2 (Guy et al. 2007),and 3) MLCS2k2 (Jha et al. 2007). The results of thesefits are listed in Table 2. In terms of the decline rate, aslightly different picture emerges: all three fitters clas-sify SN 2013aa as a faster decliner than SN 2017cbv. InFigure 4 we show how the different decline-rate parame-ters ( SNooPy ’s s BV , SALT2 ’s x , and MLCS2k2 ’s ∆) relateto each other for the CSP-I sample (Krisciunas et al.2017). It is clear that the three decline rate parametersfor the two SNe follow the general trend and are there-fore measuring the same subtle differences in light-curveshapes that tell us SN 2013aa is a faster decliner, de-spite having nearly identical ∆ m ( B ) and s DBV . Thisis because ∆ m ( B ) is determined from only two pointson the B light curve (maximum and day 15) and s DBV is determined from one point on the B light curve andone point on the B − V color curve, whereas templatefitters like SNooPy , MLCS2k2 and
SALT2 use the shapes ofmulti-band light curves to determine the decline rates.Looking closely at the color-curves of both SNe in Fig-ure 3, while the peaks are nearly identical, SN 2017cbvrises more quickly prior to maximum and declines moreslowly after maximum.3.2.
Extinction
Both SN 2013aa and SN 2017cbv are located in theouter parts of NGC 5643 on opposite sides, having a host SALT version 2.4.2 is available from http://supernovae.in2p3.fr/salt/ with updated CSP photometric system files available athttps://csp.obs.carnegiescience.edu/data/filters MLCS2k2 ∼ saurabh/mlcs2k2/ and updated CSP photometricsystem files are available at https://csp.obs.carnegiescience.edu/data/filters − . − . − . − . − . − . . . l og ( F l u x ) + o ff s e t − Figure 5.
Continuum-normalized MIKE spectra ofSN 2013aa and SN 2017cbv in the vicinity of the Na i doublet.Absorption from the Milky-Way is clearly visible. The bluevertical lines denote the rest wavelengths of the Na I D25890 ˚A and Na I D1 5896 ˚A lines, and the red lines show theirlocation red-shifted to the systemic velocity of NGC 5643,i.e., 1200 kms. The names and phases of the spectra relativeto the epoch of peak brightness are indicated to the right ofeach spectrum. offset of 17 kpc and 14 kpc, respectively. We thereforeexpect minimal to no host-galaxy dust reddening for ei-ther object. However, NGC 5643 is at a relatively lowgalactic latitude ( l = 15 . ◦ ) with a predicted Milky-Way color excess of E ( B − V ) MW = 0 .
15 mag (Schlafly& Finkbeiner 2011).Figure 5 shows the continuum-normalized spectra ofSN 2013aa and SN 2017cbv taken with MIKE in theNa I D region. Absorption from the Milky-Way is clearlyvisible in both cases, while absorption at the systemicvelocity of NGC 5643 is only detected in the spectrumof SN 2017cbv. Measuring the equivalent width of thecombined Na I D lines and using the conversion from ibling Type Ia Supernovae E ( B − V ) MW = 0 . ± .
16 mag, somewhathigher than the Schlafly & Finkbeiner (2011) value, butwithin the uncertainty. The equivalent width of the hostNa I D corresponds to E ( B − V ) host = 0 . ± . I D absorption has beenshown to be a poor predictor of the amount of extinctionin SN Ia hosts, the absence of Na I D nevertheless seemsto be a reliable indicator of a lack of dust reddening(Phillips et al. 2013). In the remainder of this paperthe Schlafly & Finkbeiner (2011) value is adopted forthe Milky-Way reddening along with a reddening lawcharacterized by R V = 3 .
1, in order to correct the SNphotometry for the effects of Milky-Way dust.The photometric colors of both objects are very blueat maximum, with SN 2017cbv being only slightly bluer.However, when the correlation between intrinsic colorand decline rate (e.g. Burns et al. 2014) is taken into ac-count, one infers slightly more extinction in SN 2017cbvthan SN 2013aa.
SALT2 , which does not estimate the ex-tinction but rather a rest-frame color parameter, givesa bluer color ( c = − .
20) for SN 2013aa than forSN 2017cbv ( c = − . MLCS2k2 provides estimatesof the visual extinction, A V , and gives a significanthost extinction for SN 2017cbv ( A V = 0 . ± . E ( B − V ) = 0 . ± .
03 mag. However, like
SNooPy , MLCS2K2
K-corrects its template light curves tofit the CSP filters and our u band is significantly differ-ent from Johnson/Cousins U , resulting in rather largecorrections. Eliminating u from the fit brings the ex-tinction estimate down to A V = 0 . ± .
06 mag or E ( B − V ) = 0 . ± .
03 mag, consistent to within theerrors with the
SNooPy value.Given the positions of the two SNe, their lack of Na I Dabsorption at the velocity of the host, and that
SNooPy predicts zero color excess, we conclude that SN 2013aaand SN 2017cbv experience minimal to no significanthost-galaxy dust extinction, making them ideal objectsto improve upon the zero-point calibration of SNe Ia.3.3.
UV Diversity
An outstanding feature of Figure 2 is the differenceat peak in the single Swift band
U V M
U V W
U V W
2, are consistent. Thisis likely due to the fact that both
U V W
U V W . U V M See for example, Fig. 1 of Brown et al. (2010).
Figure 6.
A comparison of the SWIFT UVOT
UV W − V and UV M − v color curves. SN 2013aa, SN 2017cbv, andthe twins SN 2011fe and SN 2011by are plotted with symbols.The red and blue shaded regions in the upper panel delimitthe NUV-red and NUV-blue regions from Milne et al. (2013),respectively. better indicator for diversity in the UV, sampling thewavelength region 2000 − U V W − v and U V M − v colorsas a function of time for SN 2013aa and SN 2017cbv, aswell as the two “twins” SN 2011fe and SN 2011by. Twoshaded regions are also shown indicating two popula-tions of SNe Ia: the NUV-blue and NUV-red as definedby Milne et al. (2013). They would seem to indicatethat both SN 2013aa and SN 2017cbv are NUV-blue.However, when comparing the U V M − v colors, thereis much more diversity and while SN 2013aa is similar toSN 2011by, SN 2017cbv is much redder and SN 2011feis much bluer, particularly near maximum light. Thisunderscores the issues with the red leaks. Milne et al.(2013) attribute these differences in NUV colors to dif-ferences in how close the burning front reaches the sur-face of the ejecta. Alternatively, near-UV differencesmay be the result of varied metal content in the ejecta(e.g, Walker et al. 2012; Brown et al. 2019), though nocorrelation has been found between the near-UV andthe host galaxy metallicity (Pan et al. 2019; Brown &Crumpler 2019). 3.4. Spectroscopy
Optical
Burns, Ashall, Contreras, Brown, et al.
Figure 7 presents our spectral time series of SN 2013aaand SN 2017cbv. Only data obtained during the photo-spheric and transitional phase are shown as the nebularphase spectra will be the highlight of a future study.In the case of both objects, three optical spectra wereobtained starting at − B -band max-imum. The last spectra of SN 2013aa and SN 2017cbvwere obtained on +44 d and +49 d, respectively.Doppler velocities and pseudo-EW (pEW) measure-ments were calculated for each object by fitting thecorresponding features with a Gaussian function. Therange over which the data were fit was manually se-lected, and the continuum in the selected region wasestimated by a straight line. The velocity was then mea-sured by fitting the minimum of the Gaussian and theerror was taken as the formal error of the Gaussian fit.Finally, pEW measurements were obtained following themethod discussed in Garavini et al. (2007).The spectra of the SNe are characteristic of normalSNe Ia. The − II λ − ± − and − ± − , respectively. This places them bothas normal-velocity SN Ia in the Wang et al. (2013) classi-fication scheme. They are also both core normal (CN) inthe Branch et al. (2006) classification system, as demon-strated in the Branch diagram plotted in Figure 8. Al-though the two objects are located close to the boundarybetween CN and shallow silicon (SS) SNe Ia.Figure 9 shows a comparison of the spectra of thetwo SNe obtained at early ( − − − II λ ± ± ξ ( p i ) parameter using the − ξ is essentially a reduced χ statistic and therefore requires a good noise model that we esti-mate by boxcar-averaging each spectrum with a box sizeof 11 wavelength bins and subtracting these smoothedspectra from the originals. We also color-match eachspectrum to match the observed photometry and onlyconsider the wavelength range (4000–9500 ˚A), corre-sponding to the wavelength coverage of our BV ri pho-tometry. The resulting value of ξ ( −
2) = 1 . ξ of 1.7 or less, indicating SN 2013aa andSN 2017cbv are not spectroscopic twins by this metric.We note, however, that the two spectra were obtainedon different instruments. This could introduce system-atic errors that are not accounted for in the noise modeland could lead to an over-estimate of ξ .3.4.2. NIR
NIR spectra of SN 2013aa and SN 2017cbv obtainedat +14 . H -band regionemerges in all SNe Ia by +10 d linked to allowed emis-sion lines of Co ii , Fe ii and Ni ii produced from theradioactive decay of Ni which is located above the pho-tosphere (Wheeler et al. 1998; Hsiao et al. 2015). Ashallet al. (2019a) found a correlation between the the outerblue-edge velocity, v edge , of this H -band break regionand s BV . Furthermore, Ashall et al. (2019b) found that v edge was a direct measurement of the sharp transitionbetween the incomplete Si-burning region and the re-gion of complete burning to Ni. v edge measures the Ni mass fraction between 0.03 to 0.10.Using the method of Ashall et al. (2019a), v edge ismeasured from the NIR spectra of both objects. At+14 . v edge of − ± − ,and at +17 . v edge of − ± − .SN 2017cbv has a lower value of v edge by 1300km s − ,however this is likely to be due to the fact that v edge decreases over time (see Ashall et al. 2019a).The fact that both SNe have similar values of v edge and absolute magnitude indicates that they probablyhave a very similar total ejecta mass. As explained inAshall et al. (2019b), for a given Ni mass, smallerejecta masses produce larger values of v edge . However, v edge is similar in both SN 2017cbv and SN 2013aa,once the phase difference in the spectra is taken intoaccount. Furthermore, the value of v edge obtainedfrom SN 2017cbv is consistent with similar SNe fromAshall et al. (2019a), as well as with predictions ofChandrasekhar mass ( M Ch ) delayed-detonation explo- ibling Type Ia Supernovae v edge was given as 10 ± v edge is slow, hence comparing to spectraat +14 . PROGENITORSSNe Ia are thought to originate from the thermonu-clear disruption of Carbon-Oxygen (C-O) white dwarfs(WD) in binary systems. There are many popular pro-genitor and explosions scenarios (see Livio & Mazzali2018, for a recent review). Two of the favored explosionscenarios, which can occur in both the single and doubledegenerate progenitor system, are the double detona-tion (He-det) and delayed detonation scenarios (DDT).In the He-det scenario a sub- M Ch WD accretes He froma companion, either a He star or another WD with aHe layer, the surface He layer detonates and drives ashock wave into the WD producing a central detonation(Livne & Arnett 1995; Shen & Moore 2014; Shen et al.2018). In the DDT scenario, a WD accretes materialuntil it approaches the M Ch , after which compressionalheating near the WD center initiates a thermonuclearrunaway, with the burning first traveling as a subsonicdeflagration wave and then transitions into a supersonicdetonation wave.It has been predicted that SNe Ia spectra can looksimilar at maximum light, and their light curves canhave the same shape, but their absolute magnitudes candiffer by ∼ Ni heats the photosphere; that is, where the Ni islocated with respect to the photosphere. However, oncethe photosphere is within the Ni region the exact lo-cation of the Ni can differ and alter the light curveand spectra after 30 days past maximum light (Hoeflichet al. 2017).Figure 3 shows that the B − V color curves for bothobjects are very similar. Furthermore, the optical spec-tra of SNe 2013aa and 2017cbv are nearly identical at ∼ +43 d when the photosphere is well within the Ni re-gion. This demonstrates that these two objects are sim-ilar in the inner regions, and have similar ignition mech-anisms. However, at − II λ Ni production and hence the luminosity, asis seen with the extended Si region in SN 2013aa. How-ever, overall it is likely that the explosion mechanismfor both objects is very similar, and the observations ofthese SNe are largely consistent with the DDT and/orHe-Det scenarios. DISTANCEWith a heliocentric redshift z hel = 0 . µ = 30 . ± .
08 mag and µ = 30 . ± .
08 mag forSN 2013aa and SN 2017cbv, respectively. The differencein distance modulus is 0 . ± .
11, and therefore is in-significant. This compares well with the distance deter-mined by Sand et al. (2018), who used
MLCS2k2 on theirown photometry to derive a distance of µ = 30 . ± . MLCS2k2 infers a distance of µ = 30 . ± .
08 mag forSN 2013aa and µ = 30 . ± .
08 for SN 2017cbv, whichis a difference of 0 . ± .
11 mag, so within the uncer-tainty of the fitter.
SALT2 , however, infers distances of µ = 30 . ± .
11 for SN 2013aa and µ = 30 . ± .
12 forSN 2017cbv, a difference of 0 . ± .
16 mag. The largerdifferences for
SALT2 and
MLCS2k2 are primarily due todifferences in their color parameters (
MLCS2k2 inferringSN 2017cbv to have significant A V and SALT2 inferringSN 2013aa to have very blue color).Table 3 gives a list of the current sample of SNe Ia sib-lings in the literature. Using
SNooPy , we have deriveddecline rates, extinctions and distance estimates to eachobject. We then compare the inferred host distances,which are tabulated in the column labeled ∆ µ . In all, itwas found that 14 host galaxies have hosted two SNe Iaand one (NGC 1316) has hosted 4 SNe Ia. The differ-ences in distance estimate range from 0.02 mag to 0.43mag.Of particular interest are the siblings that were ob-served with the same telescopes and instruments, elimi-nating the systematic error of transforming photometry0 Burns, Ashall, Contreras, Brown, et al.
Table 3.
SN Ia siblings in the literature.
Host SN s BV E ( B − V ) µ ∆ µ Photometric system referencemag mag magNGC 105 SN1997cw 1.30(04) 0.29(07) 34.34(11) CfA a SN2007A 1.01(04) 0.24(06) 34.38(10) 0.04(15) CfA b SN2007A 1.10(02) 0.24(06) 34.44(11) 0.10(17) CSP c NGC 1316 SN2006mr 0.25(03) 0.03(04) 31.26(04) CSP c (Burns et al. 2018)SN1980N 0.88(03) 0.14(06) 31.27(09) 0.01(10) CTIO 1m d - photographicSN2006dd 0.93(03) 0.09(06) 31.29(09) 0.03(10) ANDICAM (Stritzinger et al. 2010)SN1981D 0.77(05) 0.05(09) 31.32(10) 0.06(11) CTIO 1m d NGC 1404 SN2011iv 0.64(03) − − − e SN2010ko 0.57(04) − e NGC 3190 SN2002cv 0.85(04) 5.40(09) 31.91(61) Standard (Elias-Rosa et al. 2008)SN2002bo 0.89(03) 0.40(06) 32.03(13) 0.12(62) CfA f SN2002bo 0.94(03) 0.43(06) 32.11(13) 0.20(62) KAIT g SN2002bo 0.92(03) 0.42(06) 32.11(14) 0.20(63) Standard + LCO NIR (Krisciunas et al. 2004)NGC 3905 SN2009ds 1.05(03) 0.07(06) 34.69(09) CfA b , PAIRITEL h SN2001E 1.02(04) 0.47(06) 34.85(14) 0.16(17) KAIT g NGC 4493 SN2004br 1.12(04) 0.01(06) 34.82(08) KAIT g SN1994M 0.88(04) 0.17(06) 35.07(09) 0.25(12) CfA (Riess et al. 1999)NGC 4708 SN2016cvn 1.25(12) 0.91(13) 33.71(24) Foundation (Foley et al. 2018b)SN2005bo 0.79(03) 0.28(06) 33.95(11) 0.24(26) CSP c SN2005bo 0.86(03) 0.37(06) 33.96(12) 0.25(27) KAIT g NGC 5468 SN1999cp 0.98(03) 0.06(06) 33.10(08) KAIT g , 2MASS, (Krisciunas et al. 2000)SN2002cr 0.91(03) 0.11(06) 33.17(08) 0.07(11) KAIT g SN2002cr 0.93(03) 0.10(06) 33.21(09) 0.11(12) CfA f NGC 5490 SN2015bo 0.41(08) 0.11(13) 34.18(19) Swift UVOT e SN1997cn 0.62(04) 0.12(06) 34.57(10) 0.39(21) CfA a NGC 5643 SN2017cbv 1.09(04) 0.08(06) 30.38(09) Swift UVOT e SN2013aa 0.95(03) − − − e NGC 6240 PS1-14xw 0.95(04) 0.26(06) 34.79(12) Swift UVOT e SN2010gp 1.06(07) 0.00(07) 35.22(12) 0.43(17) Swift UVOT e NGC 6261 SN2008dt 0.87(05) 0.49(09) 35.85(18) CfA b SN2008dt 0.81(05) 0.14(07) 35.87(15) 0.02(23) KAIT g SN2007hu 0.80(05) 0.39(08) 35.91(14) 0.06(22) CfA b UGC 3218 SN2011M 0.93(02) 0.08(06) 34.37(08) KAIT g SN2011M 0.85(05) − e SN2006le 1.08(04) − g SN2006le 1.20(03) − f , PAIRITEL h UGC 7228 SN2007sw 1.19(04) 0.14(07) 35.25(10) CfA b SN2012bh 1.11(04) 0.10(06) 35.39(09) 0.14(13) PANSTARRS (Jones et al. 2018) a Jha et al. (2006). b Hicken et al. (2012). c Contreras et al. (2010). d Hamuy et al. (1991). e Brown et al. (2014). f Hicken et al. (2009). g Silverman et al. (2012). h Friedman et al. (2015). ibling Type Ia Supernovae Rest wavelength Å F + C o n s t . Rest wavelength Å F + C o n s t . Figure 7.
A rest frame time series of spectra for SN 2013aa (left) and SN 2017cbv (right). Phases are given relative to B -bandmaximum. Telluric regions in the spectra are marked. Burns, Ashall, Contreras, Brown, et al.
EW 6355 Å E W Å Figure 8.
The Branch diagram for SNe Ia. SN 2013aa(blue star) and SN 2017cbv (orange triangle) are overlaid.The shallow silicon (purple), core normal (black), broad line(green) and cool line (red) SNe are plotted from Blondinet al. (2012). from one system to another (Stritzinger et al. 2005). Inthis regard,
Swift shows the greatest dispersion amongsiblings with ∆ µ = 0 .
43 mag for NGC 6240 and ∆ µ =0 .
28 mag for NGC 1954. In the case of NGC 6240, thetwo SNe have color excesses that differ by 0.26 mag andonly the UVOT B and V filters could be reliably fitwith SNooPy , requiring we assume the typical R V = 2for SNe Ia (Mandel et al. 2009; Burns et al. 2014), ratherthan fit for it using multi-band photometry. If R V is infact higher by an amount ∆ R V for this host, the discrep-ancy would decrease by (cid:39) . · ∆ R V . As a result, if R V were as high as 4 in NGC 6240, the discrepancy wouldbe eliminated. In the case of NGC 1954, SN 2010ko is atransitional Ia (Hsiao et al. 2013) which have been shownto be less reliable as standard candles (see discussion ofNGC 1404 below). This is also true of NGC 5490, inwhich both SNe are transitional.Further investigation comparing Swift and CSP pho-tometry has also shown systematic errors between someSNe in common which can be as high as 0.15 mag. Com-paring the UVOT photometry of local sequence starsduring the separate observing campaigns of the siblingSNe does not show a significant difference (N. Crum-pler, private communication). The
Swift sibling SNe arealso not located near bright stars or in bright regions ofthe host which can invalidate the standard non-linearitycorrections (Brown et al. 2014). Further investigation ofthese discrepancies is ongoing.The Lick Observatory Supernova Search (LOSS)observed two pairs of siblings in NGC 5468 and UGC 3218 using the Katzman Automatic ImagingTelescope (KAIT). Both pairs (SN 1999cp/SN 2002crand SN 2006le/SN 2011M) exhibit similar decline ratesand both pairs had low reddening. The distance es-timates for NGC 5468 differ by less than the error(∆ µ = 0 . ± . µ = 0 . ± . µ ∼ . µ = 0 . ± .
22 mag and∆ µ = 0 . ± .
15 mag respectively). In particular, bothSN 2007A and SN 1997cw have moderately high redden-ing E ( B − V ) host ∼ . µ = 0 . s BV > .
6) and fast-decliners( s BV < . m ( B ) as a light curvedecline-rate parameter, which was later shown to failfor fast-declining SN Ia (Burns et al. 2014). Using s BV instead leads to a SN 2006mr distance that is in com-plete agreement with the other normal SNe Ia (Burnset al. 2018). This is the distance tabulated in Table 3.We therefore have siblings in NGC 1404 that seem toindicate fast (or at least transitional) decliners are notas reliable, while NGC 1316 would indicate they are. Ifthere is a diversity in progenitors at these decline rates,then we may simply be seeing an increased dispersion ibling Type Ia Supernovae Rest wavelength Å F + C o n s t . -2d+44/49d C a II M g II S i III F e III F e II F e III S i II S II S i II S i II C a II Figure 9.
A comparison of the optical spectra of SN 2013aa and SN 2017cbv at two epochs. Phases are given relative to B -band maximum. Rest wavelength [ m] L o g ( F ) C a II F e II F e II C o II N i II C o II C o II C o II Figure 10.
A spectral comparison between SN 2013aa andSN 2017cbv in the NIR, at two different epochs. Phases aregiven in the legend. in the Phillips relation at the low s BV (high ∆ m ( B ))end, or perhaps two different progenitor scenarios. Toknow for sure will require an expanded sample of tran-sitional SNe Ia.More quantitatively, we have 34 pairs of distances thatcan be compared, including multiple observations of the Table 4.
Intrinsic dispersions in the sibling distances.
Subsample σ SN N pair N gal All pairs 0 . . s BV > . . s BV > . < .
03 (95% conf.) 21 11 same SN Ia with different telescopes/ instruments. Us-ing a simple Bayesian hierarchical model, we can solvefor an intrinsic dispersion σ SN in these distances, tak-ing into account the photometric errors, errors in theSN Ia calibration, and systematic errors due to differentphotometric systems (see appendix B for details of thismodeling). Using all pairs, we derive σ SN = 0 . ± . σ SN = 0 . ± .
03, or 3%in distance. If we further remove the fast declining SNe( s BV < . σ SN < . Burns, Ashall, Contreras, Brown, et al.
Within this landscape, we present two normal SNe Iasiblings observed with the same telescope and nearlyidentical detector response functions. Unlike most ofthe siblings in Table 3, SN 2013aa and SN 2017cbv havedense optical and NIR coverage, allowing for accuratemeasurement of the extinction, which is found to beconsistent with minimal to no reddening in both cases.The difference in distance modulus (∆ µ = 0 .
01) is lessthan the uncertainties, as was the case with the CfAand KAIT siblings. Being at such low redshift, we canexpect the SN Ia distances to differ from the Hubble dis-tance modulus by about δµ = . v pec cz hel = 0 .
54 mag for anassumed typical peculiar velocity of v pec = 300 km s − .With a CMB frame redshift z cmb = 0 . H = 72 km s − Mpc − , the distancemodulus is µ = 31 .
44 mag, nearly a magnitude larger.Applying velocity corrections for Virgo, the Great At-tractor, and the Shapley supercluster decreases the dis-tance modulus to µ = 30 .
82 mag. These differences indistance demonstrate the use of standard candles suchas SNe Ia for determining departures from the Hubbleexpansion at low redshift. With a precision of 3% indistances, we can measure deviations from the Hubbleflow at a level of δv = 0 . H d . This corresponds to δv (cid:39) − at the distance of the Virgo cluster (20Mpc) and δv (cid:39) − at the distance of Coma (100Mpc). CONCLUSIONSThe galaxy NGC 5643 is unique in that it has hostedtwo normal SNe Ia that have very similar properties andis close enough to have its distance determined indepen-dently. All photometric and spectroscopic diagnosticsindicate that SN 2013aa and SN 2017cbv are both nor-mal SNe Ia with minimal to no host-galaxy reddening.This is also consistent with their positions in the out-skirts of the host galaxy. Both objects have been ob-served in the optical with the same telescope and instru-ments and show a remarkable agreement in their lightcurves.Comparing the distances inferred by SN 2013aa andSN 2017cbv gives us the opportunity to test the rela-tive precision of SNe Ia as standard candles without anumber of systematics that typically plague such com-parisons. Given that they occurred in the same hostgalaxy, there is no additional uncertainty due to pe-culiar velocities. Having been observed by essentiallythe same telescope and instruments with the same filterset, there is also minimal systematic uncertainty due tophotometric zero-points or S-corrections. Finally, hav-ing no dust extinction, we eliminate the uncertainty dueto extinction corrections and variations in the reddening law (Mandel et al. 2009; Burns et al. 2014). When fit-ting multi-band photometry using
SNooPy , SALT2 , and
MLCS2k2 , SN 2013aa is found to be characterized by aslightly faster decline rate and bluer color at maximum.The net result is a difference in distance that is insignifi-cant compared to the measurement errors, a similar sit-uation as the other pairs of normal siblings observed inmultiple colors with the same instruments.The similarity between the spectra and light curvesof SN 2013aa and SN 2017cbv at all observed epochssuggest that they may have similar explosion mecha-nisms and progenitor scenarios. However the differ-ences between most leading explosion models are bestseen at early and late times. At these earliest epochs,SN 2017cbv showed an early blue excess, but there wereno data for SN 2013aa. Naturally the question is then:would SN 2013aa also have shown this early blue excess?Although we cannot make this comparison for these twoobjects, it is something that could be studied in the fu-ture using SN siblings.Some of the other questions that naturally arise for thefuture with twins and sibling studies are: if the spectralevolution were different, e.g. one SN was high velocitygradient, and the other low velocity gradient, or if theSNe were different spectral sub-types at maximum light(e.g. shallow silicon vs. core normal vs. cool) would thedistance calculated for each SN be different and, if so,does this point to different explosion mechanisms andprogenitor scenarios? Also, with the advent of IntegralField Spectrographs (IFS), we are beginning to studythe local host properties (Galbany et al. 2018). FutureIFS observations will allow us to investigate any corre-lation between these local properties and differences ininferred distance.With the exception of the
Swift siblings, which hadlimited wavelength coverage, we have four cases of nor-mal SNe Ia where most of the systematic errors are ab-sent and find the relative distance estimates to agree towithin 3%. Another host, NGC 1316, shows the samekind of consistency despite having multiple photomet-ric sources. This kind of internal precision rivals thatof Cepheid variables (Persson et al. 2004). The pic-ture is not as encouraging in the case of transitional-and fast-declining SNe Ia, and more examples of thesetypes of objects in the Hubble flow and/or additionalpairs of such siblings are required. Comparing siblingsfrom multiple telescopes also shows that we can expectdisagreements on order of 5% to 10%, higher than typi-cal systematic errors in the photometric systems them-selves. A likely reason for this is that these systematicsin the photometric systems don’t just affect the observedbrightness, but also the observed colors , which are mul- ibling Type Ia Supernovae R λ when correcting fordust. It is therefore advantageous to use the reddest fil-ters possible to avoid this systematic when determiningdistances to SNe Ia (Freedman et al. 2009; Mandel et al.2009; Avelino et al. 2019).After submission of this paper, Scolnic et al. (2020) re-leased a preprint detailing the analysis of sibling SNe Iafrom the DES survey. Unlike our results, they findthe dispersion among siblings to be no less than thenon-sibling SNe. Their sample, though, is quite differ-ent from ours. Their SNe are photometrically classifiedwhile ours are spectroscopically classified. They couldtherefore have peculiar SNe Ia in their sample. Theirs isalso a higher redshift sample, ranging from z = 0 .
228 to z = 0 . Facilities:
Swope (SITe3 and e2V imaging CCDs),Du Pont (Tek No. 5 imaging CCD, WFCCD), Magel-lan:Clay (MIKE), Swift (UVOT), La Silla-QUEST
Software:
SNooPy (Burnsetal.2014),
SALT2 (Guyetal.2007),
MLCS2k2 (Jha et al. 2007), astropy (The AstropyCollaboration et al. 2018), matplotlib (Hunter 2007)6
Burns, Ashall, Contreras, Brown, et al.
Table 5.
Natural system photometry ofSN 2013aa and SN 2017cbv.
MJD Filter Magnitude Phasedays mag daysSN2013aa56338.360 B . − . B . − . B . − . B . − . B . − . B . − . B . . B . . B . . B . − . B . − . B . − . B . − . B . − . B . − . B . − . B . − . B . − . B . − . Note —Table 5 is published in its entirety inthe machine-readable format. A portion isshown here for guidance regarding its formand content.
APPENDIX
Table 6 . Optical local sequence photometry for SN 2013aa and SN 2017cbv. ID α (2000) δ (2000) B V u (cid:48) g (cid:48) r (cid:48) i (cid:48) SN 2013aa1 218.161804 -44.242950 14.452(003) 13.842(003) 15.363(007) 14.114(003) 13.710(003) 13.553(003)2 218.234177 -44.159451 15.595(004) 14.391(003) 17.675(018) 14.960(003) 13.991(003) 13.608(003)3 218.143051 -44.280270 15.458(004) 14.664(004) 16.676(011) 15.037(004) 14.440(003) 14.244(003)4 218.170456 -44.211262 15.951(004) 15.146(004) 17.178(014) 15.520(004) 14.910(003) 14.702(004)5 218.151825 -44.275311 16.909(007) 16.231(005) 17.785(019) 16.551(005) 16.051(006) 15.836(006)6 218.065216 -44.236328 17.081(007) 16.307(005) 18.054(024) 16.668(005) 16.075(005) 15.833(005)7 218.226013 -44.254551 17.351(009) 16.450(006) 18.776(043) 16.890(005) 16.165(006) 15.915(006)8 218.230164 -44.216572 17.646(011) 16.794(007) 18.770(047) 17.211(006) 16.505(006) 16.218(005)9 218.213226 -44.240639 17.705(011) 16.924(007) 18.904(051) 17.330(007) 16.686(006) 16.359(006)10 218.218124 -44.214561 17.924(013) 16.953(007) 19.079(094) 17.427(007) 16.625(006) 16.233(005)
Table 6 continued ibling Type Ia Supernovae Table 6 (continued) ID α (2000) δ (2000) B V u (cid:48) g (cid:48) r (cid:48) i (cid:48)
11 218.178635 -44.235592 18.010(014) 17.062(008) 19.413(111) 17.541(007) 16.749(007) 16.460(006)12 218.088196 -44.292938 17.881(013) 17.003(008) 19.238(069) 17.413(007) 16.724(007) 16.439(007)13 218.062271 -44.280991 18.048(015) 17.130(009) 19.354(086) 17.551(008) 16.614(006) 16.520(007)14 218.219574 -44.284592 18.302(018) 17.345(010) 18.988(182) 17.796(009) 16.938(008) 16.597(007)15 218.114899 -44.264271 18.517(021) 17.703(013) . . . 18.103(011) 17.409(010) 17.141(010)16 218.077866 -44.175289 17.639(011) 17.619(011) 17.998(024) 17.557(008) 17.715(012) 17.879(016)17 218.221558 -44.179260 18.492(021) 17.833(014) 19.464(122) 18.191(013) 17.676(012) 17.444(012)18 218.122116 -44.163929 18.675(026) 17.909(017) 19.428(179) 18.332(017) 17.731(014) 17.434(013)19 218.084106 -44.192081 18.747(025) 17.768(013) . . . 18.260(013) 17.385(010) 17.053(009)20 218.161133 -44.292419 18.954(031) 18.137(019) 19.034(118) 18.496(017) 17.919(015) 17.663(015)21 218.083908 -44.224899 18.568(022) 17.833(014) 19.159(068) 18.190(012) 17.583(011) 17.315(010)22 218.110077 -44.206421 18.025(014) 16.957(007) . . . 17.484(007) 16.515(006) 16.143(005)23 218.153397 -44.260262 12.893(003) 12.342(003) 13.746(006) 12.558(003) 12.191(003) 12.092(003)24 218.109406 -44.252369 12.242(002) 11.677(003) 13.116(005) 11.900(003) 11.510(003) 11.388(003)25 218.052002 -44.260151 15.225(004) 14.612(005) 16.126(010) 14.893(004) 14.456(004) 14.290(004)26 218.105301 -44.261421 14.268(003) 13.596(003) 15.183(007) 13.898(003) 13.421(003) 13.235(003)27 218.175705 -44.218399 15.070(003) 14.348(003) 16.076(009) 14.679(003) 14.154(003) 13.958(003)28 218.168304 -44.201851 14.148(003) 13.100(003) 15.629(008) 13.591(003) 12.759(003) 12.414(002)29 218.160904 -44.185322 13.640(003) 12.977(003) 14.600(006) 13.272(003) 12.783(003) 12.610(002)30 218.066696 -44.166279 14.534(003) 13.849(003) 15.485(008) 14.157(003) 13.672(003) 13.488(003)31 218.229706 -44.292591 13.372(003) 12.630(003) 14.468(009) 12.950(003) 12.394(004) 12.216(003)SN 2017cbv1 218.161804 -44.242950 14.403(014) 13.812(006) 15.315(023) 14.091(007) 13.650(008) 13.518(010)2 218.234177 -44.159451 15.602(030) 14.372(006) 17.465(086) 14.944(006) 13.963(007) 13.601(011)3 218.143051 -44.280270 . . . . . . . . . . . . . . . . . .4 218.170456 -44.211262 15.955(037) 15.132(006) 16.968(055) 15.499(007) 14.888(006) 14.673(008)5 218.151825 -44.275311 16.934(023) 16.223(007) . . . 16.509(009) 16.027(011) 15.814(011)6 218.065216 -44.236328 17.087(057) 16.291(007) . . . 16.652(006) 16.044(006) 15.814(008)7 218.226013 -44.254551 17.370(050) 16.429(007) . . . 16.868(007) 16.132(007) 15.905(009)8 218.230164 -44.216572 17.667(115) 16.784(008) . . . 17.203(008) 16.467(006) 16.209(006)9 218.213242 -44.240639 17.794(031) 16.917(009) . . . 17.325(008) 16.618(007) 16.356(007)10 218.218124 -44.214561 17.925(066) 16.944(009) . . . 17.427(008) 16.553(007) 16.231(006)11 218.178635 -44.235592 18.077(045) 17.057(010) . . . 17.538(009) 16.731(007) 16.458(007)12 218.088196 -44.292938 17.890(013) 17.000(010) . . . 17.393(009) 16.693(008) 16.428(008)13 218.062271 -44.280991 18.020(070) 17.121(011) . . . 17.538(010) 16.792(008) 16.515(008)14 218.219559 -44.284592 18.347(109) 17.329(012) . . . 17.789(012) 16.906(009) 16.595(009)15 218.114899 -44.264271 18.552(162) 17.694(015) . . . 18.088(014) 17.378(011) 17.134(011)16 218.077881 -44.175289 17.642(079) 17.599(013) 17.307(135) 17.549(009) 17.705(013) 17.862(017)17 218.221558 -44.179260 18.511(065) 17.811(016) . . . 18.191(015) 17.666(013) 17.453(013)18 218.122131 -44.163929 18.738(066) 17.904(019) . . . 18.312(020) 17.707(016) 17.428(015)19 218.084106 -44.192081 18.811(081) 17.771(016) . . . 18.242(015) 17.373(010) 17.056(009)20 218.161133 -44.292419 18.918(010) 18.115(022) . . . 18.479(020) 17.886(015) 17.660(017)21 218.083908 -44.224899 18.616(052) 17.807(016) . . . 18.186(014) 17.563(012) 17.312(011)22 218.110062 -44.206421 18.074(040) 16.951(009) . . . 17.485(009) 16.486(006) 16.132(006)23 218.153397 -44.260262 . . . . . . . . . . . . . . . . . .24 218.109406 -44.252369 . . . . . . . . . . . . . . . . . .25 218.052002 -44.260151 . . . . . . . . . . . . . . . . . .
Table 6 continued Burns, Ashall, Contreras, Brown, et al.
Table 6 (continued) ID α (2000) δ (2000) B V u (cid:48) g (cid:48) r (cid:48) i (cid:48)
26 218.105301 -44.261421 . . . . . . . . . . . . . . . . . .27 218.175705 -44.218399 15.052(039) 14.325(006) 16.031(029) 14.641(006) 14.109(007) 13.930(008)28 218.168289 -44.201851 . . . . . . . . . . . . . . . . . .29 218.160904 -44.185322 . . . . . . . . . . . . . . . . . .30 218.066696 -44.166279 . . . . . . . . . . . . . . . . . .31 218.229706 -44.292591 . . . . . . . . . . . . . . . . . .32 218.062912 -44.154339 15.588(032) 14.885(005) 16.648(043) 15.191(006) 14.674(006) 14.474(007)33 218.084000 -44.143681 18.297(058) 17.534(014) . . . 17.930(012) 17.319(011) 17.081(010)34 218.090591 -44.139259 17.739(050) 16.692(008) . . . 17.168(008) 16.353(007) 16.056(006)35 218.115189 -44.143639 18.473(040) 17.633(015) . . . 18.026(014) 17.341(011) 17.100(011)36 218.050507 -44.161770 16.711(047) 15.877(006) 17.595(113) 16.258(007) 15.607(006) 15.319(007)37 218.015289 -44.142010 16.065(065) 15.369(006) 17.025(057) 15.694(007) 15.156(005) 14.981(007)38 218.223007 -44.184189 16.707(068) 16.111(006) 17.566(111) 16.368(007) 15.957(007) 15.811(008)39 218.219589 -44.150181 16.727(034) 15.770(006) 17.759(123) 16.229(009) 15.456(006) 15.135(007)40 218.259705 -44.146191 18.403(139) 17.674(014) . . . 18.035(012) 17.461(011) 17.261(011)41 218.190201 -44.145988 17.464(039) 16.742(009) . . . 17.053(008) 16.513(007) 16.326(007)42 218.245605 -44.146679 16.749(032) 16.019(006) 17.404(106) 16.334(007) 15.811(007) 15.623(007)43 218.259705 -44.146191 . . . . . . . . . . . . . . . . . .44 218.259399 -44.152180 18.892(104) 18.123(020) . . . 18.495(018) 17.848(013) 17.622(014)
Note —For convenience, table 6 shows the standard photometry of the local sequence stars. The natural Swope photometry is availablein the machine-readable format.
Table 7.
NIR local sequence photometry for SN2013aa ID α (2000) δ (2000) Y J H
103 218.152359 -44.199512 14.986(032) 14.661(020) 14.138(036)104 218.158768 -44.201530 14.610(035) 14.191(033) 13.492(022)105 218.156769 -44.231560 15.664(037) 15.351(037) 14.926(038)107 218.161942 -44.242981 12.919(030) 12.696(023) 12.370(028)108 218.125824 -44.245266 14.360(034) 14.062(023) 13.685(036)109 218.111145 -44.244247 14.173(042) 13.866(020) 13.488(020)110 218.106705 -44.232910 15.116(033) 14.852(075) 14.437(020)111 218.119980 -44.222977 12.416(032) 12.017(031) 11.369(116)112 218.109863 -44.206455 15.261(044) 14.908(028) 14.260(037)113 218.127289 -44.201855 16.274(052) 15.992(055) 15.752(073)114 218.135681 -44.211494 16.138(047) 15.897(062) 15.453(083)
Note —For convenience, table 7 shows the standard photometry of the localsequence stars. The natural du Pont photometry for J and H is availablein the machine-readable format. A. PHOTOMETRY OF SN 2013AA ANDSN 2017CBVIn this appendix we present the photometry for thetwo CSP-II objects used in this paper. The CSP-II was acontinuation of the
Carnegie Supernova Project (Hamuyet al. 2006), with the goal of obtaining NIR observations B MJD (days) V O b s e r v e d M a g n i t u d e g r i CSPHosseinzadeh+2017
Figure 11.
Comparison between the CSP photometry (bluecircles) and that of Hosseinzadeh et al. (2017b) (orange cir-cles). No S-corrections were applied to either dataset. at higher red-shift on average than in the CSP-I. Themethods for obtaining and reducing the optical and NIRphotometry are detailed in Phillips et al. (2019). ibling Type Ia Supernovae and therefore have slightly different colorterms (see table 5 of Phillips et al. (2019)), yieldingdifferent standard photometry. The filter functions andphotometric zero-points zp λ of the CSP-I and CSP-II natural systems are available at the CSP website. These can be used to S-correct (Stritzinger et al. 2002)the photometry to other systems. Figure 11 shows acomparison of the CSP photometry with that of Sandet al. (2018), showing very good agreement. B. BAYESIAN HIERARCHICAL MODEL FORINTRINSIC DISPERSIONIn order to quantify the intrinsic dispersion in the dis-tances to sibling host galaxies, we construct a simpleBayesian model. For each pair of siblings, we have thedifference in the distance estimate ∆ µ i,j and an associ-ated error σ ( µ i,j ) which we assume to be given by: σ ( µ i,j ) = (cid:113) σ ( µ i ) + σ ( µ j ) + σ ( sys ) i,j , (B1)where σ ( µ i ) and σ ( µ j ) are the formal errors in the dis-tances, including photometric uncertainties and errors inthe calibration parameters (Phillips relation, extinction,etc). We also add additional systematic errors σ ( sys ) i,j when comparing distances using different photometricsystems. These were estimated by fitting SNe Ia thatwere observed by two or more surveys and computingthe standard deviation of the differences in distance esti-mates. We summarize these in Table 8. We also includethe mean difference in distance (cid:104) ∆ µ (cid:105) , the standard devi-ation in the light-curve parameters ( s BV and E ( B − V )),the Pearson correlation coefficients, and the number ofSNe Ia used. The mean offsets were not applied to thedistances in table 3 nor in the analysis to follow, butrather are kept as part of the error in σ ( sys ). Generallyspeaking, the difference in distance estimates are moststrongly correlated with differences in estimates in theextinction.We model the true distribution of ∆ µ Ti,j as a normaldistribution centered at zero with scale σ SN . The ob-served ∆ µ i,j are modeled as normal distributions cen- SN 2013aa was observed with SITe3 and SN 2017cbv wasobserved with e2V. https://csp.obs.carnegiescience.edu tered at ∆ µ Ti,j with scale σ ( µ i,j ). Symbolically: σ SN ∼ U (0 , ∞ ) (B2)∆ µ Ti,J ∼ N (0 , σ SN )∆ µ i,j ∼ N (cid:0) ∆ µ Ti,j , σ ( µ i,j ) (cid:1) . We fit for the value of σ SN using Markov Chain MonteCarlo (MCMC) methods. Table 4 lists the results withdifferent subsets of the sibling SNe.0 Burns, Ashall, Contreras, Brown, et al.
Table 8.
Statistical Comparison between Surveys.
Surveys (cid:104) ∆ µ (cid:105) σ ( sys ) σ ( s BV ) σ ( E ( B − V )) ρ ( µ, s BV ) ρ ( µ, E ( B − V )) NCSP/LOSS − . .
16 0 .
07 0 .
06 0 . − .
69 32CSP/Swift − . .
19 0 .
07 0 .
06 0 . − .
69 11CSP/CfA − . .
09 0 .
07 0 .
07 0 . − .
67 61LOSS/Swift − . .
21 0 .
10 0 .
09 0 . − .
76 26LOSS/CfA 0 . .
16 0 .
10 0 .
08 0 . − .
71 87CFA/Swift − . .
15 0 .
07 0 .
08 0 . − .
93 19 ibling Type Ia Supernovae