Blast From the Past: Constraining Progenitor Models of SN 1972E
Aaron Do, Benjamin J. Shappee, Jean-Pierre De Cuyper, John L. Tonry, Cynthia Hunt, François Schweizer, Mark M. Phillips, Christopher R. Burns, Rachael Beaton, Olivier Hainaut
MMNRAS , 1–14 (0000) Preprint February 19, 2021 Compiled using MNRAS L A TEX style file v3.0
Blast From the Past: Constraining Progenitor Models of SN1972E
Aaron Do, Benjamin J. Shappee, Jean-Pierre De Cuyper, John L. Tonry, Cynthia Hunt, , François Schweizer, Mark M. Phillips, Christopher R. Burns, Rachael Beaton, , Olivier Hainaut Institute for Astronomy, University of Hawai‘i, 2680 Woodlawn Dr., Honolulu, HI 96822, USA Royal Observatory of Belgium, Ringlaan 3, B-1180 Ukkel, Belgium The Observatories of the Carnegie Institution for Science, 813 Santa Barbara St., Pasadena, CA, 91101, USA Giant Magellan Telescope (GMTO Corporation), 465 N Halstead St Carnegie Observatories, Las Campanas Observatory, Casilla 601, La Serena, Chile Department of Astrophysical Sciences, Princeton University, 4 Ivy Lane, Princeton, NJ, 08544, USA European Southern Observatory, Karl-Schwarzschild-Strasse 2, 85748, Garching bei München, Germany
February 19, 2021
ABSTRACT
We present a novel technique to study Type Ia supernovae by constraining surviving companions of historical ex-tragalactic SN by combining archival photographic plates and
Hubble Space Telescope imaging. We demonstrate thistechnique for Supernova 1972E, the nearest known SN Ia in 125 years. Some models of SNe Ia describe a white dwarfwith a non-degenerate companion that donates enough mass to trigger thermonuclear detonation. Hydrodynamicsimulations and stellar evolution models show that these donor stars will survive the explosion, and show increasedluminosity for at least a thousand years. Thus, late-time observations of the exact location of a supernova can con-strain the presence of a surviving donor star and progenitor models. We find the explosion site of SN 1972E byanalyzing 17 digitized photographic plates taken with the European Southern Observatory 1m Schmidt and 1 platetaken with the Cerro Tololo Inter-American Observatory 1.5m telescope. Using the
Gaia eDR3 catalog to determineSupernova 1972E’s location yields: α = 13 h m s ± s and δ = − ◦
40’ 8 . (cid:48)(cid:48) ± . (cid:48)(cid:48)
04 (ICRS). In 2005,
HST /ACS imaged the host galaxy of SN 1972E with the F W , F W , and F W filters covering the explosionsite. The nearest detected source is offset by 3.1 times our positional precision, and is inconsistent with the colorsexpected of a surviving donor star. Thus, the HST observation rules out all Helium-star companion models and themost luminous main-sequence companion model currently in the literature. The remaining main-sequence companionmodels could be tested with deeper
HST imaging.
Key words: astrometry – transients: supernovae – methods: observational
Type Ia supernovae (SNe Ia) have become a fundamental cos-mological probe for two reasons: they release enough energyto be seen from halfway across the observable Universe (e.g.Rodney et al. 2012)), and their peak luminosities are mono-tonically correlated with their decline rate, making themstandardizable candles using what is now called the Phillipsrelation (Phillips 1993; Hamuy et al. 1995; Riess et al. 1995).The distance measurements from SNe Ia can then be usedto measure cosmological parameters (Riess et al. 1998; Perl-mutter et al. 1999). However, as SNe Ia-based cosmologicalconstraints become tighter, systematic uncertainties have be-come the dominant source of error (e.g., Wood-Vasey et al.2007; Kessler et al. 2009; Guy et al. 2010; Conley et al. 2011).Despite the empirical success of the Phillips relation, theexact processes involved in the explosions of SNe Ia have been difficult to identify both theoretically and observation-ally (for a review see Livio & Mazzali 2018). It is commonlyaccepted that a SN Ia is the runaway thermonuclear explo-sion of a carbon-oxygen white dwarf (WD) in a close binarysystem, but there is no consensus on the nature of the binarycompanion or the mechanism inducing the explosion. Thereare two dominant models regarding the binary companion.The first is the double degenerate (DD) scenario where thecompanion is also a WD. The explosion is triggered by themerger or collision of the two WDs due to the removal ofenergy and angular momentum from the binary through ei-ther gravitational radiation (e.g. Tutukov & Yungelson 1979;Iben & Tutukov 1984; Webbink 1984), or perturbations bya third (e.g. Thompson 2011; Katz & Dong 2012; Shappee& Thompson 2013; Antognini et al. 2014) or fourth (Pejchaet al. 2013; Fang et al. 2018) body. The second is the sin-gle degenerate (SD) scenario where the companion is a non- © 0000 The Authors a r X i v : . [ a s t r o - ph . H E ] F e b Do et al. degenerate object: a main sequence (MS) star, a red giant(RG), a sub-giant, or a Helium (He) star (Whelan & Iben1973; Nomoto 1982a). The explosion is triggered when one ofseveral conditions are met: • Mass from the companion accretes onto the WD, push-ing the WD towards the Chandrasekhar limit, igniting a de-generate thermonuclear runaway near the center of the WD(Röpke et al. 2007; Ma et al. 2013). • He-rich matter is accreted and detonates, causing shock-waves to trigger a carbon detonation at sub-Chandrasekharmass (double detonation; Nomoto 1982a,b; Nomoto et al.1984). • A rapidly spinning WD that grows to a super-Chandrasekhar mass remains stable through rotational sup-port. If the WD loses angular momentum, it becomes unsta-ble and explodes (Piersanti et al. 2003; Hachisu et al. 2012;Benvenuto et al. 2015).Both progenitor models have been shown to lead to SNeIa-like explosions in the literature, and both may contributeto the sum of all SNe Ia. However, determining the fractionof SNe Ia from each channel is observationally difficult. TheDD scenario is thought to not leave behind any gravitation-ally bound remnant, while many versions of the SD scenarioare thought to have observable consequences tied to the donorstar. Some of these versions have been ruled out as a domi-nant channel based on non-detections: RG companions wouldproduce shocks as the ejecta impact the companion (e.g. Hay-den et al. 2010; Bianco et al. 2011; Bloom et al. 2012; Brownet al. 2012) or the circumstellar material lost by the com-panion (e.g. Chomiuk et al. 2012, 2016; Horesh et al. 2012;Margutti et al. 2012, 2014; Shappee et al. 2018; Cendes et al.2020), and there would be hydrogen in the nebular phase(e.g. Leonard 2007; Shappee et al. 2013a; Maguire et al. 2016;Shappee et al. 2018; Tucker et al. 2019, 2020). Although non-detections abound in the literature, without positive evidencesupporting the DD scenario, the best we can do is place upperlimits on the prevalence of the SD scenario. With only upperlimits, many systems remain consistent with at least one ver-sion of the SD scenario (e.g. U Sco and V445 Pup, Li et al.2011; over 20% of the sample analyzed by Maguire et al. 2013;3C 397, Yamaguchi et al. 2015; SN2018fhw/ASASSN-18tb,Vallely et al. 2019; SN2018cqj/ATLAS18qtd, Prieto et al.2020). Without knowing whether any given SN Ia is fromthe DD or SD scenario, it is difficult to determine what ef-fect the progenitor system has on the observable propertiesof SNe Ia. This systematic uncertainty translates directly toany SNe Ia-based cosmological constraint.Most observations seeking to detect the signature of a SDscenario occur during the first year of evolution of the su-pernova, but very late time observations can also reveal thenature of the progenitor system. In SD SN Ia explosions, bothMS and He-star donors will survive the explosion, but will beshock heated and stripped of mass (Marietta et al. 2000; Panet al. 2013; Shappee et al. 2013b). Despite this loss, a surviv-ing companion star will increase in luminosity as the enve-lope expands. For MS companions this overextended envelopethen collapses on a Kelvin-Helmholtz time-scale, leaving theMS companions significantly more luminous (10 − L (cid:12) )for a long period of time ( − years) but with a mod-estly decreased effective temperature. When mass is lost byablation (overheating), it is not physically possible to avoid these effects (Shappee et al. 2013b). Surviving He stars alsoincrease in luminosity but remain blue (Pan et al. 2013). Inboth of these cases, the SN ejecta is initially optically thickand will completely hide the companion for the first ∼ year,and then outshines the companion for the next ∼ years(Seitenzahl et al. 2009; Röpke et al. 2012; Graur et al. 2016;Shappee et al. 2017; Graur et al. 2018). However, at evenlater times (e.g. > 10 years) the ejecta is both optically thinand less luminous, allowing late-time observations to directlyconstrain the presence of a surviving companion.Thus far we have been restricted to search for survivingcompanions near the centers of supernova remnants (SNRs)to distinguish between progenitor models (e.g. Canal et al.2001; Ruiz-Lapuente et al. 2004; Schaefer & Pagnotta 2012;Ruiz-Lapuente 2018). Given SN Ia rates and SNR lifetimes,there are only a handful of young examples in the Galaxyand the Magellanic Clouds. A further challenge is to pre-cisely identify the expected location of the star. While TypeIa SNRs are generally spherical, there are asymmetries due tothe differing densities of the surrounding interstellar medium(ISM) and/or bubbles or clumps of gas and dust in the sur-rounding environment, which adds uncertainty to the mea-surement of the geometric center of the SNR (e.g. Edwardset al. 2012). Finally, crowding can be a significant challengeespecially when looking through the Galactic disk (Ruiz-Lapuente 2018). In this paper, we avoid these difficulties byanalyzing the nearest normal SN Ia for which imaging exists,both near the time of its initial detection and 33 years af-ter. Beyond this paper, our technique could be applied to thehandful of other close historical SNe Ia, such as SN 1895B,1937C, 1986G, 1991T, 1991bg, and 2011fe. Supernova 1972E, hereafter 72E, was discovered on May 14,1972 (UT) by Kowal (1972) in the outer regions ( ∼ "from the center) of NGC 5253 as part of the Palomar Su-pernova Search (Kowal et al. 1973) at a visual magnitude of ∼ . mag. 72E was promptly classified as a Type I supernova(Herbig et al. 1972; Barbon et al. 1972). At just 3.15 Mpc ( µ = 27.49 mag; Freedman et al. 2001), 72E is still the nearestnormal SN Ia since the previous SN in the same galaxy, SN1895B . Furthermore, 72E became the best observed super-nova with observations in the optical and infrared and wasthe first supernova observed with modern photoelectric de-tectors (e.g. Osmer et al. 1972; Lee et al. 1972; Walker et al.1972; Przybylski 1972; Cousins 1972; Kirshner et al. 1973a;Ardeberg & de Groot 1973; Kirshner et al. 1973b; Kirshner &Oke 1975; van Genderen 1975; Caldwell & Phillips 1989). 72Eis the 4th brightest extragalactic supernova (of any type) everobserved and helped shape our understanding of supernovaephysics.72E not only exploded in the outskirts of its host galaxy– thus reducing photometric uncertainties due to crowding– but it also suffered minimal host-galaxy reddening. We fitthe published light curves of 72E with the SNe Ia light curve The nearest extragalactic SN Ia occurred in the Andromedagalaxy in 1885, but as a peculiar Type I, it cannot be used tostudy SNe Ia (de Vaucouleurs & Corwin 1985)MNRAS , 1–14 (0000)
N 1972E fitter SNooPy (Burns et al. 2011), and found that 72E is con-sistent with having no host-galaxy reddening. This suggestsit exploded into a clean environment and, combined with itsnearness, makes it an exceptional target for late-time obser-vations.72E provides us with a unique opportunity. The recent SNIa 2014J (M82), the nearest Ia since 72E, will not be a goodcandidate for this kind of study because it is highly obscured,located in a crowded environment, and is already showing ev-idence of light echoes (Crotts 2015). SN Ia 2011fe (M101), thebrightest SN Ia since 72E, is more than twice as distant andthus the upper limits we can place on any surviving compan-ion will be over 1.5 mag less strict. The reason why othershave not already examined 72E is because the position wasnot known to better than ∼ " (Barbon et al. 2008).High quality images of 72E are found on 17 photographicplates, taken for photometric use by Hans-Emil Schuster withthe European Southern Observatory (ESO) 1m Schmidt tele-scope in 1973 and 1974 (Schuster 1980), requested by ArneArdeberg and one plate taken by James Hesser and MarioZemelman with the Cerro Tololo Inter-American Observatory(CTIO) 1.5m telescope. An appendix recounts the non-trivialprocess of locating and digitizing the astronomical plates,highlighting the importance of long-term record keeping anddata archiving.We used the upgraded Royal Observatory of Belguim(ROB) Digitiser 2 (De Cuyper et al. 2012) to digitize theregion around NGC 5253, archived as negative photographicdensities on original ESO Schmidt (ESOS) 300mmx300mmand 0.8mm thick bended glass plates. Each Schmidt platewas mounted on the Aerotech ABL3600 air-bearing XY-tableby pressing it pneumatically between a supporting and acounter-pressure plate in order to flatten the glass for pre-cise focusing of the emulsion. The air bearing XY-axes hada positional stability in time in both axes of better than 35nm. NGC 5253 was centered in the field of view of the 12bitgrey-scale CMOS BCi4-6600 camera (2208x3000 pixels of3.5 µ mx3.5 µ m) mounted on the new purpose build two-sided1:1 telecentric objective with no measurable (<0.1 µ m) opti-cal distortion in this 7.7 mm by 10.5 mm field of view, corre-sponding to the same area on the plates. With the plate-scaleof 67.4 arcsec/mm of the ESO Schmidt telescope this corre-sponds to 236 mas/pixel and a sky coverage of 521"x708".We used an adapted integration time to keep the sky den-sity just below the saturation limit to maximize the dynamicgrey-scale density range. Each digitized image is composedof 32 identical camera exposures in order to minimize the il-lumination, rolling shutter, and shot noise. We digitized theCTIO plate using a Gagne Porta-trace LED light panel, aCanon 5DS EOS DSLR, and a 100 mm Zeiss Makro-Planarlens. Table 1 provides the details of each plate. The digiti-zation of ESO plate 253 is incomplete, and this plate is notincluded in our analysis. In the next sections we describehow we used the plates to solve for the position of 72E inArchival Hubble Space Telescope ( HST ) images taken withthe Advanced Camera for Surveys (ACS) on December 27th,2005 (Proposal 10765; PI Zezas).
Our goal in this section is to determine the equatorial co-ordinates of 72E using the digitized photographic plates. Insection 3.1 we use
Gaia eDR3 to determine the equatorialcoordinates of stars near 72E. In section 3.2 we use a super-Gaussian function to determine the pixel coordinates of starson the digitized plates. In section 3.3 we correct for a system-atic effect known as magnitude equation. Lastly, in section 3.4we combine the previous sections to calculate astrometric so-lutions for each plate, and transform the pixel coordinates of72E into equatorial coordinates.
We use
Gaia eDR3 as the reference star catalog in our as-trometric solution for each photographic plate (Gaia Collab-oration et al. 2016, 2018). Therefore, we work in the Interna-tional Celestial Reference System (ICRS). Due to the ∼ year difference between the plate exposures and the J2015.5epoch of Gaia eDR3, proper motion at the ∼ mas/yearlevel can lead to positional differences at the arcsecond level.We use the Gaia
Archive’s epoch propagation function toapply proper motion corrections for each star in each plateaccurate to the day of the exposure. For Gaia eDR3 sourceswithout proper motions this function assumes a proper mo-tion of 0 milliarcsecond (mas) / year with an error of 0 mas/ year. These sources are not perfectly measured, but lack aproper motion and uncertainty because they failed a qualitycheck as detailed in section 4 of Lindegren et al. (2018). Forthis reason, we do not use these sources in our analysis. Forthe ∼ Gaia eDR3 sources within a 0.4 degrees of thecenter of NGC 5253, the median uncertainty in RA and Decis on the order of 0.3 mas at the J2015.5 epoch. However,the uncertainty in proper motion is multiplied by ∼ years,and becomes the dominant source of error at ∼ mas (0.22 HST pixels).
To make use of the reference catalog, we need to determinethe pixel coordinates of each source for which we have equato-rial coordinates. We begin with initial astrometric solutionsfrom Astrometry.net (Lang et al. 2010). We transform all
Gaia eDR3 sources within 0.4 degrees of NGC 5253 usingthese initial solutions to identify which
Gaia eDR3 sourceslie in each plate. Many sources are saturated, necessitatinga careful centering treatment. Additionally, the noise comesfrom two rounds of imaging. The first occurs during the ini-tial exposure, when photons from the sky strike the plate toproduce a negative image (dark objects on light background).The second occurs during the digitization of the negative pho-tographic density plate, when photons from a back-light passthrough the plate. The digitizing camera receives few pho-tons from the opaque regions, meaning the digitized imageswill have the best signal-to-noise ratio (S/N) when the pho-ton count is neither near 0, nor near the saturation limit. see https://gea.esac.esa.int/archive-help/adql/epochprop/index.html for details MNRAS , 1–14 (0000) Do et al.
We first attempted to fit each source with a 2 dimensionalelliptical Gaussian, truncating any value above the satura-tion limit. However, this functional form did not reproducethe smooth transition from the saturated core to the wings.To reproduce the saturated plateau in a smooth function, wecharacterize each source as a super-Gaussian function with 8free parameters. N ( x, y ) = A exp ( − ( a ( x − x ) + 2 b ( x − x )( y − y )+ c ( y − y ) ) P ) + k (1)where a = cos θ w x + sin θ w y b = − sin 2 θ w x + sin 2 θ w y c = sin θ w x + cos θ w y This functional form fits for the amplitude ( A ), the centerin pixel coordinates ( x , y ), the scale and orientation of theprincipal axes ( w x , w y , θ ), the shape parameter ( P ), and thelocal sky background ( k ). We manually evaluate each fit usingdiagnostic images like Figure 1, which shows an acceptablefit. The residual image retains an area of low noise due to thenegative imaging, but the area itself is consistent with no fluxafter the model is subtracted. We reject fits where the ampli-tude of the star is comparable to the noise, where there arestars separated by a distance less than the recorded seeing,or where the residual image exceeds the background noise.We do not fit sources within about an arcminute of NGC5253’s center because they would require accurate modelingand subtraction of the galaxy. Our fits of 72E are unaffectedby galaxy light. However, there are a few issues with some ofthe plates. For the CTIO plate we reject the fit for 72E be-cause it is heavily saturated, leading to a non-zero residual.Additionally the noise around the Northern edge of 72E inESO plate 252 appears to be contaminating the fit, resultingin a non-zero residual. Lastly, we reject the fits for ESO plates489, 493, and 530 because the amplitude of the supernova isconsistent with background noise in these images.We quantify our uncertainty in source centering by Markovchain Monte Carlo (MCMC) sampling the posterior probabil-ity distribution of our fit using the package emcee (Foreman-Mackey et al. 2013). Figure 2 shows the correlations be-tween the 8 free parameters, as well as their marginalizedone dimensional histograms. x and y , the second andthird columns respectively, describe the pixel position of thesource’s center in a 120 pixel by 120 pixel cutout of 72E.The figures of merit are not the values themselves, but theuncertainties of x and y . Our image centers are precise to ∼ . pixels in our digitized ESO images, about 5 mas giventhe plate scale of 0.236 arcsec/pix, and ∼ . pixels in ourdigitized CTIO images, about 8 mas given the plate scale of0.159 arcsec/pix. In photographic plates there is a correlation between asource’s intensity and its measured position on the plate.This effect is called magnitude equation. It means that anysingle astrometric solution is only accurate for sources of a magnitude similar to those used to create the solution. Italso means increasing the range of magnitudes used increasesthe root mean square of the residuals in the astrometric fit(RMS). While it is possible to carefully model and correct forthis effect (Eichhorn 1956; Jefferys 1962), we find it sufficientto restrict the stars used to those with magnitudes similar tothat of 72E in the U , B , and V filters of the plates. Withoutcorrecting for magnitude equation, the RMS in each plateroughly doubles.Although we cannot infer magnitudes directly from the 17ESO plates because there were no sensitometer spots, and wecannot transform photographic density to intensity, we cancombine photometric catalogs with estimates of 72E’s mag-nitude at each observation epoch to determine which starsto use in each plate to construct an accurate astrometric so-lution. For uniformity’s sake, we use the same methodologyfor the CTIO plate. The color-color transformations providedin Evans et al. (2018) allow us to convert Gaia passbandsinto the V passband, but there are degeneracies when tryingto convert to B magnitudes. For that reason, we use Pan-STARRS1 g and r photometry to constrain the B magni-tudes of stars in the field (Tonry et al. 2012; Chambers et al.2016). Neither the Gaia bandpasses, nor the Pan-STARRS1bandpasses can be transformed into U magnitudes, so weuse the second data release of the SkyMapper Southern Sur-vey (Onken et al. 2019). SkyMapper uses a set of filters thatroughly correspond to the SDSS system, except the ultravio-let band which is split into a u and v filter. We find that us-ing the SkyMapper u band as a rough proxy for the Johnson-Cousins U band allows us to infer positions for 72E consistentwith the other plates, suggesting that magnitude equation issufficiently corrected for.We estimate 72E’s magnitude at each observation epochby fitting a univariate spline to existing photometry (Przy-bylski 1972; Lee et al. 1972; Osmer et al. 1972; Cousins 1972;Jarrett & Eksteen 1973; Ardeberg & de Groot 1973, 1974;Kirshner & Oke 1975; van Genderen 1975). We subtract thespline from the photometric data, and use the standard devi-ation of the residual magnitudes as the uncertainty for eachfilter. We assume rough correspondence between UG1 glasswith an -O type emulsion and the Johnson U filter, GG385glass with an -O type emulsion and the Johnson B filter, andGG495 glass with a -D type emulsion and the Johnson V filter (Fiorucci & Munari 2002, 2003). Five of the observa-tions were made between 415 and 763 days after maximumlight, while the existing photometry only extends to about ayear after maximum light. For these five epochs we estimate72E’s magnitude using the nebular phase Phillips’ relationpresented in Tucker et al. (2020). The relation provides amethod of approximating magnitude to within 20% based onthe peak magnitude and the decline rate. Although the re-lation was calibrated with photometry from supernovae be-tween 150 and 500 days after maximum light, the consistentlinear increase in magnitude during the nebular phase of SNeIa allows us to reasonably assume the relationship holds tothe epoch of our latest plate (Shappee et al. 2017).We present the estimated magnitudes of 72E in Table 1.We note that in four of the plates, 72E’s magnitude is out-side the magnitude range of all nearby Gaia eDR3 sources.Fortunately, these are CTIO 2207, and ESO plates 489, 493,and 530, plates in which we could not adequately determinethe pixel coordinates of 72E.
MNRAS , 1–14 (0000)
N 1972E Figure 1.
A randomly selected example of a successfully modeled source. Top left: 120 pixel by 120 pixel cutout of a
Gaia eDR3 source onESO plate 328. Top right: super-Gaussian model fit. Bottom left: comparison of model (blue) and data (red) as a function of distance fromthe function center. The vertical axis defines 0 as the median flux with arbitrary linear scaling. Bottom right: residual upon subtraction.
We compute the gnomonic projection coefficients that mapthe ICRS coordinates of our reference star catalogs with thepixel coordinates we measure on each plate. We follow theSimple Imaging Polynomial (SIP) convention of Shupe et al.(2005), specifying 2 × • The effects of magnitude equation scale with the range of magnitudes used in the reference catalog, meaning we wantto use a limited number of stars. • Higher order polynomials can model more sophisticateddistortion patterns. • The number of coefficients in a 2-dimensional completepolynomial scales roughly with the square of the polynomialdegree, increasing the number of free parameters to be fit.To avoid overfitting, we use at least three stars for eachfree parameter in the distortion polynomial. We find that asecond order polynomial models the distortion field well, andwith only 6 free parameters, the 18 stars we require are typi-cally within a magnitude of 72E. Third, fourth, and fifth orderpolynomials reduce the RMS, but require about 30, 45, and63 stars respectively. There are not enough stars with mag-
MNRAS , 1–14 (0000)
Do et al.
Figure 2.
A corner plot demonstrating the covariances between the 8 free parameters of our model of 72E in a plate selected randomly;ESO 291. The amplitude A is relative to an arbitrary gain factor. The values of x and y are in pixels in a 120 pix by 120 pix cutoutroughly centered on 72E. Therefore, the expected values should be near 60. The angle θ is in degrees. The autocorrelation estimator in emcee indicated convergence in about 6 000 steps (Foreman-Mackey et al. 2013). nitudes comparable to 72E in our digitized images, meaningto avoid overfitting we would exacerbate the effects of mag-nitude equation.To determine the ICRS coordinates of 72E, we performa modified bootstrap resampling of the detected Gaia eDR3sources for each plate. For any given plate, we give 72E a ran-dom magnitude from a normal distribution centered at themagnitude listed in Table 1 with a standard deviation equalto 0.12 mag in U , 0.08 mag in B , 0.07 mag in V , and 0.2mag for plates 318 and 328. We define a bin size around eachrandomly generated magnitude such that at least 25 Gaia eDR3 sources have mean G band magnitudes within the bin. This number is chosen somewhat arbitrarily, but attempts toprovide at least 18 unique sources after selecting N sources atrandom with replacement, where N is the number of sourceswithin the magnitude bin. We adjust each source’s RA andDec based on the uncertainties listed in the Gaia eDR3 cat-alog, assuming the uncertainties are distributed normally.We create astrometric solutions according to the measuredpixel coordinates and adjusted ICRS coordinates, weightingsources according to both the inverse of their uncertainties,and how many times they were selected. Lastly, we use thesesolutions to transform the detected pixel coordinates of 72Eto ICRS coordinates, completing one iteration of the boot-
MNRAS , 1–14 (0000)
N 1972E strap resampling. We ran about 5 000 iterations of this pro-cess for each plate. Figure 3 shows the distribution of rightascensions and declinations for the bootstrapped positions of72E. Each plate produces a small spread in both RA anddec. This is a systematic uncertainty inherent to the creationof our astrometric solutions. It describes the combined ef-fect of the uncertainties in 72E’s magnitude, the positionaluncertainties in Gaia eDR3, and the combinatoric diversityof choosing with replacement. It is distinct from the RMSwhich describes the amount which any single source can de-viate from the astrometric solution.We find our final ICRS coordinates by calculating aninverse-variance-weighted average of each plate’s coordinatesfor 72E. The variance comes from a number of sources addedin quadrature: centering precision,
Gaia eDR3 uncertaintiessampled through bootstrapping, and the RMS, all of whichare listed in Table 2. The magnitude of the uncertaintiesvaries from plate to plate, but the RMS in the astromet-ric solution is always the dominant source of error. The in-verse variance weighted average position is 13 h m s ± s and − ◦
40’ 9 . (cid:48)(cid:48) ± . (cid:48)(cid:48)
04 (ICRS). For each ofthe 14 plates in which we could determine 72E’s ICRS coor-dinates, we present the RA and Dec determined using thatplate alone in Table 1.
Our late-time observation comes from Archival
HST /ACSimaging of NGC 5253, 72E’s host galaxy, taken on December27th, 2005. The integration times are 1759 seconds in theF435W filter, 2400s in the F555W filter, and 2360 secondsin the F814W filter. By this date the SN was 33 years old.The ejecta will have become optically thin after ∼ days(e.g. Mattila, S. et al. 2005), and the luminosity of a typicalSN Ia will have declined below that of a shock-heated MScompanion star after ∼ years (Shappee et al. 2017).The astrometric solution from the Space Telescope ScienceInstitute was created using the Hubble Source Catalog (HSC)version 3.1 (Whitmore et al. 2016). The HSC cross-matchessources detected in HST images with those detected in other
HST images, and those in several standard catalogs. No-tably, HSC version 3.1 was created before
Gaia eDR3 andeDR3, and does not use those catalogs. The relative astro-metric residuals between objects detected in multiple
HST
ACS/WFC images have a median of 7.8 mas, and a mode of3.4 mas. The absolute astrometric median error is reportedto be 10 mas for images near the
Gaia
DR1 epoch (2015.0),increasing by 5 mas/yr for earlier and later epochs due toproper motion. This would lead to an absolute astrometricerror of about 55 mas. We independently verify this by cross-matching the HSC sources with proper-motion subtracted
Gaia eDR3 sources within 0.2 arcseconds. Given 183 matchesbetween the two catalogs, the average offset would appear tobe 51 mas. However, this average includes the sources thatlack proper motions and uncertainties. We did not use thesesources for our analysis. Removing them brings the averageoffset from 51 mas to 38 mas, and weighting each source byits positional uncertainty brings the average offset to 32 mas.We add this last value in quadrature to the uncertainties inRA and Dec when determining pixel coordinates. With ourcalculated ICRS coordinates for 72E, the pixel coordinates of 72E are 3283.96 ± ± j9k501040_drc.chip1.fits . We present thepixel coordinates in the convention that assumes the centerof the lower left pixel has coordinates of (0.5, 0.5) rather than(1, 1). We show this location in the F435W/F555W/F814Wfalse color image in Figure 4.We note that kick velocity will not affect the pixel coordi-nates even though companion stars of SNe Ia have been pro-posed as sources of hypervelocity stars (Wang & Han 2009).At 3.15 Mpc, a tangential velocity of 800 km s − maintainedfor 33 years would cover an angular distance of only ∼ . milliarcseconds ( . pixels). We used
DOLPHOT to identify sources in the F435W, F555W,and F814W images (Dolphin 2000). and detected no sourceat the location of 72E. The nearest detected source is 2.9pixels (0.146 arcsecs) away, about 3.1 standard deviations. Itwas detected at a S/N of 3.5, 7.0, and 16.1 in the F435W,F555W, and F814W filters respectively. This source is in theHubble Source Catalog version 3.1 under MatchID 27647143.We derive a limiting magnitude based on the faintest stel-lar source detected by
DOLPHOT (as opposed to hot pixels andelongated or extended objects). Requiring a S/N of at least3, the faintest sources have instrumental Vega magnitudes of27.9 in F435W, 28.0 in F555W, and 27.4 in F814W. Increas-ing the S/N requirement to 5, the faintest sources are at 27.3in F435W, 27.4 in F555W, and 27.9 in F814W. This rulesout some progenitor models, as shown in Figure 5.Pan et al. (2013) provided evolutionary tracks for four mod-els of He donor stars in the SD scenario. The stars in thesemodels have initial masses equal to 1.25, 1.35, 1.4 and 1.8 M (cid:12) , and an initial metallicity of Z = 0.02. Thirty years after72E, any surviving He star would be contracting from a lumi-nous OB-like star to a hot blue-subdwarf-like (sdO-like) star.These stars would have luminosities between 4 000 and 14000 L (cid:12) depending on their initial masses. This would placeall models within our limiting magnitude in the F555W filter,and all but the least massive model within our limiting mag-nitude in the F815W filter. The lack of any detected sourcein the HST /ACS imaging indicates that any surviving com-panion cannot have been a He star with a mass of 0.7 M (cid:12) ormore.MS donor stars in the SD scenario would be overluminous,but less so than He stars. We rule out the most luminousMS star model, which comes from Shappee et al. (2013b).They explore a 1 M (cid:12) star with four parameters: mass loss,thermal energy added, truncation radius, and the timescaleof ablation and heating. Given the set that best matchedhydrodynamical simulations, a MS companion could have aluminosity of ∼ L (cid:12)
30 years after explosion. The appar-ent visual magnitude would be significantly brighter than ourlimiting magnitude. Conversely, the MS donor stars modeledin Pan et al. (2012) reach lower luminosities ( ∼ L (cid:12) ). Intheir models the expansion of the photosphere and increase ineffective temperature may be delayed by a century or more,meaning 72E’s companion would be far from its peak lumi-nosity. MNRAS , 1–14 (0000)
Do et al.
Table 1.
Digitized Plate Derived ICRS Coordinates of SN1972EPlate Date Exposure Seeing Glass Emulsion Equivalent Filter Magnitude RA (ICRS) Dec (ICRS)YY-MM-DD minutes Johnson-Cousins (13:39:XX) ( − V V .
706 09 . ESO 251 73-02-10 10 3-4 GG385 103a-O B .
711 09 . ESO 252 73-02-10 12 3-4 UG1 103a-O U V B .
712 09 . ESO 255 73-02-11 15 3 UG1 103a-O U .
704 09 . ESO 290 73-03-07 15 4 GG385 IIa-O B .
704 08 . ESO 291 73-03-07 25 4 UG1 IIa-O U .
701 09 . ESO 292 73-03-07 15 3-4 GG495 103a-D V .
711 09 . ESO 294 73-03-13 15 2 GG495 103a-D V .
706 09 . ESO 295 73-03-13 25 2 UG1 IIa-O U .
707 09 . ESO 296 73-03-13 15 2 GG385 IIa-O B .
705 09 . ESO 318 73-06-28 25 3 GG385 IIa-O B .
714 08 . ESO 328 73-07-02 25 2 GG385 II-O B .
709 08 . ESO 489 74-02-23 25 2 GG385 IIa-O B B B Uncertainties are 0.12 mag in U , 0.08 in B , and 0.07 mag in V . The Uncertainties in B increase to 20%, or about 0.2 mag forplates 318, 328, 489, 493, and 530. GG495 glass defines the lower wavelength limit of the transmission curve, while the emulsion defines the upper wavelength limit.We assume the CTIO 2207 plate used an a-D type emulsion to produce a V filter, but we cannot be sure.
Figure 3.
This normalized histogram of bootstrapped coordinates shows the precision of our astrometric solutions. Each plate wasresampled about 5 000 times. Left: The median right ascension is 13 h m s (ICRS). The standard deviation is about 56 mas, orabout 48 mas when multiplied by the cosine of the declination. Right: The median declination is − ◦
40’ 9 . (cid:48)(cid:48)
01 (ICRS). The standarddeviation is 42 mas.
Our objective was to determine the presence or absence ofa surviving companion for Type Ia supernova 1972E. Wehave analyzed 18 digitized plates to find the RA and Decof 72E. We used a super-Gaussian centering kernel to findpixel coordinates and the
Gaia eDR3 catalog for equatorial coordinates. There were six plates we did not include in ouranalysis: 72E was too saturated in the first plate, CTIO 2207,too faint in the last three plates, ESO plates 489, 493, and530, and did not subtract properly in ESO plate 252. Usingthe 12 remaining plates, the median values for 72E’s positionare α = 13 h m s and δ = − ◦
40’ 9 . (cid:48)(cid:48)
01 (ICRS). We
MNRAS , 1–14 (0000)
N 1972E Table 2.
Plate Specific Sources of UncertaintyPlate Equivalent Filter 72E centering
Gaia eDR3 Astrometric Solution WeightJohnson-Cousins mas mas mas Normalized to 100ESO 250 V B B U B U V V U B B B Figure 4.
Archival
HST /ACS image taken at the position of SN 1972E (Proposal 10765; PI Zezas) with the location of SN 1972E andindicated by the cyan circles. Their radii are 1 and 3 times the positional uncertainty of 72E. There is no source visible to the depth ofthe archival image 33 years after explosion. The nearest detected source is in the black circle Northwest of 72E. performed a bootstrap analysis to estimate our uncertainties.The standard deviations are 48 mas in right ascension, and 42mas in declination. Archival
HST /ACS imaging of NGC 5253from 2005 (Proposal 10765; PI Zezas) allows us to examinethe explosion site of 72E to a depth of 27.9 mag in F435W,28.0 mag in F555W, and 27.4 mag in F814W. We detect nosource at the given location. The nearest source is 0.146 arc-seconds away, about 3.1 standard deviations. Furthermore,this source is more luminous than all the MS models, andredder than all the MS and He-star models described in Panet al. (2012, 2013) and Shappee et al. (2013b). Thus, thissource is neither consistent with the location of 72E nor theexpected color from theoretical models and we rule it out asassociated with 72E.While previous studies in the literature have constrainedthe occurrence rate of MS companions (e.g. Hayden et al.2010; Bianco et al. 2011; Brown et al. 2012; Chomiuk et al. 2016; Tucker et al. 2020), the He star channel has provenmore difficult to test due its blue progenitor color and highsurface gravity reducing the effects of the SN ejecta inter-action. The exclusion of He-rich companions was attemptedfor the brightest SN Ia since 72E, SN 2011fe (Li et al. 2011),and the nearest since 72E, SN 2014J (Kelly et al. 2014). Forboth supernovae, pre-explosion imaging placed upper limitson the luminosity of any companion star. However, neitherstudy was able to fully exclude the He-rich channel as de-scribed in Liu et al. (2010). In contrast, we are able to fullyexclude this channel due to the increase in luminosity fol-lowing the explosion (Pan et al. 2013). Tucker et al. (2020)took a statistical approach to constrain SN Ia progenitor sys-tems by looking for lines in the nebular phase spectra of 111SNe Ia. This method relies on radiative transfer models thathave not been consistently performed for He star companions(Botyánszki & Kasen 2017; Botyánszki et al. 2018).
MNRAS , 1–14 (0000) Do et al.
Figure 5.
Color-Magnitude diagram showing what progenitor models are constrained by the Archival
HST image. The numbers in eachmodel give the post-explosion mass of the companion star in solar masses. The Pan et al. (2012) MS stars all use a mass of 1.17 M (cid:12) .There are five models that we would have been able to detect; the Helium star models and one main sequence model. The nearest twosources are redder than any of the main sequence models.
We rule out two progenitor systems. 72E was not the resultof a white dwarf and a He-rich companion as modeled by Panet al. (2012). Additionally, we rule out a MS companion asmodeled by Shappee et al. (2013b). However, we do not havesufficiently deep photometric data to rule out the MS modelsfrom Pan et al. (2012). The faintest of these models wouldhave a magnitude of 29.8 in the
HST /ACS Broad V (F606W)filter which the
HST would be able to detect at a S/N of 3.5in 10 orbits. Looking forward, the
James Webb Space Tele-scope ( JWST ) has 6.25 times the light gathering power of the
HST and its Near Infrared Camera (NIRCAM) has a finerscale at 32 mas pixel − (compared to HST /ACS ’s 50 maspixel − ) (Perrin et al. 2014). However, the angular resolutionof NIRCAM is not significantly better than that of HST /ACS(Nelan 2005).
JWST ’s key science goals involve studying highredshift, early universe phenomena, but do not necessitateincreased astrometric performance. Conversely, several of thekey science cases of the
Large UV/Optical/Infrared Telescope ( LUVOIR ) involve incredibly fine astrometric measurements(e.g. exoplanets causing stars to wobble, stellar proper mo-tions due to dark matter in dwarf galaxies). As such, the HighDefinition Imager (HDI) instrument has a precision astrome-try mode that will be able to make sub-microarcsecond mea-surements (The LUVOIR Team 2019). This far exceeds the accuracy of current astrometric catalogs, but future missionslike
GaiaNIR (Hobbs & Høg 2017) are planning to improvethe precision of the
Gaia eDR3 catalog by over an order ofmagnitude. With the light gathering power of
LUVOIR be-ing 40 times greater than that of the
HST , and with enoughastrometric precision to mitigate the effects of crowding, wewill be able to use the techniques in this paper on most SNeIa in galaxies within ∼
20 Mpc. Such SNe Ia include histor-ical supernovae like 1895B, 1937C, 1986G, 1991T, 1991bg,and 2011fe. Applying our technique to such a sample wouldplace firm constraints on the He star channel that has provendifficult to probe through other methods.Understanding the progenitor systems behind SNe Ia iscritical as modern surveys begin recording these transientevents with increasing depth and frequency. With systematicerrors dominating the uncertainties of SNe Ia based measure-ments, precision cosmology depends on our ability to fullyunderstand the process behind events like supernova 1972E.
ACKNOWLEDGMENTS
This work relied on the acquisition and accurate digitiza-tion of astronomical plates. For the former, we would like
MNRAS000
MNRAS000 , 1–14 (0000)
N 1972E to thank many people: Krzysztof Z. Stanek, Scott Gaudi,and Patrick Osmer from Ohio State University (OSU); JoséPrieto from the Astronomy Nucleus at Universidad DiegoPortales; Shrinivas Kulkarni and Jean Mueller from the Cal-ifornia Institute for Technology (Caltech); John Johnsonfrom Harvard University; Fernando Comerón, and Chris-telle Pluciennik from the European Southern Observatory(ESO); Oscar Duhalde from the Carnegie Observatories; Dai-nis Dravins and Arne Ardeberg from Lund Observatory; andSteve Heathcote and Malcolm Smith from the Cerro TololoInter-American Observatory (CTIO).A.D. would like to thank Michael A. Tucker, Jason T.Hinkle, Norbert Zacharias, and Chris Ashall, for reviewingthis work and providing valuable feedback, Or Graur andDavid Rubin for very insightful conversations. B.J.S. wouldlike to thanks Krzysztof Z. Stanek, Christopher S. Kochanek,and Jennifer van Saders for useful discussion during earlyphases of this project. B.J.S. is supported by NSF grantsAST-1908952, AST-1920392, and AST-1911074.This work has made use of data from the EuropeanSpace Agency (ESA) mission Gaia ( ), processed by the Gaia
Data Processing andAnalysis Consortium (DPAC, ). Funding for the DPAChas been provided by national institutions, in particular theinstitutions participating in the
Gaia
Multilateral Agree-ment.The Pan-STARRS1 Surveys (PS1) and the PS1 public sci-ence archive have been made possible through contributionsby the Institute for Astronomy, the University of Hawaii, thePan-STARRS Project Office, the Max-Planck Society andits participating institutes, the Max Planck Institute for As-tronomy, Heidelberg and the Max Planck Institute for Ex-traterrestrial Physics, Garching, The Johns Hopkins Univer-sity, Durham University, the University of Edinburgh, theQueen’s University Belfast, the Harvard-Smithsonian Cen-ter for Astrophysics, the Las Cumbres Observatory GlobalTelescope Network Incorporated, the National Central Uni-versity of Taiwan, the Space Telescope Science Institute, theNational Aeronautics and Space Administration under GrantNo. NNX08AR22G issued through the Planetary Science Di-vision of the NASA Science Mission Directorate, the Na-tional Science Foundation Grant No. AST-1238877, the Uni-versity of Maryland, Eotvos Lorand University (ELTE), theLos Alamos National Laboratory, and the Gordon and BettyMoore Foundation.The national facility capability for SkyMapper has beenfunded through ARC LIEF grant LE130100104 from the Aus-tralian Research Council, awarded to the University of Syd-ney, the Australian National University, Swinburne Univer-sity of Technology, the University of Queensland, the Uni-versity of Western Australia, the University of Melbourne,Curtin University of Technology, Monash University and theAustralian Astronomical Observatory. SkyMapper is ownedand operated by The Australian National University’s Re-search School of Astronomy and Astrophysics. The surveydata were processed and provided by the SkyMapper Teamat ANU. The SkyMapper node of the All-Sky Virtual Ob-servatory (ASVO) is hosted at the National ComputationalInfrastructure (NCI). Development and support the SkyMap-per node of the ASVO has been funded in part by Astron-omy Australia Limited (AAL) and the Australian Govern- ment through the Commonwealth’s Education InvestmentFund (EIF) and National Collaborative Research Infrastruc-ture Strategy (NCRIS), particularly the National eResearchCollaboration Tools and Resources (NeCTAR) and the Aus-tralian National Data Service Projects (ANDS).Based on observations made with the NASA/ESA
Hub-ble Space Telescope , and obtained from the Hubble LegacyArchive, which is a collaboration between the Space Tele-scope Science Institute (STScI/NASA), the Space TelescopeEuropean Coordinating Facility (ST-ECF/ESAC/ESA) andthe Canadian Astronomy Data Centre (CADC/NRC/CSA).
MNRAS , 1–14 (0000) Do et al.
APPENDIX A: ACQUIRING AND DIGITIZINGTHE PLATES
The analysis performed in this paper was partially conceived30 years ago in Caldwell & Phillips (1989), where they iden-tified a 24th magnitude object within 1 arcsecond of thelocation they calculated for 72E (13 h m s − ◦ . (cid:48)(cid:48) Palomar Discovery and 48-inch Schmidt Plates
Charles Kowal discovered 72E on a photograph that he hadobtained with the Palomar 18-inch Schmidt camera on 1972May 14 (UT), and he confirmed it via three photographicplates that he asked one of us (F.S.) to take for him with the48-inch Schmidt on 1972 May 16 (UT). He announced hisdiscovery in IAU Circular No. 2405 without specifying thetelescopes or photographic materials used (Kowal 1972). Thetwo dates he gave in that Circular (May 13 and May 15) werein local time (PDT), rather than in UT.The discovery photograph is a circular piece of film andcentered approximately on M83 (NGC 5236), a barred spiralgalaxy of type SBc II that was part of the 1972 supernovasearch in luminous nearby Sc galaxies (Kowal et al. 1973).NGC 5253 and 72E are only ∼ . ◦ away from M83 and liewell within the 8 . ◦ V ≈ . magwas an unforgettable experience. Shortly thereafter, Kowaldropped by to pick up his plates and go develop them. Thethree plates all bear the local date of 1972 May 15 and are:PS-17962, a 2 min exposure on Kodak 103a-O emulsion takenwithout filter; PS-17963, a 2 min exposure on 103a-O emul-sion taken with a Schott GG13 (now called GG385) filter;and PS-17964, a 5 min exposure on 103a-D emulsion takenwith a Wratten 12 filter. Leaving the mountain the next day,F.S. never saw the developed plates.One of these three 48-inch plates is shown in a “PalomarSkies” blog of 2010 Aug 6 and claimed to be the 72E discov-ery plate, yet it clearly is not. Upon request by B.J.S., Scott https://palomarskies.blogspot.com/2010/08/astrophoto-friday-supernova-1972e.html Gaudi contacted John A. Johnson (at Caltech at the time) insearch of an archive of Palomar plates. Johnson indicated thatJean Mueller maintained Palomar’s plate archive, verified byShrivinas Kulkarni in conversation with B.J.S. When B.J.S.contacted Jean Mueller, she reported that there were three48-inch Schmidt plates, but that none of them remained inthe archive. Unfortunately, their present location and ownerare unknown, but efforts to locate them will continue.
ESO 1-m Schmidt Plates
Arne Ardeberg, with the help of Mart de Groot (Ardeberg& de Groot 1973) immediately after the announcement of72E on May 17, 1972 started standard UBV photometric ob-servations at the Cassegrain focus of the 1m ESO telescopeon Cerro La Silla in Chile, using a cooled EMI 6256 photo-multiplier. In early October, 1973, Arne Ardeberg convincedHans-Emil Schuster, who was in charge of the new ESO 1mSchmidt telescope (ESOS), to take plates of NGC 5253 amongthose used for the testing and initial operation of the ESOSprior to the start of the ESO Southern QBS atlas program.During the first three nights a U, B, and V filter was usedand later on only a B filter in combination with differentplate emulsions. Near the end of February 1977, when ArneArdeberg left La Silla for Lund Observatory he arranged forthe 17 ESOS plates H.-E. Schuster had taken of 72E in 1973and 1974 to be sent by ESO to Lund Observatory for him.These plates were never used because there were no sensito-meter spots exposed on the plates, hence the density to in-tensity transformation could not be done, and also becauseof problems with the standardisation using standard stars(Christiaan Sterken, 2020, Private communication). The ESOScience Archive Facility lists twenty plates featuring NGC5253. Seventeen of them were taken less than two years af-ter 72E. B.J.S. contacted Fernando Comerón about access-ing the plates, but was told that they were in transit. TheESO archive was to be moved from the ESO headquartersin Garching, Bavaria, Germany to the Royal Observatory ofBelgium (ROB) in Ukkel, Brussels, Belgium. In early 2013B.J.S., then an astronomy graduate student at the Ohio StateUniversity, contacted Christelle Pluciennik re-inquiring as in-structed. She reached out to Jean-Pierre De Cuyper and puthim in direct contact with B.J.S.At this point a number of roadblocks emerged. In July, J.-P. De Cuyper asked for the cabinet containing the non-atlasplates numbered from 1 to 2000 in advance of the entire ship-ment. When the cabinet arrived, it did not contain the 72Eplates, in contradiction with the information ESO had givenbeforehand. Closer inspection of the archive revealed thatthe plates were on loan to ’*4’ and had not been returnedto ESO. The data division at ESO, one of the best obser-vatories in archiving most completely their observations andmeta data sets, had no record of persons for which non-atlasplates had been taken with the ESOS and who were markedas ’*number’ in the online database. Also the maps with en-velope copies of loaned Schmidt plates listed borrowers byname, not number and did not contain the 72E plates. Thus,this lead went cold until 2016 when B.J.S. contacted J.-P.De Cuyper again hoping things had settled. J.-P. De Cuyperrepeated his search, finding no new results as the logbooksthat were now at ROB do not mention anything about theperson for whom the 72E plates were taken. Only for the last
MNRAS , 1–14 (0000)
N 1972E plates an ’SN’ or one time an ’SN5253’ is mentioned as ob-ject. In desperation he redid a full search in the ESOS onlinedatabase for ’*4’ non-survey plates, that returned 605 entriescontaining a ’4’. It turned out that a series of 23 plates ofCarina also had been taken for ’*4’. The logbook entrees formost of these plates do not mention as usual the object ’Ca-rina’ next to the coordinates, but something that resembleda name like ’Adeley’ or ’Andeley’. Only for the last 3 sets ofplates taken on 1978/03/08,09,10 does the logbook show astandard entry, with the left side of the logbook mentioningan ’ARDEBERG’ for the first plate of that night and nextto the coordinates on the right side, the object ’Carine’. J.-P. De Cuyper phoned Arne Ardeberg, a professor emeritus ofLund University, who replied not to know anything of SN5253nor Carina ESO Schmidt plates. He advised to take contactwith Hans-Emil Schuster, who was in charge of the ESOS,now retired from ESO and living in his hometown Hamburgagain.Contacting Lund University, De Cuyper was directed toDainis Dravins who immediately found both series of ESOSplates in their plate-archive still in the original ESO shippingboxes. After waiting out the winter of 2016, Dravins shippedthe two boxes to Brussels. The digitization of the 17 platesof 72E was delayed by the upgrade of the ROB digitizer andserious illness, but was completed in 2019. It may be naturalto view historic observations as relics tracing the evolution ofastronomy, but even today they can provide scientific utilitythrough time-domain astronomy. If there is a lesson to belearned, it is that proper data management can prevent agreat deal of hardship in the future. We encourage such effortsto be undertaken and funded. CTIO Plates
In 2016, B.J.S. contacted Patrick Osmer, who worked withCTIO data (Osmer et al. 1972). Osmer could not remem-ber what became of the plates, but directed B.J.S. to SteveHeathcote, director of CTIO. Simultaneously, B.J.S. (atCarnegie at the time) contacted Mark Phillips who by co-incidence happened to have worked with the same CTIOdata in Caldwell & Phillips (1989). Phillips contacted Heath-cote, who quickly found plate number 2207, taken by Hesserand Zemelman with the CTIO 1.5-m telescope. B.J.S. foundrecords of several plates taken with the Curtis Schmidt 0.6-m telescope, the CTIo 1.5 meter telescope, and the VictorM Blanco 4 meter telescope, between 1972 and 1975 , butHeathcote reported large gaps in the archives surroundingtheir plate numbers. He claims Osmer borrowed many of theplates in these missing regions, and Osmer suggests that hemay have given them to a collaborator. Attempts to reachout to his collaborators have not been successful.Oscar Duhalde (Las Campanas Observatory) volunteeredto create an appropriate box to use when transporting the1.5-m plate. After an observing run to Las Campanas Obser-vatory B.J.S. brought the plate back to Carnegie Observato-ries in Pasadena to be digitized by C.H. B.J.S then returnedthe CTIO plate on a subsequent observing run to Chile. Plate logs at
References
Antognini J. M., Shappee B. J., Thompson T. A., Amaro-SeoaneP., 2014, MNRAS, 439, 1079Ardeberg A., de Groot M., 1973, A&A, 28, 295Ardeberg A., de Groot M., 1974, Photometry and Spectroscopy ofthe 1972 Supernova in NGC 5253. p. 103, doi:10.1007/978-94-010-2166-1_12Barbon R., Ciatti F., Ardeberg A. L., de Groot M. J. H., 1972,IAU Circ., 2411Barbon R., Buondi V., Cappellaro E., Turatto M., 2008, VizieROnline Data Catalog, p. II/283Benvenuto O. G., Panei J. A., Nomoto K., Kitamura H., HachisuI., 2015, The Astrophysical Journal, 809, L6Bianco F. B., et al., 2011, ApJ, 741, 20Bloom J. S., et al., 2012, ApJ, 744, L17Botyánszki J., Kasen D., 2017, ApJ, 845, 176Botyánszki J., Kasen D., Plewa T., 2018, ApJ, 852, L6Brown P. J., Dawson K. S., Harris D. W., Olmstead M., Milne P.,Roming P. W. A., 2012, ApJ, 749, 18Burns C. R., et al., 2011, AJ, 141, 19Caldwell N., Phillips M. M., 1989, ApJ, 338, 789Canal R., Méndez J., Ruiz-Lapuente P., 2001, ApJ, 550, L53Cendes Y., Drout M. R., Chomiuk L., Sarbadhicary S. K., 2020,arXiv e-prints, p. arXiv:2001.03558Chambers K. C., et al., 2016, arXiv e-prints, p. arXiv:1612.05560Chomiuk L., et al., 2012, ApJ, 750, 164Chomiuk L., et al., 2016, ApJ, 821, 119Conley A., et al., 2011, ApJS, 192, 1Cousins A. W. J., 1972, Information Bulletin on Variable Stars,700Crotts A. P. S., 2015, ApJ, 804, L37De Cuyper J., de Decker G., Laux U., Winter L., Zacharias N.,2012, in Ballester P., Egret D., Lorente N. P. F., eds, Astro-nomical Society of the Pacific Conference Series Vol. 461, As-tronomical Data Analysis Software and Systems XXI. p. 315Dolphin A. E., 2000, PASP, 112, 1383Edwards Z. I., Pagnotta A., Schaefer B. E., 2012, ApJ, 747, L19Eichhorn H. K., 1956, AJ, 61, 3Evans D. W., et al., 2018, A&A, 616, A4Fang X., Thompson T. A., Hirata C. M., 2018, MNRAS, 476, 4234Fiorucci M., Munari U., 2002, Ap&SS, 280, 77Fiorucci M., Munari U., 2003, A&A, 401, 781Foreman-Mackey D., Hogg D. W., Lang D., Goodman J., 2013,PASP, 125, 306Freedman W. L., et al., 2001, ApJ, 553, 47Gaia Collaboration et al., 2016, A&A, 595, A1Gaia Collaboration et al., 2018, A&A, 616, A1Graur O., Zurek D., Shara M. M., Riess A. G., Seitenzahl I. R.,Rest A., 2016, ApJ, 819, 31Graur O., et al., 2018, ApJ, 859, 79Guy J., et al., 2010, A&A, 523, A7Hachisu I., Kato M., Nomoto K., 2012, ApJ, 756, L4Hamuy M., Phillips M. M., Maza J., Suntzeff N. B., SchommerR. A., Aviles R., 1995, AJ, 109, 1Hayden B. T., et al., 2010, ApJ, 722, 1691Herbig G. H., Stephenson C. B., Wisniewski W., Lee T., WdowiakT., Michlovic J., Matchett V. L., Mayer E. H., 1972, IAU Circ.,2407Hobbs D., Høg E., 2017, Proceedings of the International Astro-nomical Union, 12, 67–70Horesh A., et al., 2012, ApJ, 746, 21Iben Jr. I., Tutukov A. V., 1984, ApJS, 54, 335Jarrett A. H., Eksteen J. P., 1973, Monthly Notes of the Astro-nomical Society of South Africa, 32, 54Jefferys W. H., 1962, AJ, 67, 532Katz B., Dong S., 2012, arXiv:astro-ph/1211.4584,Kelly P. L., et al., 2014, Astrophys. J., 790, 3MNRAS , 1–14 (0000) Do et al.
Kessler R., et al., 2009, ApJS, 185, 32Kirshner R. P., Oke J. B., 1975, ApJ, 200, 574Kirshner R. P., Willner S. P., Becklin E. E., Neugebauer G., OkeJ. B., 1973a, ApJ, 180, L97Kirshner R. P., Oke J. B., Penston M. V., Searle L., 1973b, ApJ,185, 303Kowal C. T., 1972, IAU Circ., 2405Kowal C. T., Zwicky F., Sargent W. L. W., Searle L., 1973, PASP,85, 427Lang D., Hogg D. W., Mierle K., Blanton M., Roweis S., 2010, AJ,139, 1782Lee T. A., Wamsteker W., Wisniewski W. Z., Wdowiak T. J., 1972,ApJ, 177, L59Leonard D. C., 2007, ApJ, 670, 1275Li W., et al., 2011, Nature, 480, 348Lindegren L., et al., 2018, A&A, 616, A2Liu W.-M., Chen W.-C., Wang B., Han Z. W., 2010, A&A, 523,A3Livio M., Mazzali P., 2018, Phys. Rep., 736, 1Ma H., Woosley S. E., Malone C. M., Almgren A., Bell J., 2013,ApJ, 771, 58Maguire K., et al., 2013, MNRAS, 436, 222Maguire K., Taubenberger S., Sullivan M., Mazzali P. A., 2016,MNRAS, 457, 3254Margutti R., et al., 2012, ApJ, 751, 134Margutti R., Parrent J., Kamble A., Soderberg A. M., Foley R. J.,Milisavljevic D., Drout M. R., Kirshner R., 2014, ApJ, 790, 52Marietta E., Burrows A., Fryxell B., 2000, ApJS, 128, 615Mattila, S. Lundqvist, P. Sollerman, J. Kozma, C. Baron, E. Frans-son, C. Leibundgut, B. Nomoto, K. 2005, A&A, 443, 649Nelan E., 2005, in Seidelmann P. K., Monet A. K. B., eds, As-tronomical Society of the Pacific Conference Series Vol. 338,Astrometry in the Age of the Next Generation of Large Tele-scopes. p. 59Nomoto K., 1982a, ApJ, 253, 798Nomoto K., 1982b, ApJ, 257, 780Nomoto K., Thielemann F. K., Yokoi K., 1984, ApJ, 286, 644Onken C. A., et al., 2019, Publ. Astron. Soc. Australia, 36, e033Osmer P. S., Hesser J. E., Kunkel W. E., Lasker B. M., McCarthyM. F., Landolt A. U., 1972, Nature Physical Science, 238, 21Pan K.-C., Ricker P. M., Taam R. E., 2012, ApJ, 760, 21Pan K.-C., Ricker P. M., Taam R. E., 2013, ApJ, 773, 49Pejcha O., Antognini J. M., Shappee B. J., Thompson T. A., 2013,MNRAS, 435, 943Perlmutter S., et al., 1999, ApJ, 517, 565Perrin M. D., Sivaramakrishnan A., Lajoie C.-P., Elliott E., PueyoL., Ravindranath S., Albert L., 2014, in Oschmann Jacobus M.J., Clampin M., Fazio G. G., MacEwen H. A., eds, Soci-ety of Photo-Optical Instrumentation Engineers (SPIE) Con-ference Series Vol. 9143, Space Telescopes and Instrumenta-tion 2014: Optical, Infrared, and Millimeter Wave. p. 91433X,doi:10.1117/12.2056689Phillips M. M., 1993, ApJ, 413, L105Piersanti L., Gagliardi S., Icko Iben J., Tornambe A., 2003, TheAstrophysical Journal, 583, 885Prieto J. L., et al., 2020, ApJ, 889, 100Przybylski A., 1972, IAU Circ., 2434, 1Riess A. G., Press W. H., Kirshner R. P., 1995, ApJ, 438, L17Riess A. G., et al., 1998, AJ, 116, 1009Rodney S. A., et al., 2012, ApJ, 746, 5Röpke F. K., Hillebrandt W., Schmidt W., Niemeyer J. C., Blin-nikov S. I., Mazzali P. A., 2007, ApJ, 668, 1132Röpke F. K., et al., 2012, ApJ, 750, L19Ruiz-Lapuente P., 2018, arXiv e-prints,Ruiz-Lapuente P., et al., 2004, Nature, 431, 1069Schaefer B. E., Pagnotta A., 2012, Nature, 481, 164Schuster H. E., 1980, The Messenger, 22, 7 Seitenzahl I. R., Taubenberger S., Sim S. A., 2009, MNRAS, 400,531Shappee B. J., Thompson T. A., 2013, ApJ, 766, 64Shappee B. J., Stanek K. Z., Pogge R. W., Garnavich P. M., 2013a,ApJ, 762, L5Shappee B. J., Kochanek C. S., Stanek K. Z., 2013b, ApJ, 765, 150Shappee B. J., Stanek K. Z., Kochanek C. S., Garnavich P. M.,2017, ApJ, 841, 48Shappee B. J., Piro A. L., Stanek K. Z., Patel S. G., MarguttiR. A., Lipunov V. M., Pogge R. W., 2018, ApJ, 855, 6Shupe D., Moshir M., Li J., Makovoz D., Narron R., Hook R.,2005, ASP Conf. Ser., 347, 491The LUVOIR Team 2019, arXiv e-prints, p. arXiv:1912.06219Thompson T. A., 2011, ApJ, 741, 82Tonry J. L., et al., 2012, The Astrophysical Journal, 750, 99Tucker M. A., Shappee B. J., Wisniewski J. P., 2019, ApJ, 872,L22Tucker M. A., et al., 2020, MNRAS, 493, 1044Tutukov A. V., Yungelson L. R., 1979, Acta Astron., 29, 665Vallely P. J., et al., 2019, MNRAS, 487, 2372Walker Christie Feasey Freeth Locher K., Bortle J. E., 1972,IAU Circ., 2413Wang B., Han Z., 2009, A&A, 508, L27Webbink R. F., 1984, ApJ, 277, 355Whelan J., Iben Jr. I., 1973, ApJ, 186, 1007Whitmore B. C., et al., 2016, AJ, 151, 134Wood-Vasey W. M., et al., 2007, ApJ, 666, 694Yamaguchi H., et al., 2015, ApJ, 801, L31de Vaucouleurs G., Corwin H. G. J., 1985, ApJ, 295, 287van Genderen A. M., 1975, A&A, 45, 429MNRAS000