THE CIRCUMSTELLAR ENVIRONMENT OF TYPE IA SUPERNOVAE

Raphael Ferretti

The Circumstellar Environment ofType Ia Supernovae

Raphael Ferretti

©Raphael Ferretti, Stockholm University 2017 ISBN print 978-91-7649-996-2ISBN PDF 978-91-7649-997-9 Printed in Sweden by Universitetsservice US-AB, Stockholm 2017Distributor: Department of Physics, Stockholm University Cover image: Dusty Supernova over Brunnsviken (by Tanja Petrushevska)

Abstract

Type Ia supernovae (SNe Ia) have proven to be extremely useful for measur-ing cosmological distances and were used for the discovery of the acceleratedexpansion of the universe. Although thousands of SNe Ia have been observedto date, many questions surrounding the physics of the explosions and thenature of their progenitor systems remain unanswered. An notable propertyof many SNe Ia is the relation between extinction due to dust and theircolour. For example SN 2014J, the nearest SN Ia in recent years, has anextinction relation which would be very unusual to observe in the Milky Way.One possible explanation to the peculiar extinction could be the presence ofcircumstellar (CS) dust surrounding the explosions. Incidentally, some pro-posed progenitor models of SNe Ia suggest that the explosions are surroundedby shells of matter, which could account for the unusual extinction.

CS gas would be ionised, if it is exposed to the intense ultraviolet (UV)radiation of a SN Ia. The research presented in this thesis focuses on thesearch for CS gas by observing the effects of photoionisation on absorptionlines commonly detected in optical spectra. Simple models suggest that thefrequently studied sodium doublet (Na I D) should significantly decrease oreven disappear if the gas is in the CS environment. Conversely, the absenceof variations implies that the absorbing gas clouds must be far from theexplosion, in the interstellar medium (ISM). To date, few SNe Ia have beenshown to have variable absorption lines, to which we have added another casewith SN 2013gh. Yet, we have also shown that most observations searchingfor variable absorption lines have been taken at too late phases, when mostCS gas will have already been ionised. Setting out to obtain the earliestpossible coverage of a SN Ia with high-resolution spectra, we have been ableto set strong limits on the presence of CS gas surrounding SN 2017cbv.

Along with evidence from other observational methods, these results haveshown that there is little matter in the CS environments of SNe Ia, suggestingthat the peculiar extinction likely results from the dust properties of theirhost galaxy ISM. Although the progenitor question cannot be resolved bythese observations, nondetections of CS gas point to models which do notdeposit large amounts of matter in their surroundings.

Sammanfattning

Typ Ia supernovor (SNe Ia) har visats sig vara väldigt lämpliga för att be-stämma avstånd i rymden och de har använts för upptäckten av universumsaccelererande expansion. även om tusentals SNe Ia har observerats hittills,finns det många obesvarade frågor om explosionens fysik och det ännu intekänt vilka stjärnsystem som genererar explosionerna. En ovanlig egenskap avsomliga SNe Ia är sambandet mellan extintion på grund av stoft och derasfärg. Till exempel SN 2014J, den närmaste SNe Ia de senaste åren, har enextintion som skulle vara mycket ovanligt att observera i vår galax. En möjligförklaring till denna egendomliga extintion kan vara närvaro av mycket stoftsom omger explosionen. Dessutom finns det en del teoretiska modeller förstjärnsystem som genererar SNe Ia där att stoft bildas kring systemen.

Gas, som omger en SN Ia, skulle joniseras, om det vore exponerat tillexplosionens ultravioletta strålning. Forskningen i den här avhandlingen fo-kuserar på att observera effekterna av jonisering i absorptionslinjer i optiskaspektra. Våra modeller förutspår att natrium absorption försvinner om ga-sen är nära explosionerna. Om absorptionslinjer istället inte förändras, måstegasen vara långt borta från explosionen. Hittills har man upptäckt få SNe Iasom har variabla absorptionslinjer, men vi hittade ett nytt fall i spektra avSN 2013gh. Dessutom har vi visat, att de flesta observationer som söker ef-ter variabla absorptionslinjer har tagits för sent för att observera jonisering.Därför hade vi som mål att få ett spektrum inom en vecka av en SN Ia ex-plosion. Vi uppnådde målet med SN 2017cbv, som visade att lite gas kundeomge explosionen.

Tillsammans med andra observationsmetoder vet man nu att lite materiakan omge SNe Ia. Därför måste ovanlig extintion komma från stoft som ärlångt borta från explosionen. även om våra observationer inte löser frågan omvilka stjärnsystem som genererar SNe Ia, understödjer våra resultat modellersom är inte omgivna av mycket gas och stoft.

I dedicate this thesis to my friend Felix who is aninspiration to everyone and hope he too can soon

pursue his dreams again.

List of Accompanying Publications

The following papers, referred to in the text by their Roman numerals, areincluded in this thesis.

PAPER I: The Rise of SN 2014J in the Nearby Galaxy M82A. Goobar, J. Johansson, R. Amanullah, Y. Cao, D. A. Perley,M. M. Kasliwal, R. Ferretti, P. E. Nugent, C. Harris, A. Gal-Yam, E. O. Ofek, S. P. Tendulkar, M. Dennefeld, S. Valenti,I. Arcavi, D. P. K. Banerjee, V. Venkataraman, V. Joshi,N. M. Ashok, S. B. Cenko, R. F. Diaz, C. Fremling, A. Horesh,D. A. Howell, S. R. Kulkarni, S. Papadogiannakis, T. Petru-shevska, D. Sand, J. Sollerman, V. Stanishev, J. S. Bloom,J. Surace, T. J. Dupuy, and M. C. Liu, ApJL, 784 (1), L12(2014).DOI: 10.1088/2041-8205/784/1/L12

PAPER II: Diversity in extinction laws of Type Ia supernovae mea-sured between 0.2 and 2 μmR. Amanullah, J. Johansson, A. Goobar, R. Ferretti, S. Pa-padogiannakis, T. Petrushevska, P. J. Brown, Y. Cao, C. Con-treras, H. Dahle, N. Elias-Rosa, J. P. U. Fynbo, J. Gorosabel,L. Guaita, L. Hangard, D. A. Howell, E. Y. Hsiao, E. Kankare,M. Kasliwal, G. Leloudas, P. Lundqvist, S. Mattila, P. Nu-gent, M. M. Phillips, A. Sandberg, V. Stanishev, M. Sullivan,F. Taddia, G. Östlin, S. Asadi, R. Herrero-Illana, J. J. Jensen,K. Karhunen, S. Lazarevic, E. Varenius, P. Santos, S. S. Srid-har, S. H. J. Wallström, and J. Wiegert, MNRAS, 453 (3),3300-3328 (2015).DOI: 10.1093/mnras/stv1505

PAPER III: Time-Varying Sodium Absorption in the Type Ia Super-nova 2013ghR. Ferretti, R. Amanullah, A. Goobar, J. Johansson, P. M. Vrees-wijk, R. P. Butler, Y. Cao, S. B. Cenko, G. Doran, A. V. Filip-penko, E. Freeland, G. Hosseinzadeh, D. A. Howell, P. Lundqvist,

S. Mattila, J. Nordin, P. E. Nugent, T. Petrushevska, S. Valenti,S. Vogt, and P.Wozniak, A&A, 592, A40 (2016).DOI: 10.1051/0004-6361/201628351

PAPER IV: Probing gas and dust in the tidal tail of NGC 5221 withthe type Ia supernova iPTF16abcR. Ferretti, R. Amanullah, A. Goobar, T. Petrushevska, S. Bor-thakur, M. Bulla, O. Fox, E. Freeland, C. Fremling, L. Hangard,and M. Hayes, forthcoming article at A&A, (2017).DOI: 10.1051/0004-6361/201731409

PAPER V: No evidence of circumstellar gas surrounding type Ia su-pernova SN 2017cbvR. Ferretti, R. Amanullah, M. Bulla, A. Goobar, J. Johansson,and P. Lundqvist, submitted, arXiv:1708.05394.

Reprints were made with permission from the publishers.

List of Publications Not Included

The following papers, referred to in the text by their letters, are not includedin this thesis.

PAPER A: High-redshift supernova rates measured with the gravita-tional telescope A 1689T. Petrushevska and 14 additional authors including R. Ferretti,A&A, 594, A54 (2016).DOI: 10.1051/0004-6361/201628925

PAPER B: iPTF search for an optical counterpart to gravitational-wave transient GW150914M. M. Kasliwal and 30 additional authors including R. Ferretti,ApJL, 824 (2), L24 (2016).DOI: 10.3847/2041-8205/824/2/L24

PAPER C: Localization and broadband follow-up of the gravitational-wave transient GW150914B. P. Abbott and 1576 additional authors including R. Ferretti(name misspelt in publication), ApJL, 826 (1), L13 (2016).DOI: 10.3847/2041-8205/826/1/L13

PAPER D: iPTF16geu: A multiply imaged, gravitationally lensedtype Ia supernovaA. Goobar and 33 additional authors including R. Ferretti,Science, 356 (6335), 291-295 (2017).DOI: 10.1126/science.aal2729

PAPER E: Testing for redshift evolution of Type Ia supernovae usingthe strongly lensed PS1-10afx at z=1.4T. Petrushevska, R. Amanullah, M. Bulla, M. Kromer, R. Fer-retti, A. Goobar and S. Papadogiannakis, A&A, 603, A136(2017).DOI: 10.1051/0004-6361/201730989

PAPER F: Estimating dust distances to Type Ia supernovae fromcolour excess time-evolutionM. Bulla, A. Goobar, R. Amanullah, U. Feindt and R. Ferretti,forthcoming article at MNRAS, (2017).DOI: 10.1093/mnras/stx2291

PAPER G: Early Observations of the Type Ia Supernova iPTF 16abc:Evidence for Strong Ejecta Mixing or Interaction with Dif-fuse MaterialA. A. Miller and 21 additional authors including R. Ferretti,submitted, arXiv:1708.07124.

Author’s Contributions

Most of my research focused on the analysis of high-resolution spectra, whichinvolved identifying absorption lines, normalising the continua, measuringequivalent widths and fitting profiles to the absorption lines. Then the funpart of describing the data followed. The results had to be interpreted anddescribed by models, such as the photoionisation model for CS gas. Althoughthe above tasks are those primarily presented in the accompanying research,a lot of work is done to be able to acquire the data in the first place. I wasinvolved in some of the steps.

To start with, supernovae need to be discovered and made public. Forthis purpose our research group is part of the intermediate Palomar TransientFactory (iPTF), where I was a regular scanner. The work encompassed visu-ally identifying and saving supernova candidates from nightly images takenwith the 48-inch telescope at Palomar Observatory. Promising candidateshad to be assigned for further observations and made public in AstronomersTelegrams (ATels) or the Transient Name Server (TNS) reports. In total Iwas involved in the publication of 22 ATels and 3 TNS reports. Although Iwas not involved in the discovery of these specific supernovae, it is the sameprocess that resulted in the discoveries of iPTF16abc (Papers IV and G) andiPTF16geu (Paper D). Through iPTF, I was also involved in the scanningfor optical counter parts of gravitational wave triggers by the Laser Interfer-ometer Gravitational-Wave Observatory (LIGO), which resulted in Papers Band C.

In order to obtain high-resolution spectra, we needed observing timeat telescopes with suitable instruments. On the Nordic Optical Telescope(NOT) we had access to the FIbre-fed Echelle Spectrograph (FIES) and wehad observing runs at the Very Large Telescope (VLT) using the Ultravioletand Visual Echelle Spectrograph (UVES). While I was involved in writing ourproposals and planing the observation runs, I also was principle investigator(PI) of the VLT proposal to observe iPTF16abc, which was approved inDirector’s Discretionary Time (DDT).

My specific contributions to each of the accompanying papers were:

PAPER I: The analysis of the absorption features in the high-resolutionspectra of SN 2014J. We obtained multiple epochs of high-

resolution spectra, of which I reduced the FIES spectra. Iplotted Figure 4 in the paper, which shows the absorptionlines identified, and computed the corresponding equivalentwidth values presented in the text.

PAPER II: The analysis of the available high-resolution spectra and theadaptations of the photoionisation model to determine an ex-clusion range for gas in the CS environment. In Figure 10, Iplotted the FIES spectra of SN 2012cg and used the absenceof variations for Figure 11, which defines the exclusion range.My contributions further include Figure 17, Table 11 and textsections which discuss narrow absorption lines.

PAPER III: The analysis of the high-resolution spectra of SN 2013gh andiPTF13dge. The work involved the reduction of the UVEShigh-resolution spectra and developing the photoionisation modelfurther. I led the analysis and write-up of this paper.

PAPER IV: The analysis of the high-resolution spectroscopy of iPTF16abc.I was PI of the DDT proposal through which we obtained thedata presented, and led the analysis and write-up of this paper.

PAPER V: The analysis of high-resolution spectra of SN 2017cbv. I ledthe analysis and write-up of this paper, which is currentlysubmitted for peer-review.

Contents

Abstract i

Sammanfattning iii

List of Accompanying Publications vii

List of Publications Not Included ix

Author’s Contributions xi

Abbreviations xv

List of Figures xvii

List of Tables xix

1 Introduction 21

2 Type Ia Supernovae 252.1 Supernovae types . . . . . . . . . . . . . . . . . . . . . . . 252.2 Thermonuclear explosions of white dwarfs . . . . . . . . . . 262.3 SN Ia progenitor models . . . . . . . . . . . . . . . . . . . . 28

2.3.1 Observational evidence . . . . . . . . . . . . . . . . 292.3.2 CS environments . . . . . . . . . . . . . . . . . . . . 30

3 Cosmology 333.1 General Relativity . . . . . . . . . . . . . . . . . . . . . . . 33

3.1.1 FLRW metric and Friedmann equations . . . . . . . 343.1.2 ΛCDM . . . . . . . . . . . . . . . . . . . . . . . . . 36

3.2 Supernova cosmology . . . . . . . . . . . . . . . . . . . . . 373.2.1 Extinction and reddening . . . . . . . . . . . . . . . 403.2.2 Low RV and CS dust . . . . . . . . . . . . . . . . . 42

4 Circumstellar Environments of SNe Ia 454.1 Narrow absorption lines . . . . . . . . . . . . . . . . . . . . 454.2 Photoionisation . . . . . . . . . . . . . . . . . . . . . . . . 474.3 Other observational evidence . . . . . . . . . . . . . . . . . 49

5 Conclusions 53

Acknowledgements lv

References lvii

Abbreviations

ATel Astronomer’s Telegram

Ca II Singly ionised calcium

Ca II H Calcium H line at λ 3968

Ca II K Calcium K line at λ 3934

CMB Cosmic microwave background

CS Circumstellar

DD Double degenerate

DDT Director’s Discretionary Time

FLRW Friedmann-Lemaître-Robertson-Walker (referring to the metric)

GR General relativity

K I Neutral potassium

ΛCDM Λ cold dark matter

LIGO Laser Interferometer Gravitational-Wave Observatory

Na I D Sodium D doublet at λλ 5890 and 5896

Na I Neutral sodium

NOT Nordic Optical Telescope

iPTF Intermediate Palomar Transient Factory

IR Infrared

ISM Interstellar Medium

SD Single degenerate

SN Supernova

SN Ia Type Ia Supernova

TNS Transient Name Server

UV Ultraviolet

UVES Ultraviolet and Visual Echelle Spectrograph

VLT Very Large Telescope

WD White dwarf

List of Figures

1.1 An artistic impression of a dusty supernova (credit T. Petru-shevska). . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

2.1 iPTF13dge and SN 2013gh, two SNe Ia as they appeared intheir host galaxies. Images reproduced from Paper III. . . . . 26

2.2 Illustrations of SN Ia progenitor models. In the single de-generate system on the left, a white dwarf is accreting gasfrom nondegenerate companion star. On the right, two whitedwarfs in a double degenerate system are in each others orbit,which slowly decays by gravitational wave emission. (Illus-trated by R.F.) . . . . . . . . . . . . . . . . . . . . . . . . . 28

3.1 B and V -band lightcurves of a typical SN Ia are characterisedby a stretch factor and colour with respect to a template. . . 39

3.2 Conceptual illustration of extinction. Photons are scatteredby dust out of the line-of-sight of an observer. (Illustratedby R.F.) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

3.3 Extinction curves with RV = 1.4 and 3.0. It can be seen thatthe colours of SN 2014J follow the unusually steep extinctioncurve. Plot adopted from Paper II. . . . . . . . . . . . . . . 41

3.4 Conceptual illustration of photons scattered by CS dust intothe line-of-sight of an observer. (Illustrated by R.F.) . . . . . 42

4.1 Geometric effects of an expanding photosphere. A small gascloud can initially absorb a larger fraction of the photonsemitted towards an observer to the right of the diagram.As the photosphere expands, more photons pass around thegas cloud and the fractional amount of photons reaching theobserver increases. Depending on the geometry of the gasclouds along the line-of-sight, this effect can lead to increas-ing or decreasing absorption line profiles in supernova spectra.(Illustrated by R.F.) . . . . . . . . . . . . . . . . . . . . . . 46

4.2 Na I D absorption profile of SN 2013gh. The most redshiftedfeature has decreased from the first epoch in blue to the lastepoch in red. For convincing, sceptical readers are referredto Paper III, from where the data is adopted. . . . . . . . . 47

4.3 Photoionisation of Na I gas shells with different radii. Thephase is given with respect to Bmax, with the explosion oc-curring at −20 days. The vertical axis shows the fractionalNa I column with respect to the initial value at explosion.Within 0.1 parsec, most Na I gas is ionised before −13 days. 48

List of Tables

4.1 List of SNe Ia with multi-epoch high-resolution spectroscopystarting before maximum brightness. The earliest epoch isgiven with respect to Bmax and the computed inner radiuslimit of photoionisation of Na I and Ca II gas are presented. . 49

4.2 Observational approaches to detect CS matter surroundingSNe Ia. The timing refers to the approximate phase of theexplosion, when the observations must be made and the sen-sitivity range indicates at what distances from the supernovaethe method can detect matter. The methods are describedin detail in the text. . . . . . . . . . . . . . . . . . . . . . . 51

1. Introduction

Type Ia supernovae (SNe Ia) have been of great importance to our currentunderstanding of cosmology as well as being interesting astrophysical phe-nomena by themselves. Since SNe Ia were first identified as a distinct type ofexplosion in the early 20th century, many thousands of examples have beenobserved and astrophysicists have taken advantage of their uniform proper-ties to use them to measure cosmological distances. Because the absolutebrightness of the explosions are known, SNe Ia are sought after cosmologicalstandard candles, with which the expansion of space-time can be measured.This resulted in the surprising discovery of the accelerating expansion rateof the universe [1; 2].

Despite being intensively studied, many fundamental questions surround-ing the physics of SNe Ia and their use in cosmology still remain. Currently,research in SNe Ia generally addresses one or both of the following questions:

• Can their standard candle nature be further improved to determinecosmological parameters more accurately?

• Can the illusive progenitor system, the star system from which theexplosions occur, be determined?

To answer both of these questions, astrophysicists have obtained large sam-ples of thousands of SNe Ia and try to study individual explosions in evergreater detail. To date, SNe Ia samples have reached sizes where systematicuncertainties dominate the statistical errors of cosmological measurements[3; 4]. Thus an improved understanding of the explosions themselves isnecessary to advance supernova cosmology. For this reason, much currentresearch, including that presented in this thesis, focuses on studying theproperties of individual SNe Ia.

While there are many approaches to address the two questions posedabove, this thesis aims at observationally characterising the circumstellar(CS) environment of SNe Ia. Specifically, the research presented analyseshigh-resolution spectra of individual supernovae in which narrow absorptionlines due to gas along the line-of-sight are visible. If gas is located closeto a supernova, or in the CS medium, it should be ionised by ultraviolet

21

Figure 1.1: An artistic impression of a dusty supernova (credit T. Petru-shevska).

(UV) radiation [5]. Photoionisation should cause detectable changes in theabsorption line profiles of the gas being ionised.

One of the initial motivations for this research was the suspicion that CSdust surrounding SNe Ia could be responsible for an unusual relation betweenreddening and extinction in cosmological samples [6; 7]. Studies of individualSNe Ia [for example in Paper II, 8], have revealed extinction relations, whichwould be very unusual to observe in our own galaxy, the Milky Way. Itwas thus fortunate that the nearest SN Ia since modern telescopes havebeen built occurred in the nearby galaxy M82, while we were investigatingthis subject. SN 2014J happened to also be a heavily reddened supernova[Paper I, 9] and have a peculiar extinction relation [10].

Much research has focused on determining whether the extinction ofSN 2014J originated from dust in the CS environment of the supernova orin the interstellar medium (ISM) of M82 [10–12]. Dust can be traced bygases, including neutral sodium (Na I) and singly ionised calcium (Ca II),which can be identified by narrow Na I D and Ca II H&K absorption linesin optical spectra. The spectra of SN 2014J showed many deep Na I Dand Ca II H&K absorption lines. If any of this gas was located in the CSmedium of the supernova, photoionisation should have led to detectablechanges in the profiles. While a potassium (K I) line of SN 2014J was found

22

to vary slightly [13], it was shown that the bulk of gas and dust along theline-of-sight must be located in the ISM of M82 [11; 12].

Detecting variations in absorption line profiles requires obtaining seriesof high-resolution spectra, which are time consuming and disruptive to theschedules of large telescopes. For this reason the published sample of time-series of high-resolution spectra of SNe Ia to date is small, with ∼ 23 exam-ples [8; 13–18]. In only four normal SNe Ia, absorption line variations havebeen detected [13–16]. To add a fifth example to this list, we discovered avariable absorption line in the Na I D profile of SN 2013gh [Paper III, 18].

It turns out that these detections do not firmly imply the presence of CSmatter surrounding the supernovae. In particular, geometric effects couldexplain the variations in each of these [19; 20]. Furthermore, in Paper III wehave been able to show that most published time-series of high-resolutionspectra have been obtained after all CS gas would have already been ionised.Thus the observations are insufficient to determine whether there is gasin the CS environment by searching for absorption line variations due tophotoionisation. To search for gas at distances where CS dust could beaccountable for peculiar extinction, the first spectrum of the time-seriesneeds to be obtained within a week after explosion. Setting this as thebenchmark, we aimed to discover and observe supernovae as early as possibleand successfully obtained high-resolution spectra of iPTF16abc [Paper IV,21] and SN 2017cbv [Paper V, 22]. In the latter case, we achieved the goal ofobtaining very early spectra, with which we were able to exclude the presenceof significant amounts of Na I and Ca II gas from the CS environment ofSN 2017cbv.

The limits of CS gas around SN 2017cbv happen to also provide cluesto its progenitor system. Some SNe Ia progenitor models suggest that theexplosions should be surrounded by shells of matter at the distances we aresensitive to with photoionisation [23; 24]. Stellar winds coming from the pro-genitor are predicted to blow cavities into the surrounding ISM and depositmatter at the edges of the cavities. Although the absence of variable ab-sorption lines is insufficient to exclude any progenitor model, it does provideevidence against models which predict gas in the CS environment.

In the following I will motivate studying photoionisation of CS gas aroundSNe Ia in order to explore the two questions posed at the beginning of theintroduction. The outline of the thesis1 will be as follows:

• Chapter 2 introduces SNe Ia and what we know about their physics.In the description of SNe Ia the open progenitor question is addressed

1Chapters 2 and 3 have been adopted from my licentiate thesis in a slightly modifiedform.

23

along with the different CS environments the models are expected tohave.

• Chapter 3 describes how SNe Ia are used as cosmological standardcandles. Starting from basic general relativity (GR), the current Λ–Cold Dark Matter (ΛCDM) model of the universe is described and therole SNe Ia have played in determining the parameters of the model.Finally, the focus is shifted to the unusual extinction relations thatseems to apply to cosmological SNe Ia samples and describe how CSdust can lead to a steep extinction law.

• Chapter 4 discusses the evidence for CS matter from various obser-vational methods, including photoionisation, the focus of most of myresearch.

• Finally, in the concluding Chapter 5, the current status of the illusiveprogenitor and peculiar extinction questions will be address.

24

2. Type Ia Supernovae

Unlike most astrophysical phenomena, supernovae are transient – they ap-pear and disappear again over a short time period of less than a few months.Historically, newly appearing stellar objects were described as novae. Todaya broad collection of different phenomena are classified as different types ofnovae, whereby these objects roughly attain stellar brightnesses. Recurrentnovae will be mentioned later on, as they are possible progenitor systems ofSNe Ia.

Supernovae were later recognised as a distinct group of phenomena whichappear to be several orders of magnitude brighter than classical novae [25].Although the modern description is based on observations of events in othergalaxies, there are several historical observations of supernovae in the MilkyWay [26]. Famous examples include:

• the supernova of 1054, resulting in the Crab nebula,

• Tycho’s supernova in 1572, which is believed to have been a SN Ia[27],

• and Kepler’s supernova in 1604 [28], which is also believed to havebeen a SN Ia [29] and the last supernova to have been observed in theMilky Way.

These supernovae must have been spectacular witness, while they were thebrightest objects in the night sky and maybe even visible during daylight.The images in Figure 2.1 depict how supernovae can appear, when they areobserved in other galaxies.

2.1 Supernovae types

Once it was recognised that supernovae are not a homogenous group, severaltypes were distinguished by their spectral and photometric properties [30].An early system of typing supernovae is still used to this day. Within thisclassification system, one group, termed type Ia supernovae, stand out dueto their incredibly uniform properties. SNe Ia are defined by their spectralcharacteristics, specifically, the absence of hydrogen emission and presence

25

N

E60"

iPTF13dge

N

E60"

SN2013gh

Figure 2.1: iPTF13dge and SN 2013gh, two SNe Ia as they appeared in theirhost galaxies. Images reproduced from Paper III.

of silicon absorption [31; 32]. Although there exist many more spectraltypes of supernovae, they are most easily distinguished from SNe Ia by theirexplosion mechanism.

Most supernovae (except for SNe Ia) have in common that their progen-itor stars fall out of equilibrium between radiation pressure and gravitationalcontraction [33]. This results in the gravitational collapse of their core to ei-ther neutron stars or black holes, while the outer layers of the star are blownoff in a burst of neutrinos. It is the blown off envelope of the explosion thatis subsequently detected in electromagnetic radiation. These supernovae arecollectively termed core collapse supernovae.

For reasons we will discuss below, it is believed that SNe Ia have a distinctexplosion mechanism from other types of supernovae and are probably thethermonuclear explosions of white dwarfs.

2.2 Thermonuclear explosions of white dwarfs

There is strong evidence for SNe Ia to be the explosions of white dwarf (WD)stars. WDs are the end state in stellar evolution of most main sequence starswhich have used up their nuclear fuel. The analysis of the rise time of thenearby SN 2011fe has shown that the object that exploded was very compactand smaller than < 0.02 Rsun [34]. Although this is also consistent with aneutron star, typical SNe Ia spectra show that the ejecta contains silicon,calcium and iron group elements, suggesting an explosive fusion process ofnuclear matter. In SN 2014J, the characteristic gamma ray emission lines

26

of the radioactive decay of 56Ni and 56Co have been detected [35–37]. 56Niis a fusion product of each two 12C and 16O in any order. Furthermore, theluminosity and duration of SN Ia lightcurves can be neatly described by theenergy release of this fusion process and the subsequent decay of 56Ni →56Co → 56Fe [38]. Since, carbon and oxygen are the main constituents ofcarbon-oxygen (C/O) WDs, this suggests that SNe Ia are the thermonuclearexplosions of these stars.

WDs are numerous, so why do only some explode? Thermonuclear fusionof carbon and oxygen will only occur when a critical temperature and pressureare reached. Once fusion is ignited in the core of a WD, a run-away processcommences which releases more energy than the binding energy of the entirestar. However, it is not obvious how the critical point to start thermonuclearfusion is reached!

WDs do not have an internal energy source. They merely radiate thethermal energy retained from the stars they evolved from. With no radiationpressure to counteract the gravitational compression, the electrons in theWD approach the state of a Fermi gas, where they occupy every availablequantum state. In this limit, the electron density is so large that the Pauliexclusion principle comes into effect, which limits the electrons from occu-pying degenerate states. When pressure is applied to a Fermi gas, electronsmust occupy higher energy states, which translates to an electron degen-eracy pressure. It is the electron degeneracy pressure which keeps the WDfrom further collapsing. In the limit in which electrons occupy states withrelativistic energies however, the degeneracy pressure rapidly decreases, be-cause there are no more energy levels to occupy. For a gravitating objectsupported by electron degeneracy pressure, this sets a finite mass limit atabout MCh ≈ 1.4 M�, known as the Chandrasekhar limit [39].

Thus a WD should collapse if its mass increases beyond the Chan-drasekhar limit. As the core density of the WD increases however, a numberof nuclear or particle reactions could step in. One conceivable scenario, isthat electron and protons undergo inverse β -decay and become neutrons[40]. In this scenario, the WD collapses to a neutron star, which itself issupported by neutron degeneracy pressure. Another possibility is the ignitionof nuclear fusion in the core of the white dwarf, which is the desired processto describe SNe Ia. Thus the progenitor system of a SN Ia is C/O WD,which in some way increases in mass to approach the Chandrasekhar limit,where a thermonuclear fusion ignites and the star explodes.

27

Figure 2.2: Illustrations of SN Ia progenitor models. In the single degeneratesystem on the left, a white dwarf is accreting gas from nondegenerate com-panion star. On the right, two white dwarfs in a double degenerate systemare in each others orbit, which slowly decays by gravitational wave emission.(Illustrated by R.F.)

2.3 SN Ia progenitor models

The explosion of a WD approaching the Chandrasekhar limit provides a neatdescription for the observed uniform properties of SNe Ia. If the ignitionoccurs once a critical amount of nuclear fuel is attained, the energy outputwill always be comparable. Thus, all proposed progenitor models involve aWD close to the Chandrasekhar mass with a companion star from which itaccretes matter or with which it merges.

To understand the diversity of progenitor models, it is helpful to considerthe evolution of a binary main sequence star system. As such a system ages,there are multiple points at which a SN Ia explosion could conceivably occur.The more massive star in the binary system will become a giant star earlierthan its companion. Radiation can blow off the outer gas layers of giant starsand if the radius of the star approaches the Roche limit, matter can also betransferred to the companion star. Provided the giant star has a carbonand oxygen core, a C/O WD will remain. The companion star, which mayhave become more massive in this process, will eventually reach a giantphase itself. At this stage the WD can accrete gas from its companion star,increasing its mass towards the Chandrasekhar limit. This system is calleda single degenerate (SD) model [41; 42] and is composed of a WD and anynondegenerate star, which can be a giant, sub-giant, helium star or even amain sequence star. The left panel of Figure 2.2 shows an illustration of a

28

WD accreting gas from its companion star.SD degenerate progenitor systems resemble recurrent novae, which have

the same stellar constituents and are known to exist in our Milky Way. Anexample is RS Ophiuchi, which reappears every ∼ 20 years and is believedto contain a WD close to the Chandrasekhar mass, which is accreting gasfrom a red giant star [43]. Because of the regular fusion explosions on theWD in a recurrent nova, a lot of gas can be blown off from its surface. It isthus debated whether the WD can gain enough mass through this process toreach the Chandrasekhar limit, but models indicate that it is possible [44].

An unusual, but probably observed scenario, is that the companion giantstar becomes large enough to engulf the WD in a common envelope. A smallnumber of SNe Ia, known as SNe Ia-CSM, for appearing to have very denseCS media, show strong interaction signals with the surrounding gas [45; 46].These are believed to be such common envelope or symbiotic progenitorssystems.

If the binary system outlives the giant star phase of the companion star,the companion star itself will become a WD. This system is called a doubledegenerate (DD) system for containing two WDs [47]. These will orbit eachother and spiral down over a long time period through gravitational waveemission. The right panel of Figure 2.2 shows an illustration of two WDsorbiting each other. Once the WDs get close enough to one another, theycan merge or the more massive WD disrupts the less massive companionand accretes the matter [48]. A DD progenitor system can be composedof two C/O WDs or C/O+He WD system. In the latter case helium fusionexplosions can occur on the surface of the C/O WD, when it is accretingfrom its companion.

Lastly, a unique progenitor scenario consists of a system of multiple (morethan two) stars, in which two WDs collide with one another [49].

2.3.1 Observational evidence

Currently, it is unclear which progenitor model produces normal SNe Ia.It is in fact possible that a combination of progenitor models produce theobserved population of explosions. An increasing number of outliers in spec-troscopic and photometric properties could be associated with some of theprogenitor models. To name two extremes:

• PTF11kx [46], while spectroscopically like a SN Ia, it also showednarrow hydrogen emission lines suggesting that it had a SD commonenvelope progenitor.

• SN 2003fg [50] was shown to have produced more than the Chan-

29

drasekhar mass of 56Ni, which points to a DD merger.

Nevertheless, the progenitor systems of normal SNe Ia are yet to be deter-mined and there is is observational evidence supporting both SD and DDmodels.

For example recent analyses of early lightcurves of some supernovae showan excess of blue and UV light [51; 52]. This has been interpreted as evidencefor a companion star being shocked by the ejecta of the supernovae [53].However, shocking a dense CS medium could have a similar effect [54].

Currently, the evidence for DD progenitor systems appear to outweigh:

• Pre-explosion images of nearby SNe 2011fe [55] and 2014J [PaperI, 9; 56] set strong limits on the maximum brightness of a potentialcompanion star, excluding any giant companions.

• There appears to be no indication of surviving companion stars inGalactic supernova remnants, such as Kepler’s supernova [57].

• DD systems have a long delay time from the time the progenitor starswere formed. The delay-time distribution between the cosmic starformation rate and SNe Ia rate points to the DD scenario, as the shortdelay time of the SD scenario cannot account for it [58].

Lastly, the density and matter content of the CS environment of SNe Iashould provide further indications to what progenitor system shaped them.

2.3.2 CS environments

There are different predictions for the CS environments around SD and DDprogenitor systems. Roughly speaking, SD progenitors resemble recurrentnovae, which are WDs accreting from nondegenerate companion stars. A SDprogenitor system will likely have powerful winds, sustained by the radiationof the companion star and from fusion reactions on the surface of the WD.Thus, such a system is predicted to produce a lot of outflowing matter, whichwould blow cavities in the surrounding ISM. At the edges of the cavities,where the wind hits the ISM, matter will be deposited, which effectivelyresults in a shell of matter surrounding the progenitor [23]. The cavities arepredicted to have the sizes up to ∼ 10 parsec.

Interestingly, a cavity is not a unique feature of a SD progenitor system.In C/O+He WD binaries, helium is accreted onto the surface of the primaryWD, where fusion explosions could blow off matter from the surface. Sucha DD progenitor system could produce similar cavities to those of SD pro-genitors. However these cavities are predicted to have smaller radii of up to∼ 1 parsec [24].

30

Finally, DD mergers are thought to have normal ISM surrounding them,because they do not sustain strong stellar winds. Nevertheless, tidal inter-actions shortly before the supernova explosion can enrich the immediate CSmedium with matter [59].

In Chapter 4, the observational evidence for matter in the CS environ-ments of SNe Ia will be returned to.

31

32

3. Cosmology

SNe Ia have been of great importance to cosmology. As described below,their uniform properties make it possible to use them cosmological standardcandles, with which distances in the universe can be accurately determined.To understand the methodology and the significance of using SNe Ia asstandard candles, a brief description of general relativity and our currentcosmological model are necessary.

3.1 General Relativity

Einstein noted that gravity is indistinguishable from inertial motion, a con-cept which is now known as the weak equivalent principle. He furthermoredefined a strong equivalence principle, which describes that an observer canlocally not distinguish between being accelerated or standing on the surfaceof a gravitating object such as the Earth. Similarly, floating far from a grav-itational object in space cannot be locally distinguished from a free fall ina gravitational field. From this Einstein conjectured that rather than beinga force, gravity is simply inertial motion in a curved space-time [60]. Anobject follows a so called geodesic, which an observer in the same referenceframe experiences as a free fall. Viewed from a external frame, the motionhowever can appear curved, if a flat spacetime is presumed. In this context,the elliptical orbits of the planets around the sun, are in fact closed geodesicsthat connect to themselves.

In a given spacetime, a line element, ds2, between two points with space-time separation dxμ is defined as

ds2 = gμνdxμdxν , (3.1)

where μ runs from 0 to 3 and the Einstein summation convention, the sum-mation over repeated indices, is used. The metric gμν defines how distancesare measured in a given spacetime, taking its curvature into account.

A theory of gravity needs to relate the spacetime curvature to the con-tained energy density. This must be done by relating the spacetime geometryto the stress-energy tensor Tμν . The stress-energy tensor has two basic prop-erties, it is symmetric (Tμν = Tνμ) and divergenceless (∇νT μν = 0). These

33

properties must be matched by a geometric construction to describe a grav-itational theory. The simplest solution found by Einstein is a tensor Gμν ,which is named after him. The derivation of Einsteins gravitational fieldequation is lengthy, but results in the following equation:

Gμν = Rμν − 12

gμνR+Λgμν =8πGc4 Tμν , (3.2)

where G is the gravitational constant, c the speed of light and the Ricciscalar is the contraction of the Ricci tensor R = gμνRμν , which in turn isthe contraction of the Riemann curvature tensor Rμν = Rα

μαν . The Rie-mann tensor can be computed for a given metric. Equation (3.2) includes agravitational constant term Λ, which will be discussed below.

3.1.1 FLRW metric and Friedmann equations

Until the begining of the 20th century, when GR was developed, it was stillgenerally believed that the universe is in a static state, implying that theseparation between distant objects is only determined by the kinematics ofthe objects themselves. This however cannot be the case in a universethat is filled with matter, which curves spacetime. For this reason, Einsteinbrought a cosmological constant into the discussion, which could (with somefine tuning) counteract the gravitational effect of matter [61]. The gravita-tional field equations allow for a cosmological constant term denoted Λ inEquation (3.2).

Around the same period, methods in astronomical observation were ad-vancing significantly as well. It was discovered that objects beyond our MilkyWay, then called nebulae, but now known as separate galaxies, are recedingfrom us [62; 63]. The local expansion rate is today known as the Hubbleconstant H0 which relates the distance of an object to its recession velocityby

v = H0d. (3.3)

The velocity of this recession could be measured from the redshifted spectraof these objects, where redshift z is defined through the ratio between theobserved, λobs, and emitted, λ0, wavelengths,

z =λobs

λ0−1 (3.4)

It was thus found that the local universe is expanding.In an obviously non-static universe, a cosmological constant appeared no

longer necessary and the notion was abandoned. Depending on the exact

34

energy density, the universe is bound to expand for ever, or if the density isgreater than a critical value

ρc =3H2

08πG

, (3.5)

eventually collapse. The ratio of the true matter density to the critical matterdensity today, ΩM ≡ ρM/ρc, is thus an interesting parameter to measure,which should determine the ‘fate’ of the universe. That is the case unlessthere is a non-zero cosmological constant term.

The gravitational field Equations (3.2) have simple analytic solutions ifone assumes a space geometry which is isotropic and homogenous. We thusrequire a metric which is homogenous, isotropic, that can expand in time andcan have curvature. Such a metric is the Friedmann-Lemaître-Robertson-Walker [FLRW, 62; 64] metric with line element

ds2 = c2dt2−a(t)2(

dr2

1− kr2 + r2dθ + r2 sin2 θdφ 2), (3.6)

in spherical coordinates t,r,θ and φ . Here a scale factor a(t), which is afunction of time only, and a curvature parameter k have been defined. Thecurvature can take different values, k =−1,0,+1 depending on the whetherthe overall geometry of the universe is open, flat or closed, respectively.

To solve the gravitational field equations for the above metric, one makesthe assumption that the contents of the universe can be described as perfectfluids. This implies that the energy density, T00 = ρ and pressure Tii = pare the only non-zero terms in the stress-energy tensor, in the rest-frameof the fluid. Since both gμν and Tμν are diagonal, there are four possibleequations that can be found, of which two are independent. These equationsare known as the Friedmann equations,

H2 =

(aa

)2

=8πG

3ρ− kc2

a2 +Λc2

3(3.7)

andaa=−4πG

3

(ρ +

3pc2

)+

Λc2

3. (3.8)

Above, we have defined H = aa which is the expansion rate of the universe

at a given time. The local expansion rate denoted H0 = H(t0) is the same asdefined in Equation (3.3). Equation (3.7) can further be rewritten in termsof different energy densities,

H2 = H20

(ΩR

a4 +ΩM

a3 +Ωk

a2 +ΩΛ

), (3.9)

35

where one distinguishes between the radiation, matter and vacuum (or cos-mological constant) energy density ΩR, ΩM and ΩΛ, respectively, and Ωkcan be viewed as the potential energy contained in spatial curvature. TheFriedmann equations have very simple solutions if we assume that the cur-vature k = 0. For reasons not discussed here it is known from data of thecosmic microwave background [CMB, 65], that the energy densities of theuniverse are consistent with zero curvature. To understand the different lim-its of the solutions to Equations (3.7) and (3.8), we will define the equationof state parameter w, which relates the energy density to pressure,

pc2 = wρ. (3.10)

One can show that

a(t)∼{

t2

3(1+w) , if w >−1,eHt , if w =−1

(3.11)

are solutions to Equation (3.8). The equation of state parameter can takethe values,

w =

⎧⎪⎨⎪⎩

0, for pressureless matter,1/3, for radiation,−1, for Λ.

(3.12)

The overall equation of state parameter of the universe will take the valuethat corresponds to the dominant component of the fluid.

3.1.2 ΛCDM

Much effort has been invested into measuring the cosmological parametersusing dedicated ground and space based telescopes. The first direct ev-idence for a non-zero cosmological constant came from measurements ofthe luminosity distances of supernovae. Further cosmological probes suchas the CMB power spectrum and spatial correlations in the distribution ofvisible matter, known as baryonic acoustic oscillations (BAOs) complementthe supernova results. Today, the best concordance model suggests thatΩΛ ≈ 0.69 and ΩM ≈ 0.31, implying that the expansion of the universe isaccelerating [65]. This model is known as the Λ-Cold Dark Matter (ΛCDM)model, for being dominated by ΩΛ. The CDM extension arrises from theobservation that most of ΩM must be made of matter which has yet to bedetected beyond its gravitational effects.

The existence of a cosmological constant is equivalent to a constantenergy density of space. A physical fluid that with w=−1 could be described

36

by a vacuum energy. Quantum field theory predicts a vacuum energy, dueto the non-zero ground state energy of all particles. However, the energydensity computed from quantum field theory requires the choice of arbitraryintegration limits, which results in a very different value than the measuredvacuum energy density [66].

Alternative dark energy models could give rise to a corresponding energydensity that would vary with time, similar to matter or radiation. A redshiftdependence can be parameterised in several different ways. One ansatz [67]typically used is

w(z) = w0 +waz

1+ z. (3.13)

Today, the best fit value of the equation of state parameter is consistent withw0 =−1, or a cosmological constant, while the the redshift dependent wa isconstrained around zero [68]. Future measurements using SNe Ia should beable to constrain the equation of state better.

3.2 Supernova cosmology

Due to the well constrained distribution of absolute brightness, SNe Ia wererecognised as potential standard candles early on [32]. The photometric andspectroscopic evolution of SNe Ia are so uniform that one can determinerelative luminosity distances to the explosions. Using a nearby sample ofSNe Ia in host galaxies to which one has independent distance measures(for example by using Cepheid variable stars), it was found that the peakbrightness of the explosions are standardisable by determining only a fewadditional parameters [69]. To use SNe Ia for cosmology, it is presumed thatthe standardisability of SNe Ia extends to distances beyond which indepen-dent distance measures exist. The simplicity of this approach is that it isentirely based in empirical correlations and no additional physical knowledgeof the explosions are required.

The standardisation of SNe Ia lightcurves involves two simple corrections.The lightcurves and spectra have distribution in their respective duration andoverall colour, which correlate neatly with the peak luminosity of the SNe.The lightcurve of a SNe Ia typically has a well defined peak, denoted Bmax

in Figure 3.1. Furthermore, a lightcurve width can be defined which one candefine as a temporal stretch parameter s. Viewed through different filters,colour differences at maximum light can further be defined. In general,broader lightcurves appear brighter and supernovae that appear redder arefainter.

In order to measure the luminosity distance to an observed object, onemust know its intrinsic luminosity (L). The observed flux of an object is

37

defined

F =L

4πd2L, (3.14)

where the luminosity distance is defined

dL(z) = r(1+ z)a(t0)≈ czH0

, (3.15)

in the local universe, with the comoving distance r. The brightness of an as-tronomical object through some filter X is typically measured in magnitudesdefined as

mX(z) = MX +5log10 dL(z)+25+KX +AMW , (3.16)

where MX is the intrinsic absolute magnitude of the object (the magnitude ifit were at a distance of 10 parsec) and dL is measured in Mpc. Furthermore,so called K-corrections KX and a correction for dust extinction in the MilkyWay, AMW , must be applied. An observation through a given band (suchas an optical filter) is like a weighted integration over the spectrum. Sincethe SNe observed are redshifted, the section of the spectra seen through aparticular band depends on the redshift of the object. The intrinsic spectra ofSNe Ia are well known, therefore K-corrections can be applied to compensatefor the redshift dependent shifts in photometry.

As mentioned earlier, supernovae have a distribution in stretch andcolour, both of which are related to the absolute peak magnitude MX . Therelations of stretch and colour to peak magnitude are determined empirically.Thus a possible parameterisation is,

MX = Mavg +α(s−1)−βc, (3.17)

where Mavg, α and β are constant parameters determined from larger sam-ples and s and c are stretch and colour parameters, respectively [70]. Havingapplied all the corrections to a sample of SNe, a Hubble diagram showingthe expansion rate of the universe can be constructed.

Today, sample sizes are large enough for the errors in the cosmologicalfits to be dominated by the systematic errors in using SNe Ia as distanceindicators. Next generation surveys, such as the Large Synoptic SurveyTelescope [LSST, 71] aim at constraining the equation of state parameter.For these tasks systematic errors must be reduced, as well as identifyingpossible redshift dependent properties.

Although the colour distribution is in part intrinsic, it is known that manySNe are reddened by dust in their host galaxy. It is the reddening due todust that we will focus on below.

38

3.2.1 Extinction and reddening

Correcting for reddening is one of the main steps applied to ‘standardise’SNe Ia. When an object is observed, dust along the line-of-sight absorbsand scatters part of the light. The light that does not arrive at the ob-server, is said to be extinct. Figure 3.2 conceptually illustrates extinctiondue to dust. The extinction in a particular band X (e.g. the filter in aninstrument) is typically defined as the difference between the intrinsic andobserved magnitude of a source,

AX = mX ,obs−mX ,intr (3.18)

Extinction results in reddening, because blue light is preferentially scatteredor absorbed by dust, resulting in an overall redder appearance. The exactwavelength-dependence of the extinction depends on the dust compositionand is only known from empirical measurements. Because the extinction lawis typically not known, one generally focuses on the bands through which anobject is being observed. For example, reddening across B and V bands isdefined by comparing the difference in extinction between them,

E(B−V ) = AB−AV . (3.19)

An extinction law defines the extinction as a function of wavelength. Toparameterise the extinction law around a given band, the ratio

RV =AV

E(B−V )(3.20)

is typically defined. In general, smaller RV values are said to have a steeperextinction law, implying that bluer are more extinct than redder wavelengths.

The difficulty in determining the extinction of any given object, is thatthe intrinsic brightness and colours are often not known. In the case ofSNe Ia, we presume to know the intrinsic properties from a sample of nearbyobjects, which are believed to be unreddened by dust.

When we construct a Hubble diagram in e.g. V -band, we would like tocorrect the supernova peak magnitudes by host extinction AV , whereby

AV = RV × ((mB−mV )obs− (mB−mV )intr). (3.21)

We must assume values for (mB−mV )intr, while RV can also be assumed orbe determined if more colour information is known.

To understand intrinsic colours, one can consider a distribution of asample of what one believes to be unreddened supernovae. Supernovaewhich are believed to be unreddened still have a distribution in colours, which

40

Figure 3.4: Conceptual illustration of photons scattered by CS dust into theline-of-sight of an observer. (Illustrated by R.F.)

3.2.2 Low RV and CS dust

The RV of any given line-of-sight is generally not known. In the Milky Way,dust has been studied and mapped in detail and has an average value ofRV ≈ 3.1. However, the exact value of a given line-of-sight can range betweenRV = 2–5 [see for instance 72]. It is often assumed that this property is thesame for other galaxies, including SNe hosts. Nevertheless, studies overbroad range of wavelengths have shown that the RV can have very differentvalues for different SNe. Paper II studies the extinction laws of a small sampleof SNe in detail, where surprisingly steep extinctions laws are found, withRV values which would be very unusual to observe in the Milky Way. Thelower than average RV of many SNe Ia hints at the explosions occurring inunusual environments. It is unclear whether the unusual extinction laws areinfluenced or even caused by the supernova progenitor systems themselves.If this is not the case, the extinction law determined in our Milky Way, isnot representative for many SNe Ia host galaxies. Since age, morphologyand star formation rate of galaxies vary, it would not be surprising if theseaffect the dust properties as well.

It is believed that the reddening law correlates with the grain size ofdust. Generally speaking, small dust grains imply a lower RV . Thus the dustgrain size in some SNe Ia host galaxies might be smaller than the usual grainsize in the Milky Way. For example, it has been proposed that the intenseUV light from star forming regions could break up dust grains, resulting inthe observed extinction laws. To date, it has however not been possible tocorrelate SNe Ia with low RV to star formation regions.

Another possibility to obtain a low RV is by scattering light multiple times[6; 7]. Normally, dust scatters light out of the line-of-sight, thereby causingextinction. If however the geometry of the dust cloud allows for light to be

42

scattered back into the line-of-sight, it can still be detected. An exampleof such a geometry is a CS dust shell, illustrated in Figure 3.4. A solutionto the low RV problem could be the presence of CS dust shells surroundingSNe Ia. Thus the origin to the low RV is connected to the properties of theCS environment of SNe Ia and which progenitor models predict the formationof CS dust.

43

44

4. Circumstellar Environments ofSNe Ia

We have seen that there are two motivations to investigate the propertiesof the CS environment of SNe Ia:

• The properties could help distinguish between SNe Ia progenitor mod-els, as described Chapter 2.

• CS dust could explain the peculiar extinction relation seen in manySNe Ia and outlined in Chapter 3.

As a reminder, some progenitor models predict that CS matter should bedeposited at distances where dust would lead to the steep extinction rela-tions observed. CS dust needs to be within a few 10−1 parsec [6; 7], toresult in a low RV . While models of C/O+He WD binaries suggest thatmatter shells within that radius [24], SD models point to slightly larger cav-ities with matter deposited ∼ 10 parsec from the progenitor [23]. Thereare several observational approaches to characterising different properties ofthe CS environment. The methods are described below and summarised inTable 4.2.

4.1 Narrow absorption lines

To date, evidence for the presence of CS matter round SNe Ia has mainlyappeared in studying narrow absorption line features in their spectra. Ab-sorption lines appear due to gas anywhere along the line-of-sight to theexplosions. Statistical analyses of Na I D absorption lines in SNe Ia spectrahave shown that the profiles are predominantly blue-shifted, indicating thatthere is outflowing gas [73; 74]. Interestingly, this could not be observed ina comparable sample of core-collapse supernovae, suggesting that this is afeature unique to SNe Ia. It is thus possible that some SNe Ia are surroundedby CS gas. Since this method only provides indirect evidence for outflowingmatter in some SNe Ia, more direct measurements of CS gases are desirable.

In fact, the detection of CS gas surrounding SN 2006X had previouslybeen claimed, based on the profile variations of Na I D absorption lines

45

Figure 4.1: Geometric effects of an expanding photosphere. A small gas cloudcan initially absorb a larger fraction of the photons emitted towards an observerto the right of the diagram. As the photosphere expands, more photonspass around the gas cloud and the fractional amount of photons reaching theobserver increases. Depending on the geometry of the gas clouds along theline-of-sight, this effect can lead to increasing or decreasing absorption lineprofiles in supernova spectra. (Illustrated by R.F.)

[14]. In the weeks following maximum brightness of SN 2006X, severalblue-shifted components of the Na I D profile appeared and increased overtime. This was interpreted as the recombination of gas to Na I in a denseCS gas shell, after the gas had been ionised by the early UV flux of thesupernova. Intriguingly, the absorption line profile, and the changes to it,resembled those detected in an outburst of the recurrent nova RS Ophiuchi,mentioned as a SN Ia progenitor candidate in Chapter 2 [75]. However,the time-scale of the variations detected in SN 2006X have been difficult toexplain by recombination and it has been suggested that geometric effectsof ISM clouds are responsible for the variations [20].

The main cause of geometric effects are gas clouds which have a smallersize than the projected surface of the the supernova photosphere [19]. Be-cause the photosphere is rapidly expanding behind the gas clouds, the frac-tion of absorbed light varies with time, which is observed as an absorptionline variation. Figure 4.1 illustrates how the fractional amount of absorptioncan change due to an expansing photosphere.

While another case of an increasing Na I D component was discoveredin SN 2007le [15], a larger sample of 14 time-series of high-resolution spec-tra failed to uncover any more examples of variable absorption lines [17].Although the well covered SN 2011fe showed slight variations, those wereattributed to geometric effects as well [16].

Thus, at the outset of the research presented in this thesis, the evidencefor matter in the CS environment of SNe Ia was unclear. The cases ofincreasing absorption lines do not have a single simple explanation and couldnot be easily used to determine the location of the gas with respect to thesupernova.

46

SN earliest epoch RinnerNa I Rinner

Ca II references(days) (parsec) (parsec)

SN 2011fe −18 0.003 – [16; 76]SN 2017cbv −15 0.02 3×10−4 [Paper V, 22]SN 2014J −11 0.08 0.003 [13]iPTF13dge −10 0.1 0.004 [Paper III, 18]SN 2013gh −9 0.2 0.006 [Paper III, 18]SN 2010A −8 0.2 0.01 [17]SN 2012cg −7 0.3 0.01 [Paper II, 8]iPTF16abc −6 0.3 0.02 [Paper IV, 21]SN 2007le −5 0.4 0.02 [15]SN 2012fr −5 0.4 0.02 [17]SN 2006X −2 0.5 0.04 [14]SN 2012ht −2 0.5 0.04 [17]SN 2013aa −1 0.8 0.05 [17]

Table 4.1: List of SNe Ia with multi-epoch high-resolution spectroscopy start-ing before maximum brightness. The earliest epoch is given with respect toBmax and the computed inner radius limit of photoionisation of Na I and Ca IIgas are presented.

Assuming the simple photoionisation model, one can determine distancesfrom the supernova at which the presence of gas can be excluded. Thereby,the inner radius of the exclusion range is determined by the epoch at whichthe first spectrum is taken. If the gas at a given radius is ionised before thetime of the first epoch, it will be undetectable. Thus the observations aresensitive to gas at distance greater than that radius. The method used todetermine the exclusion radii is discussed in Papers II and III. In Figure 4.3,it can be seen that early spectra, two weeks before maximum, are necessaryto detect photoionisation of Na I D down to ∼ 0.1 parsec. In Table 4.1SNe Ia for which multi-epoch high-resolution spectra starting before Bmax

have been published, are listed along with the phase of the earliest spectrumand corresponding inner sensitivity limits from photoionisation. It can beseen that the best possible exclusion limits have been reached with SN 2011feand SN 2017cbv, in Paper V.

4.3 Other observational evidence

Many observational methods aim to detect outflowing matter, which willhave a characteristic Doppler shift with respect to the rest frame of the

49

supernova explosion. For example, a giant star and accreting WD in theprogenitor system should sustain a powerful stellar wind, which would bedetectable by radio and X-ray emission. In the vicinity of a SN Ia (a few10−2–10−1 parsec), electrons are intensely heated by radiation and emit X-ray and and synchrotron radiation. The signals can be distinguished fromthose of the supernovae ejecta because the winds should be Doppler shiftedat ∼ 103 km s−1 as opposed to the ejecta which expands at ∼ 105 km s−1.These methods have set strong limits on the density of outflowing gas in theCS environments of SNe Ia in radio [77; 78] and X-ray [79] from SNe 2011feand 2014J. These values are consistent with a normal warm ISM and suggestthat there were no strong stellar winds coming from the progenitor systems

A direct measurement of the density of dust content in SN Ia CS envi-ronments can be made by observing the thermal emission of dust heated bythe explosion after it has faded. Results from mid- and far-infrared suggestthat there can only be small amounts of dust surrounding normal SNe Ia[80; 81]. In particular for SN 2014J this method has shown that there isnegligible dust out to a radius of ∼ 10 parsec.

This result strongly suggests that the extinction seen in SNe Ia suchas SN 2014J comes from dust in the ISM. The conclusion is supported bypolarisation data, which shows that the dust gains causing the extinctionare aligned with the magnetic fields of the host galaxies [11]. A newlydeveloped method which tracks the colour evolution of SNe Ia can determinethe distance dust clouds along the line-of-sight to the explosion [Paper F,82]. Similar to CS dust, dust that is located in the ISM, but still closeenough to the supernova, can scatter photons back into the line-of-sightof the observer. Because of the time delay of the scattered photons, theextinction has a time dependence, which is determined by the separationbetween the dust and the explosion. In the case of SN 2014J this methodconfirms that the dust must have been at interstellar distances.

50

Obs

erva

tiona

lMet

hod

Tim

ing

CS

rang

ese

nsiti

vity

Phy

sica

ldes

crip

tion

(par

sec)

X-r

aym

axim

umbr

ight

ness

∼10−2

–10−

1Em

ission

ofhe

ated

elec

tron

s[7

9]Rad

iom

axim

umbr

ight

ness

∼10−2

–10−

1Sy

nchr

otro

nra

diat

ion

ofhe

ated

elec

tron

s[7

7;78

]m

id-

&fa

r-IR

afte

rSN

has

fade

d<

10th

erm

alem

ission

ofdu

st[8

0;81

]

colo

urex

cess

time-

evol

utio

nco

vera

geof

the

entir

elig

htcu

rve

∼10−3

–101

char

acte

ristic

time

dela

yof

phot

ons

scat

-te

red

back

into

the

line-

of-s

ight

bydu

st[P

a-pe

rF,

82]

abso

rptio

nlin

epr

ofile

san

ytim

ear

ound

max

i-m

umbr

ight

ness

entir

elin

e-of

-sig

htga

sst

ream

ing

outre

sults

inbl

ue-s

hift

edpr

o-fil

es[7

3;74

]ab

sorp

tion

line

varia

tions

durin

gth

eris

etim

e∼

10−2

–101

phot

oion

isat

ion

ofga

s[5

]

Tab

le4.

2:O

bser

vatio

nalap

proa

ches

tode

tect

CS

mat

ter

surr

ound

ing

SNe

Ia.

The

timin

gre

fers

toth

eap

prox

imat

eph

ase

ofth

eex

plos

ion,

whe

nth

eob

serv

atio

nsm

ust

bem

ade

and

the

sens

itivi

tyra

nge

indi

cate

sat

wha

tdi

stan

ces

from

the

supe

rnov

aeth

em

etho

dca

nde

tect

mat

ter.

The

met

hods

are

desc

ribed

inde

tail

inth

ete

xt.

51

52

5. Conclusions

The aim of the research presented in this thesis was to search for CS mattersurrounding SNe Ia. The search had two main motivations. On the one hand,the presence of CS dust is a possible explanation for the steep extinctioncurves observed in some SNe Ia, on the other, the presence and propertiesof CS matter should provide valuable clues to the illusive progenitor systemsof the explosions.

In the past, the detection of CS gas surrounding a SN Ia had beenclaimed by detecting variations in narrow absorption lines in multi-epochhigh-resolution spectra. Guided by past observations, we obtained time-series of spectra spanning the maximum brightness of several SNe Ia, in-cluding SNe 2014J in Papers I and II and 2013gh in Paper III. In the lattercase, we discovered a decreasing Na I D absorption feature, which could beexplained by photoionisation.

Taking advantage of the well-known spectral energy distribution of SNe Ia,we can predict the photoionisation rate of gas in the CS medium of the ex-plosions. Assuming that the absorption line in SN 2013gh varied due tophotoionisation, the model suggest that the gas was at a distance fromthe explosion where it could be an independent gas cloud in the ISM. InPaper III, we were also able to show that most of the time-series of high-resolution spectra to date start at epochs at which most of the CS gaswould have already been ionised. Thus the nondetections of absorption linevariations until this point do not conclusively exclude the presence of CSgas.

To be sensitive to photoionisation across most of the CS medium, weaimed to obtain high-resolution spectra starting two weeks before maximumbrightness of the explosions. To my knowledge, such early spectra had onlybeen obtained once before in the case of SN 2011fe, where only slight ab-sorption line variations were detected, which were consistent with geometriceffects. We achieved our goal by observing the nearby explosion SN 2017cbvmore than two weeks before maximum brightness, presented in Paper V. Thetime-span of our observations of SN 2017cbv allowed us to conclude thatvery little Na I and Ca II gas could be present in the CS environment ofthe explosion. This adds to the list of observational evidence, that only

53

undetectable amounts of CS matter surround normal SNe Ia.In conclusion, this allows us to address the initial motivations of the

research in this thesis, the illusive progenitor question introduced in Chapter 2and the steep extinction laws as discussed in Chapter 3.

In particular, the absence of thermal emission of heavily reddened SNe Ia,such as SN 2014J [80; 81], shows that the bulk of the dust must be locatedin the ISM of the host galaxies as opposed to the CS media of the explosions.Furthermore, polarisation of reddened SNe Ia indicate that dust grains alongthe line-of-sight align with the magnetic field of the host galaxies [11]. This isin agreement with the observations from high-resolution spectra, where deepnon-varying absorption lines are detected, suggesting that the gas is far fromthe supernovae where it could not be ionised. Thus the evidence suggest thatthe steep extinction laws do not arise from CS dust, but from dust in the ISMof the host galaxies. Interstellar dust can give rise to steep extinction lawsas well, however, this is very unusual in the Milky Way. Future supernovacosmology surveys need to reduce systematic uncertainties in using SNe Ia asstandard candles to improve the measurements of cosmological parameters.Thus further research will have to show how and why the dust propertiesin some SN Ia lines-of-sight differ from the properties of dust in the MilkyWay.

The absence of a dense CS environment also provides clues to answeringthe illusive progenitor question. The evidence favours progenitor models,which do not have prolonged phases with strong stellar winds, which wouldcreate shells of CS matter. This speaks against SD and C/O+He DD sce-narios, for which models suggest detectable absorption features should bepresent. This is in agreements with other observations, such as the absenceof potential companion stars in pre- and post-explosion images of normalSNe Ia. In future, multi-epoch high-resolution spectra could be used toidentify CS gas around peculiar SNe Ia, which could help associating themwith specific progenitor systems.

54

Acknowledgements

It has always been a dream of mine to research and there are many peoplewho have helped me along the way to completing this thesis. It is difficult tofind the right words here, but I hope to personally thank most of the peoplewho have supported me at some time along the way. Nevertheless, I wouldlike to thank a few people here who have had a significant role during mytime in Stockholm.

Firstly, Rahman and Ariel, thank you for the opportunity to work withyou in the Snova group. It is too early to grasp how much I have learnt overthe past years and I am sure a lot of it will be useful in the future, whateverI end up doing!

I am also grateful to have made many good friends during my time inStockholm. Semeli, Tanja and Laura, I have many great memories fromour time sharing an office together! Semeli, Tanja, Brandon, Manuel, Svea,Timur, Angnis, Knut, Suzanna, Michael and Francesca, you have becomedear friends to me. We have had many good laughs together and yousupported me when I felt down. It is difficult to describe how thankful I amfor your friendship! I hope we manage to see each other in future and youare always welcome to visit and stay with me, where ever I might be.

I also would like to thank my old friends, who have remotely always beenthere for me. Camilla, Francis, Tono and many others, our skype calls havehelped me survive the dark Swedish winters!

Lastly, I would like to thank my family for always being loving and sup-portive. Mama and Papa, needless to say that I wouldn’t have achieved anyof this without you and Micki, glad to have convinced you to also start aPhD in physics.

References

[1] A. G. Riess, A. V. Filippenko, P. Challis, A. Clocchiatti, A. Diercks, P. M.

Garnavich, R. L. Gilliland, C. J. Hogan, S. Jha, R. P. Kirshner, B. Lei-

bundgut, M. M. Phillips, D. Reiss, B. P. Schmidt, R. A. Schommer, R. C.

Smith, J. Spyromilio, C. Stubbs, N. B. Suntzeff, and J. Tonry. Observational

Evidence from Supernovae for an Accelerating Universe and a Cosmological Constant.AJ, 116:1009–1038, September 1998. 21

[2] S. Perlmutter, G. Aldering, G. Goldhaber, R. A. Knop, P. Nugent, P. G.

Castro, S. Deustua, S. Fabbro, A. Goobar, D. E. Groom, I. M. Hook,

A. G. Kim, M. Y. Kim, J. C. Lee, N. J. Nunes, R. Pain, C. R. Pennypacker,

R. Quimby, C. Lidman, R. S. Ellis, M. Irwin, R. G. McMahon, P. Ruiz-

Lapuente, N. Walton, B. Schaefer, B. J. Boyle, A. V. Filippenko, T. Math-

eson, A. S. Fruchter, N. Panagia, H. J. M. Newberg, W. J. Couch, and

Supernova Cosmology Project. Measurements of Omega and Lambda from 42

High-Redshift Supernovae. ApJ, 517:565–586, June 1999. 21

[3] N. Suzuki, D. Rubin, C. Lidman, G. Aldering, R. Amanullah, K. Barbary,

L. F. Barrientos, J. Botyanszki, M. Brodwin, N. Connolly, K. S. Dawson,

A. Dey, M. Doi, M. Donahue, S. Deustua, P. Eisenhardt, E. Ellingson,

L. Faccioli, V. Fadeyev, H. K. Fakhouri, A. S. Fruchter, D. G. Gilbank,

M. D. Gladders, G. Goldhaber, A. H. Gonzalez, A. Goobar, A. Gude,

T. Hattori, H. Hoekstra, E. Hsiao, X. Huang, Y. Ihara, M. J. Jee, D. John-

ston, N. Kashikawa, B. Koester, K. Konishi, M. Kowalski, E. V. Linder,

L. Lubin, J. Melbourne, J. Meyers, T. Morokuma, F. Munshi, C. Mullis,

T. Oda, N. Panagia, S. Perlmutter, M. Postman, T. Pritchard, J. Rhodes,

P. Ripoche, P. Rosati, D. J. Schlegel, A. Spadafora, S. A. Stanford, V. Stan-

ishev, D. Stern, M. Strovink, N. Takanashi, K. Tokita, M. Wagner, L. Wang,

N. Yasuda, H. K. C. Yee, and T. Supernova Cosmology Project. The Hubble

Space Telescope Cluster Supernova Survey. V. Improving the Dark-energy Constraints

above z > 1 and Building an Early-type-hosted Supernova Sample. ApJ, 746:85, Febru-ary 2012. 21

[4] M. Betoule, R. Kessler, J. Guy, J. Mosher, D. Hardin, R. Biswas, P. Astier,

P. El-Hage, M. Konig, S. Kuhlmann, J. Marriner, R. Pain, N. Regnault,

C. Balland, B. A. Bassett, P. J. Brown, H. Campbell, R. G. Carlberg,

F. Cellier-Holzem, D. Cinabro, A. Conley, C. B. D’Andrea, D. L. DePoy,

M. Doi, R. S. Ellis, S. Fabbro, A. V. Filippenko, R. J. Foley, J. A. Frieman,

D. Fouchez, L. Galbany, A. Goobar, R. R. Gupta, G. J. Hill, R. Hlozek, C. J.

Hogan, I. M. Hook, D. A. Howell, S. W. Jha, L. Le Guillou, G. Leloudas,

C. Lidman, J. L. Marshall, A. Möller, A. M. Mourão, J. Neveu, R. Nichol,

M. D. Olmstead, N. Palanque-Delabrouille, S. Perlmutter, J. L. Prieto,

C. J. Pritchet, M. Richmond, A. G. Riess, V. Ruhlmann-Kleider, M. Sako,

K. Schahmaneche, D. P. Schneider, M. Smith, J. Sollerman, M. Sullivan,

N. A. Walton, and C. J. Wheeler. Improved cosmological constraints from a joint

analysis of the SDSS-II and SNLS supernova samples. A&A, 568:A22, August 2014. 21

[5] K. J. Borkowski, J. M. Blondin, and S. P. Reynolds. Circumstellar Shells in

Absorption in Type Ia Supernovae. ApJL, 699:L64–L67, July 2009. 22, 47, 51

[6] L. Wang. Dust around Type Ia Supernovae. ApJL, 635:L33–L36, December 2005. 22,42, 45

[7] A. Goobar. Low RV from Circumstellar Dust around Supernovae. ApJL, 686:L103–L106, October 2008. 22, 42, 45

[8] R. Amanullah, J. Johansson, A. Goobar, R. Ferretti, S. Papadogiannakis,

T. Petrushevska, P. J. Brown, Y. Cao, C. Contreras, H. Dahle, N. Elias-

Rosa, J. P. U. Fynbo, J. Gorosabel, L. Guaita, L. Hangard, D. A. Howell,

E. Y. Hsiao, E. Kankare, M. Kasliwal, G. Leloudas, P. Lundqvist, S. Mat-

tila, P. Nugent, M. M. Phillips, A. Sandberg, V. Stanishev, M. Sullivan,

F. Taddia, G. Östlin, S. Asadi, R. Herrero-Illana, J. J. Jensen, K. Karhunen,

S. Lazarevic, E. Varenius, P. Santos, S. S. Sridhar, S. H. J. Wallström, and

J. Wiegert. Diversity in extinction laws of Type Ia supernovae measured between

0.2 and 2 μm. MNRAS, 453:3300–3328, November 2015. 22, 23, 48, 49

[9] A. Goobar, J. Johansson, R. Amanullah, Y. Cao, D. A. Perley, M. M.

Kasliwal, R. Ferretti, P. E. Nugent, C. Harris, A. Gal-Yam, E. O. Ofek,

S. P. Tendulkar, M. Dennefeld, S. Valenti, I. Arcavi, D. P. K. Banerjee,

V. Venkataraman, V. Joshi, N. M. Ashok, S. B. Cenko, R. F. Diaz, C. Frem-

ling, A. Horesh, D. A. Howell, S. R. Kulkarni, S. Papadogiannakis, T. Petru-

shevska, D. Sand, J. Sollerman, V. Stanishev, J. S. Bloom, J. Surace, T. J.

Dupuy, and M. C. Liu. The Rise of SN 2014J in the Nearby Galaxy M82. ApJL,784:L12, March 2014. 22, 30

[10] R. Amanullah, A. Goobar, J. Johansson, D. P. K. Banerjee, V. Venkatara-

man, V. Joshi, N. M. Ashok, Y. Cao, M. M. Kasliwal, S. R. Kulkarni, P. E.

Nugent, T. Petrushevska, and V. Stanishev. The Peculiar Extinction Law of SN

2014J Measured with the Hubble Space Telescope. ApJL, 788:L21, June 2014. 22

[11] F. Patat, S. Taubenberger, N. L. J. Cox, D. Baade, A. Clocchiatti,

P. Höflich, J. R. Maund, E. Reilly, J. Spyromilio, L. Wang, J. C. Wheeler,

and P. Zelaya. Properties of extragalactic dust inferred from linear polarimetry of

Type Ia Supernovae. A&A, 577:A53, May 2015. 23, 50, 54

[12] K. Maeda, A. Tajitsu, K. S. Kawabata, R. J. Foley, S. Honda, Y. Moritani,

M. Tanaka, O. Hashimoto, M. Ishigaki, J. D. Simon, M. M. Phillips, M. Ya-

manaka, D. Nogami, A. Arai, W. Aoki, K. Nomoto, D. Milisavljevic, P. A.

Mazzali, A. M. Soderberg, M. Schramm, B. Sato, H. Harakawa, N. Morrell,

and N. Arimoto. Sodium Absorption Systems toward SN Ia 2014J Originate on

Interstellar Scales. ApJ, 816:57, January 2016. 22, 23

[13] M. L. Graham, S. Valenti, B. J. Fulton, L. M. Weiss, K. J. Shen, P. L.

Kelly, W. Zheng, A. V. Filippenko, G. W. Marcy, D. A. Howell, J. Burt,

and E. J. Rivera. Time-Varying Potassium in High-Resolution Spectra of the Type

Ia Supernova 2014j. ApJ, 801:136, March 2015. 23, 47, 48, 49

[14] F. Patat, P. Chandra, R. Chevalier, S. Justham, P. Podsiadlowski, C. Wolf,

A. Gal-Yam, L. Pasquini, I. A. Crawford, P. A. Mazzali, A. W. A. Pauldrach,

K. Nomoto, S. Benetti, E. Cappellaro, N. Elias-Rosa, W. Hillebrandt, D. C.

Leonard, A. Pastorello, A. Renzini, F. Sabbadin, J. D. Simon, and M. Tu-

ratto. Detection of Circumstellar Material in a Normal Type Ia Supernova. Science,317:924–, August 2007. 46, 49

[15] J. D. Simon, A. Gal-Yam, O. Gnat, R. M. Quimby, M. Ganeshalingam, J. M.

Silverman, S. Blondin, W. Li, A. V. Filippenko, J. C. Wheeler, R. P. Kirsh-

ner, F. Patat, P. Nugent, R. J. Foley, S. S. Vogt, R. P. Butler, K. M. G.

Peek, E. Rosolowsky, G. J. Herczeg, D. N. Sauer, and P. A. Mazzali. Vari-

able Sodium Absorption in a Low-extinction Type Ia Supernova. ApJ, 702:1157–1170,September 2009. 46, 49

[16] F. Patat, M. A. Cordiner, N. L. J. Cox, R. I. Anderson, A. Harutyunyan,

R. Kotak, L. Palaversa, V. Stanishev, L. Tomasella, S. Benetti, A. Goobar,

A. Pastorello, and J. Sollerman. Multi-epoch high-resolution spectroscopy of SN

2011fe. Linking the progenitor to its environment. A&A, 549:A62, January 2013. 23,46, 49

[17] A. Sternberg, A. Gal-Yam, J. D. Simon, F. Patat, W. Hillebrandt, M. M.

Phillips, R. J. Foley, I. Thompson, N. Morrell, L. Chomiuk, A. M. Soder-

berg, D. Yong, A. L. Kraus, G. J. Herczeg, E. Y. Hsiao, S. Raskutti, J. G.

Cohen, P. A. Mazzali, and K. Nomoto. Multi-epoch high-spectral-resolution

observations of neutral sodium in 14 Type Ia supernovae. MNRAS, 443:1849–1860,September 2014. 46, 49

[18] R. Ferretti, R. Amanullah, A. Goobar, J. Johansson, P. M. Vreeswijk, R. P.

Butler, Y. Cao, S. B. Cenko, G. Doran, A. V. Filippenko, E. Freeland,

G. Hosseinzadeh, D. A. Howell, P. Lundqvist, S. Mattila, J. Nordin, P. E.

Nugent, T. Petrushevska, S. Valenti, S. Vogt, and P. Wozniak. Time-varying

sodium absorption in the Type Ia supernova 2013gh. A&A, 592:A40, July 2016. 23, 48,49

[19] F. Patat, N. L. J. Cox, J. Parrent, and D. Branch. Studying the small scale

ISM structure with supernovae. A&A, 514:A78, May 2010. 23, 46

[20] N. N. Chugai. Circumstellar Na I and Ca II absorption lines of type Ia supernovae in

the symbiotic scenario. Astronomy Letters, 34:389–396, June 2008. 23, 46

[21] R. Ferretti, R. Amanullah, A. Goobar, T. Petrushevska, S. Borthakur,

M. Bulla, O. Fox, E. Freeland, C. Fremling, L. Hangard, and M. Hayes.Probing gas and dust in the tidal tail of NGC 5221 with the type Ia supernova

iPTF16abc. ArXiv e-prints, August 2017. 23, 49

[22] R. Ferretti, R. Amanullah, M. Bulla, A. Goobar, J. Johansson, and

P. Lundqvist. No evidence of circumstellar gas surrounding Type Ia Supernova SN

2017cbv. ArXiv e-prints, August 2017. 23, 49

[23] C. Badenes, J. P. Hughes, E. Bravo, and N. Langer. Are the Models for Type Ia

Supernova Progenitors Consistent with the Properties of Supernova Remnants? ApJ,662:472–486, June 2007. 23, 30, 45

[24] K. J. Shen, J. Guillochon, and R. J. Foley. Circumstellar Absorption in Double

Detonation Type Ia Supernovae. ApJL, 770:L35, June 2013. 23, 30, 45

[25] W. Baade and F. Zwicky. Remarks on Super-Novae and Cosmic Rays. PhysicalReview, 46:76–77, July 1934. 25

[26] F. R. Stephenson and D. A. Green. Historical supernovae and their remnants. His-torical supernovae and their remnants, by F. Richard Stephenson and David A. Green. Inter-national series in astronomy and astrophysics, vol. 5. Oxford: Clarendon Press, 2002, ISBN0198507666, 5, 2002. 25

[27] C. Badenes, K. J. Borkowski, J. P. Hughes, U. Hwang, and E. Bravo. Con-

straints on the Physics of Type Ia Supernovae from the X-Ray Spectrum of the Tycho

Supernova Remnant. ApJ, 645:1373–1391, July 2006. 25

[28] J. Kepler. Ioannis Kepleri ... De Iesu Christi servatoris nostri vero anno quatuor an-nos in usitatam epocham desiderantis : (titulus currens: natalitio, consideratio novissimaesententiae Laurentii Suslygae ..., Sylva chronologica). 1606. 25

[29] S. P. Reynolds, K. J. Borkowski, U. Hwang, J. P. Hughes, C. Badenes, J. M.

Laming, and J. M. Blondin. A Deep Chandra Observation of Kepler’s Supernova

Remnant: A Type Ia Event with Circumstellar Interaction. ApJL, 668:L135–L138,October 2007. 25

[30] R. Minkowski. Spectra of Supernovae. PASP, 53:224, August 1941. 25

[31] J. C. Wheeler and R. Levreault. The peculiar Type I supernova in NGC 991.ApJL, 294:L17–L20, July 1985. 26

[32] J. H. Elias, K. Matthews, G. Neugebauer, and S. E. Persson. Type I supernovae

in the infrared and their use as distance indicators. ApJ, 296:379–389, September 1985.26, 37

[33] A. Heger, C. L. Fryer, S. E. Woosley, N. Langer, and D. H. Hartmann. How

Massive Single Stars End Their Life. ApJ, 591:288–300, July 2003. 26

[34] J. S. Bloom, D. Kasen, K. J. Shen, P. E. Nugent, N. R. Butler, M. L.

Graham, D. A. Howell, U. Kolb, S. Holmes, C. A. Haswell, V. Burwitz,

J. Rodriguez, and M. Sullivan. A Compact Degenerate Primary-star Progenitor of

SN 2011fe. ApJL, 744:L17, January 2012. 26

[35] E. Churazov, R. Sunyaev, J. Isern, J. Knödlseder, P. Jean, F. Lebrun,

N. Chugai, S. Grebenev, E. Bravo, S. Sazonov, and M. Renaud. Cobalt-56

γ-ray emission lines from the type Ia supernova 2014J. Nature, 512:406–408, August2014. 27

[36] E. Churazov, R. Sunyaev, J. Isern, I. Bikmaev, E. Bravo, N. Chugai,

S. Grebenev, P. Jean, J. Knödlseder, F. Lebrun, and E. Kuulkers. Gamma-

rays from Type Ia Supernova SN2014J. ApJ, 812:62, October 2015.

[37] J. Isern, P. Jean, E. Bravo, J. Knödlseder, F. Lebrun, E. Churazov, R. Sun-

yaev, A. Domingo, C. Badenes, D. H. Hartmann, P. Hoeflich, M. Renaud,

S. Soldi, N. Elias-Rosa, M. Hernanz, I. Domínguez, D. García-Senz, G. G.

Lichti, G. Vedrenne, and P. Von Ballmoos. Gamma-ray emission from SN2014J

near maximum optical light. A&A, 588:A67, April 2016. 27

[38] W. D. Arnett. Type I supernovae. I - Analytic solutions for the early part of the

light curve. ApJ, 253:785–797, February 1982. 27

[39] S. Chandrasekhar. The Maximum Mass of Ideal White Dwarfs. ApJ, 74:81, July1931. 27

[40] K. Nomoto and Y. Kondo. Conditions for accretion-induced collapse of white dwarfs.ApJL, 367:L19–L22, January 1991. 27

[41] J. Whelan and I. Iben, Jr. Binaries and Supernovae of Type I. ApJ, 186:1007–1014,December 1973. 28

[42] I. Iben, Jr. and A. V. Tutukov. Supernovae of type I as end products of the

evolution of binaries with components of moderate initial mass (M not greater than

about 9 solar masses). ApJS, 54:335–372, February 1984. 28

[43] M. Hernanz and J. José. The recurrent nova RS Oph: A possible scenario for type

Ia supernovae. New Astronomy Reviews, 52:386–389, October 2008. 29

[44] Z. Han and P. Podsiadlowski. The single-degenerate channel for the progenitors

of Type Ia supernovae. MNRAS, 350:1301–1309, June 2004. 29

[45] M. Hamuy, M. M. Phillips, N. B. Suntzeff, J. Maza, L. E. González, M. Roth,

K. Krisciunas, N. Morrell, E. M. Green, S. E. Persson, and P. J. McCarthy.An asymptotic-giant-branch star in the progenitor system of a type Ia supernova.Nature, 424:651–654, August 2003. 29

[46] B. Dilday, D. A. Howell, S. B. Cenko, J. M. Silverman, P. E. Nugent, M. Sul-

livan, S. Ben-Ami, L. Bildsten, M. Bolte, M. Endl, A. V. Filippenko, O. Gnat,

A. Horesh, E. Hsiao, M. M. Kasliwal, D. Kirkman, K. Maguire, G. W. Marcy,

K. Moore, Y. Pan, J. T. Parrent, P. Podsiadlowski, R. M. Quimby, A. Stern-

berg, N. Suzuki, D. R. Tytler, D. Xu, J. S. Bloom, A. Gal-Yam, I. M. Hook,

S. R. Kulkarni, N. M. Law, E. O. Ofek, D. Polishook, and D. Poznanski.PTF 11kx: A Type Ia Supernova with a Symbiotic Nova Progenitor. Science, 337:942,August 2012. 29

[47] R. F. Webbink. Double white dwarfs as progenitors of R Coronae Borealis stars and

Type I supernovae. ApJ, 277:355–360, February 1984. 29

[48] K. J. Shen, L. Bildsten, D. Kasen, and E. Quataert. The Long-term Evolution

of Double White Dwarf Mergers. ApJ, 748:35, March 2012. 29

[49] S. Dong, B. Katz, D. Kushnir, and J. L. Prieto. Type Ia supernovae with bimodal

explosions are common - possible smoking gun for direct collisions of white dwarfs.MNRAS, 454:L61–L65, November 2015. 29

[50] D. A. Howell, M. Sullivan, P. E. Nugent, R. S. Ellis, A. J. Conley, D. Le

Borgne, R. G. Carlberg, J. Guy, D. Balam, S. Basa, D. Fouchez, I. M. Hook,

E. Y. Hsiao, J. D. Neill, R. Pain, K. M. Perrett, and C. J. Pritchet. The

type Ia supernova SNLS-03D3bb from a super-Chandrasekhar-mass white dwarf star.Nature, 443:308–311, September 2006. 29

[51] Y. Cao, S. R. Kulkarni, D. A. Howell, A. Gal-Yam, M. M. Kasliwal,

S. Valenti, J. Johansson, R. Amanullah, A. Goobar, J. Sollerman, F. Taddia,

A. Horesh, I. Sagiv, S. B. Cenko, P. E. Nugent, I. Arcavi, J. Surace, P. R.

Woźniak, D. I. Moody, U. D. Rebbapragada, B. D. Bue, and N. Gehrels. A

strong ultraviolet pulse from a newborn type Ia supernova. Nature, 521:328–331, May2015. 30

[52] G. Hosseinzadeh, D. J. Sand, S. Valenti, P. Brown, D. A. Howell, C. Mc-

Cully, D. Kasen, I. Arcavi, K. Azalee Bostroem, L. Tartaglia, E. Y. Hsiao,

S. Davis, M. Shahbandeh, and M. D. Stritzinger. Early Blue Excess from the

Type Ia Supernova 2017cbv and Implications for its Progenitor. ArXiv e-prints, June2017. 30

[53] D. Kasen. Seeing the Collision of a Supernova with Its Companion Star. ApJ,708:1025–1031, January 2010. 30

[54] A. L. Piro and V. S. Morozova. Exploring the Potential Diversity of Early Type Ia

Supernova Light Curves. ApJ, 826:96, July 2016. 30

[55] W. Li, J. S. Bloom, P. Podsiadlowski, A. A. Miller, S. B. Cenko, S. W.

Jha, M. Sullivan, D. A. Howell, P. E. Nugent, N. R. Butler, E. O. Ofek,

M. M. Kasliwal, J. W. Richards, A. Stockton, H.-Y. Shih, L. Bildsten, M. M.

Shara, J. Bibby, A. V. Filippenko, M. Ganeshalingam, J. M. Silverman, S. R.

Kulkarni, N. M. Law, D. Poznanski, R. M. Quimby, C. McCully, B. Patel,

K. Maguire, and K. J. Shen. Exclusion of a luminous red giant as a companion star

to the progenitor of supernova SN 2011fe. Nature, 480:348–350, December 2011. 30

[56] P. L. Kelly, O. D. Fox, A. V. Filippenko, S. B. Cenko, L. Prato, G. Schaefer,

K. J. Shen, W. Zheng, M. L. Graham, and B. E. Tucker. Constraints on the

Progenitor System of the Type Ia Supernova 2014J from Pre-explosion Hubble Space

Telescope Imaging. ApJ, 790:3, July 2014. 30

[57] W. E. Kerzendorf, M. Childress, J. Scharwächter, T. Do, and B. P. Schmidt.A Reconnaissance of the Possible Donor Stars to the Kepler Supernova. ApJ, 782:27,February 2014. 30

[58] O. Graur, D. Poznanski, D. Maoz, N. Yasuda, T. Totani, M. Fukugita, A. V.

Filippenko, R. J. Foley, J. M. Silverman, A. Gal-Yam, A. Horesh, and B. T.

Jannuzi. Supernovae in the Subaru Deep Field: the rate and delay-time distribution

of Type Ia supernovae out to redshift 2. MNRAS, 417:916–940, October 2011. 30

[59] C. Raskin and D. Kasen. Tidal Tail Ejection as a Signature of Type Ia Supernovae

from White Dwarf Mergers. ApJ, 772:1, July 2013. 31

[60] A. Einstein. Die Grundlage der allgemeinen Relativitätstheorie. Annalen der Physik,354:769–822, 1916. 33

[61] A. Einstein. Kosmologische Betrachtungen zur allgemeinen Relativitätstheorie.Sitzungsberichte der Königlich Preußischen Akademie der Wissenschaften (Berlin), Seite142-152., 1917. 34

[62] G. Lemaître. Un Univers homogène de masse constante et de rayon croissant rendant

compte de la vitesse radiale des nébuleuses extra-galactiques. Annales de la SociétéScientifique de Bruxelles, 47:49–59, 1927. 34, 35

[63] E. Hubble. A Relation between Distance and Radial Velocity among Extra-Galactic

Nebulae. Proceedings of the National Academy of Science, 15:168–173, March 1929. 34

[64] A. Friedmann. Über die Krümmung des Raumes. Zeitschrift fur Physik, 10:377–386,1922. 35

[65] Planck Collaboration, P. A. R. Ade, N. Aghanim, M. Arnaud, M. Ashdown,

J. Aumont, C. Baccigalupi, A. J. Banday, R. B. Barreiro, J. G. Bartlett, and

et al. Planck 2015 results. XIII. Cosmological parameters. A&A, 594:A13, September2016. 36

[66] J. Martin. Everything you always wanted to know about the cosmological constant

problem (but were afraid to ask). Comptes Rendus Physique, 13:566–665, July 2012. 37

[67] E. V. Linder. Exploring the Expansion History of the Universe. Physical Review Letters,90(9):091301, March 2003. 37

[68] Planck Collaboration, P. A. R. Ade, N. Aghanim, M. Arnaud, M. Ashdown,

J. Aumont, C. Baccigalupi, A. J. Banday, R. B. Barreiro, N. Bartolo, and

et al. Planck 2015 results. XIV. Dark energy and modified gravity. A&A, 594:A14,September 2016. 37

[69] M. M. Phillips. The absolute magnitudes of Type IA supernovae. ApJL, 413:L105–L108, August 1993. 37

[70] J. Guy, P. Astier, S. Baumont, D. Hardin, R. Pain, N. Regnault, S. Basa,

R. G. Carlberg, A. Conley, S. Fabbro, D. Fouchez, I. M. Hook, D. A.

Howell, K. Perrett, C. J. Pritchet, J. Rich, M. Sullivan, P. Antilogus,

E. Aubourg, G. Bazin, J. Bronder, M. Filiol, N. Palanque-Delabrouille,

P. Ripoche, and V. Ruhlmann-Kleider. SALT2: using distant supernovae to im-

prove the use of type Ia supernovae as distance indicators. A&A, 466:11–21, April 2007.38

[71] LSST Science Collaboration, P. A. Abell, J. Allison, S. F. Anderson, J. R.

Andrew, J. R. P. Angel, L. Armus, D. Arnett, S. J. Asztalos, T. S. Axelrod,

and et al. LSST Science Book, Version 2.0. ArXiv e-prints, December 2009. 38

[72] D. M. Nataf, A. Gould, P. Fouqué, O. A. Gonzalez, J. A. Johnson,

J. Skowron, A. Udalski, M. K. Szymański, M. Kubiak, G. Pietrzyński,

I. Soszyński, K. Ulaczyk, Ł. Wyrzykowski, and R. Poleski. Reddening and

Extinction toward the Galactic Bulge from OGLE-III: The Inner Milky Way’s RV ˜ 2.5

Extinction Curve. ApJ, 769:88, June 2013. 42

[73] A. Sternberg, A. Gal-Yam, J. D. Simon, D. C. Leonard, R. M. Quimby, M. M.

Phillips, N. Morrell, I. B. Thompson, I. Ivans, J. L. Marshall, A. V. Fil-

ippenko, G. W. Marcy, J. S. Bloom, F. Patat, R. J. Foley, D. Yong, B. E.

Penprase, D. J. Beeler, C. Allende Prieto, and G. S. Stringfellow. Cir-

cumstellar Material in Type Ia Supernovae via Sodium Absorption Features. Science,333:856–, August 2011. 45, 51

[74] K. Maguire, M. Sullivan, F. Patat, A. Gal-Yam, I. M. Hook, S. Dhawan,

D. A. Howell, P. Mazzali, P. E. Nugent, Y.-C. Pan, P. Podsiadlowski, J. D.

Simon, A. Sternberg, S. Valenti, C. Baltay, D. Bersier, N. Blagorodnova,

T.-W. Chen, N. Ellman, U. Feindt, F. Förster, M. Fraser, S. González-

Gaitán, M. L. Graham, C. Gutiérrez, S. Hachinger, E. Hadjiyska, C. In-

serra, C. Knapic, R. R. Laher, G. Leloudas, S. Margheim, R. McKinnon,

M. Molinaro, N. Morrell, E. O. Ofek, D. Rabinowitz, A. Rest, D. Sand,

R. Smareglia, S. J. Smartt, F. Taddia, E. S. Walker, N. A. Walton, and

D. R. Young. A statistical analysis of circumstellar material in Type Ia supernovae.MNRAS, 436:222–240, November 2013. 45, 51

[75] F. Patat, N. N. Chugai, P. Podsiadlowski, E. Mason, C. Melo, and

L. Pasquini. Connecting RS Ophiuchi to [some] type Ia supernovae. A&A, 530:A63,June 2011. 46

[76] P. E. Nugent, M. Sullivan, S. B. Cenko, R. C. Thomas, D. Kasen, D. A.

Howell, D. Bersier, J. S. Bloom, S. R. Kulkarni, M. T. Kandrashoff, A. V.

Filippenko, J. M. Silverman, G. W. Marcy, A. W. Howard, H. T. Isaacson,

K. Maguire, N. Suzuki, J. E. Tarlton, Y.-C. Pan, L. Bildsten, B. J. Fulton,

J. T. Parrent, D. Sand, P. Podsiadlowski, F. B. Bianco, B. Dilday, M. L.

Graham, J. Lyman, P. James, M. M. Kasliwal, N. M. Law, R. M. Quimby,

I. M. Hook, E. S. Walker, P. Mazzali, E. Pian, E. O. Ofek, A. Gal-Yam, and

D. Poznanski. Supernova SN 2011fe from an exploding carbon-oxygen white dwarf

star. Nature, 480:344–347, December 2011. 49

[77] L. Chomiuk, A. M. Soderberg, R. A. Chevalier, S. Bruzewski, R. J. Foley,

J. Parrent, J. Strader, C. Badenes, C. Fransson, A. Kamble, R. Margutti,

M. P. Rupen, and J. D. Simon. A Deep Search for Prompt Radio Emission from

Thermonuclear Supernovae with the Very Large Array. ApJ, 821:119, April 2016. 50,51

[78] E. Kundu, P. Lundqvist, M. A. Pérez-Torres, R. Herrero-Illana, and A. Al-

berdi. Constraining Magnetic Field Amplification in SN Shocks Using Radio Obser-

vations of SNe 2011fe and 2014J. ApJ, 842:17, June 2017. 50, 51

[79] R. Margutti, J. Parrent, A. Kamble, A. M. Soderberg, R. J. Foley, D. Mil-

isavljevic, M. R. Drout, and R. Kirshner. No X-Rays from the Very Nearby Type

Ia SN 2014J: Constraints on Its Environment. ApJ, 790:52, July 2014. 50, 51

[80] J. Johansson, R. Amanullah, and A. Goobar. Herschel limits on far-infrared

emission from circumstellar dust around three nearby Type Ia supernovae. MNRAS,431:L43–L47, April 2013. 50, 51, 54

[81] J. Johansson, A. Goobar, M. M. Kasliwal, G. Helou, F. Masci, S. Tinyanont,

J. Jencson, Y. Cao, O. D. Fox, M. Kromer, R. Amanullah, D. P. K. Banerjee,

V. Joshi, A. Jerkstrand, E. Kankare, and T. A. Prince. Spitzer observations

of SN 2014J and properties of mid-IR emission in Type Ia supernovae. MNRAS,466:3442–3449, April 2017. 50, 51, 54

[82] M. Bulla, A. Goobar, R. Amanullah, U. Feindt, and R. Ferretti. Estimating

dust distances to Type Ia supernovae from colour excess time-evolution. ArXiv e-prints,July 2017. 50, 51