The extraplanar type II supernova ASASSN 14jb in the galaxy ESO 467 G051

Texto completo

(1)Pontificia Universidad Católica de Chile Facultad de Fı́sica Instituto de Astrofı́sica. The extraplanar type II supernova ASASSN-14jb in the galaxy ESO 467-G051. Por Nicolás Eduardo Meza Retamal Tesis presentada a la Facultad de Fı́sica de la Pontificia Universidad Católica de Chile, para optar al grado académico de de Magı́ster en Astrofı́sica.. Supervisores. :. Dr. Alejandro Clocchiatti Dr. Jose Luis Prieto. Correctores :. Dr. Franz Bauer Dr. Joe Anderson. Octubre 30 - 2018 Santiago, Chile c 2018, Nicolás Meza.

(2) c 2018, Nicolás Meza Se autoriza la reproducción total o parcial, con fines académicos, por cualquier medio o procedimiento, incluyendo la cita bibliográfica del documento.. 1.

(3) Dedico esto primero a mi familia, quien estuvo conmigo en los momentos complejos de este proceso, especialmente a mi Lela Marı́a y Tata Manuel, con quienes convivı́ estos años. Agradezco a mi Madre y Padre, Padrino Mario y tı́a Marion, quienes me apoyaron dentro de sus dificultades también en los momentos difı́ciles. No puedo dejar fuera a mis amigos, fuera y dentro de la UC, quienes me han acompañado de una u otra forma. Su recuerdo quedará conmigo siempre. Finalmente, mis agradecimientos no pueden faltar a todo el equipo del Instituto de Astrofı́sica, desde Mariela hasta las tı́as del aseo, que me ayudaron en cualquiera de mis problemas y mañas..

(4) Acknowledgement The author acknowledges the Insitute of Astrophysics at PUC, the chilean Milenium Institute of Astrophysics (MAS) and the Astronomy department at University Diego Portales (UDP) for supporting this research. Specifically I appreciate the advisory and opportunity given by Alejandro Clocchiatti at PUC, Jose L. Prieto at UDP and Joe Anderson at ESO-Chile. Also I thank the many contributions for the understanding of several parts of this work to Luc Dessart, Enrique Perez-Montero, Ondrew Pejcha, Franz Bauer, Lluis Galbany, Melina Bersten, Lin Xiao, Antonia Bevan, among others. This research has made use of data from the Public ESO Spectroscopic Survey of Transient Objects [PESSTO; Smartt et al., 2015, ESO program ID 191.D-0935] and Spitzer IRAC data from 2015-09-13 (program ID 11053), 2016-08-17 (program ID 12099) and 2014-09-05 (program ID 10139). This research has made use of the NASA/IPAC Extragalactic Database (NED) which is operated by the Jet Propulsion Laboratory, California Institute of Technology, under contract with the National Aeronautics. We acknowledge the usage of the HyperLeda database (http://leda.univ-lyon1.fr).. 3.

(5) Contents 1 Introduction. 8. 2 Data. 13. 2.1. Photometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13. 2.2. Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15. 3 Analysis 3.1. 3.2. 3.3. 17. Evolution and Comparison with other Type II Supernovae . . . . . . . . 17 3.1.1. Photometric Evolution . . . . . . . . . . . . . . . . . . . . . . . . 18. 3.1.2. Spectroscopic Evolution in the Plateau Phase . . . . . . . . . . . 19. 3.1.3. Expansion Velocities . . . . . . . . . . . . . . . . . . . . . . . . . 21. 3.1.4. Nebular Spectrum. . . . . . . . . . . . . . . . . . . . . . . . . . . 23. Distance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 3.2.1. Photospheric Magnitude Method . . . . . . . . . . . . . . . . . . 24. 3.2.2. Standard Candle Method . . . . . . . . . . . . . . . . . . . . . . . 26. 3.2.3. Photometric Colour Method . . . . . . . . . . . . . . . . . . . . . 26. 3.2.4. Adopted Distance . . . . . . . . . . . . . . . . . . . . . . . . . . . 27. Physical parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 3.3.1. Temperature Evolution and the Progenitor Radius . . . . . . . . . 27. 3.3.2. Bolometric Light Curve and the. 4. 56. Ni Mass . . . . . . . . . . . . . 32.

(6) CONTENTS. 3.4. 3.3.3. Nebular Spectra and the Progenitor Mass . . . . . . . . . . . . . 33. 3.3.4. Progenitor Metallicity from the Photospheric Spectral Lines . . . 37. The Host Galaxy of ASASSN-14jb with MUSE . . . . . . . . . . . . . . . 37 3.4.1. Overall Properties of ESO 467-G051 . . . . . . . . . . . . . . . . 38. 3.4.2. Host H II Regions Properties . . . . . . . . . . . . . . . . . . . . 40. 3.4.3. Hα and Oxygen Abundance Maps . . . . . . . . . . . . . . . . . . 41. 3.4.4. D16 and the N/O ratio . . . . . . . . . . . . . . . . . . . . . . . . 42. 4 Discussion and Conclusions 4.1. 4.2. 43. Runaways Stars as type II SNe . . . . . . . . . . . . . . . . . . . . . . . 43 4.1.1. Earlier Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44. 4.1.2. Runaway Stars as Peculiar CCSNe? . . . . . . . . . . . . . . . . . 45. Disk Thickness and Extraplanar Star Formation . . . . . . . . . . . . . . 46 4.2.1. Vertical Distribution of SNe . . . . . . . . . . . . . . . . . . . . . 46. 4.2.2. In-situ Star Formation . . . . . . . . . . . . . . . . . . . . . . . . 47. 4.2.3. Abundance Offset . . . . . . . . . . . . . . . . . . . . . . . . . . . 48. 4.3. SN-Host connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49. 4.4. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50. Appendices. 54. A Photometry. 55. B P-Cygni profile fitting. 58. 5.

(7) Abstract We present optical photometry and spectroscopy of the Type II supernova ASASSN14jb, together with VLT MUSE IFU observations of its host galaxy and a nebular-phase spectrum. This supernova, in the nearby galaxy ESO 467-G051 (z = 0.006), was discovered and followed-up by the All Sky Automated Survey for SuperNovae (ASAS-SN). We obtained well-sampled LCOGTN BVgri and Swift w2m1w1ubv optical and near-UV light curves and several optical spectra in the early photospheric phases. ASASSN-14jb exploded ∼2 kpc above the star-forming disk of ESO 467-G051, an edge-on disk galaxy. The large projected distance from the disk and non-detection of any H II region in a 1.4 kpc projected radius, based on MUSE IFU observations of the explosion site, are in conflict with the standard environment of core-collapse supernova progenitors and suggest a possible scenario whereby the progenitor received a kick during a binary interaction. We present analysis of the optical light curves and spectra, from which we derive a distance of 25 ± 1 Mpc using state of the art empirical methods for Type II SNe, physical properties of the SN explosion (56 Ni mass, explosion energy, and ejected mass) and properties of the progenitor, namely the progenitor radius, mass and metallicity. Our analysis yields a. 56. Ni mass of 0.0210 ± 0.0025 M , an explosion energy of. ≈ 0.25 foe and an ejected mass of ≈ 6 M . We also constrain the progenitor radius to be R∗ = 579 ± 28 R which seems to be consistent with the sub-solar metallicity of 0.3±0.1 Z derived from the supernova Fe II λ5018 line. The nebular spectra constrains strongly the progenitor mass to be in the range 10-12 M . From Spitzer data archives 6.

(8) CONTENTS. we detect ASASSN-14jb ≈ 330 days past explosion and we derive a total dust mass of 10−4 M from the 3.6 µm and 4.5 µm photometry. Using the FUV, NUV, BV gri,Ks , 3.6 µm, and 4.5 µm total magnitudes for the host galaxy, we fit stellar population synthesis models, which give an estimate of M∗ ≈ 1 × 109 M , an age of 3.2 Gyr, and a SFR ≈ 0.07 M /yr. We also discuss the low oxygen abundance of the host galaxy derived from the MUSE data, having an average of 12 + log (O/H) = 8.27+0.16 −0.20 using the O3N2 diagnostic, with strong line methods and compare it with the supernova spectra, which is also consistent with a sub-solar metallicity progenitor. Following recent observations of extraplanar H II regions in nearby edge-on galaxies, we derive the metallicity offset from the disk to be positive, suggesting enrichment from disk outflows. We finally discuss the possible scenarios for the unusual environment for ASASSN-14jb and conclude that either the in-situ star formation or a runaway scenario would imply a low mass progenitor, agreeing with our estimate from the supernova nebular spectra. Regardless of the true origin of ASASSN-14jb we show in this work that the detailed study of the environment approximately agrees with the stronger constraint of the transient observations.. 7.

(9) Chapter 1 Introduction Originally classified according to the absence (Type I) or presence (Type II) of hydrogen lines in their optical spectra [Minkowski, 1941], supernovae (SNe) represent the explosive ending of a star. The explosion of massive stars , with Zero Age Main Sequence (ZAMS) masses MZAM S > 8 − 10 M , that arise from the collapse of the iron core are collectively called core-collapse supernovae (CCSNe) while the explosion of white dwarfs from lower mass stars, associated with Type Ia SNe, are called thermonuclear SNe [Filippenko, 1997, Turatto et al., 2007]. For detailed reviews on explosion mechanism see, e.g., Bethe [1990] for CCSNe and Hillebrandt and Niemeyer [2000] for thermonuclear explosions. Supernovae are a critical component of galaxy evolution. The injected energy they give to the ISM regulates star formation and the chemical enrichment of galaxies and haloes, being a fundamental condition to understand galaxy formation and evolution in modern cosmological scenarios [Larson, 1974, Ferrara and Tolstoy, 2000, Hopkins et al., 2014]. Supernovae are also a vital step in the extragalactic distance ladder. High redshift (z ≥ 0.1) observations of Type Ia SNe have made a revolution in our understanding of the energy content of the Universe, showing that the cosmic expansion rate is increasing in time by an unknown energy component dubbed “Dark energy” [Frieman et al., 2008]. Although dimmer than Type Ia SNe, Type II SNe have been 8.

(10) CHAPTER 1. INTRODUCTION. proven to be an interesting independent method to measure cosmological distances, being subject of recent improvements [de Jaeger et al., 2017a, Gall et al., 2018]. Historically [Barbon et al., 1979], Type II SNe were photometrically classified in distinct sub-classes based on their decline rate in the optical bands. The slow decliners were called “plateau” (Type II-P) and the fast decliners “linear” (Type II-L). Recent studies have shown that there is continuous distribution of the decline rate and other parameters [e.g. Anderson et al., 2014b, Sanders et al., 2015, Pejcha and Prieto, 2015a, Galbany et al., 2016b] and that type II-L have similar progenitors to II-P [Valenti et al., 2015]. The great diversity of CCSNe, and in particular of Type II SNe, is explained by the continuous distribution of the massive star progenitor properties, such as initial mass, radius, metallicity, rotation, among others, like the many orbital parameters involved in binary interactions [e.g. Heger et al., 2003, Kasen and Woosley, 2009, Dessart et al., 2013, Pejcha and Prieto, 2015b]. If the progenitor star retains a significant fraction of hydrogen in its envelope at the moment of explosion, a recombination wave sets in the cooling expanding plasma that delays the release of energy. Consequently, the light curve is characterized by a slowly declining phase (or plateau) after maximum [e.g. Popov, 1993, Bersten et al., 2011]. This plateau phase has been observed lo last ∼ 100 days [Anderson et al., 2014a, Galbany et al., 2016b, Valenti et al., 2016] and justifies the name of Type II-Plateau SNe. If the progenitor mass loss is significant, the lack of hydrogen in the envelope makes the SN ejecta photosphere recede faster and the light curve declines faster as a Type II-L [e.g. Blinnikov and Bartunov, 1993, Moriya et al., 2016]. Greater mass loss and the interaction with circumstellar material (CSM) changes the observational display and the SNe event nomenclature. If only a small fraction of hydrogen is maintained at the time of core-collapse, the event becomes a transitional Type IIb, where the hydrogen is observed only at early times [Filippenko, 1988, Filippenko et al., 1993, Nomoto et al., 1993, Chevalier and Soderberg, 2010]. Type IIn 9.

(11) CHAPTER 1. INTRODUCTION. supernovae are a subclass of Type II where the spectra is characterized by the presence of strong narrow emission lines that are interpreted as the result of interaction with a dense CSM [e.g. Dopita et al., 1984, Schlegel, 1990, Stathakis and Sadler, 1991, Chugai, 1994]. Narrow emission lines have also been observed in normal Type II SNe at early times, demonstrating the range of masses where the mass loss can play an important effect on the SNe emission [e.g. SN2007pk in Inserra et al., 2013, Khazov et al., 2016, Yaron et al., 2017]. Further mass loss can make a star lose its hydrogen/helium envelope entirely and the consequent absence of hydrogen/helium lines in the SNe spectra gives the nomenclature to the SN Ib/c spectral types. Type Ib/c together with Type IIb SNe are also called stripped enveloped supernovae (SESNe). Progenitors of Type II-P like SNe have been detected using pre-explosion images as red supergiants (RSG) with ZAMS masses between ∼ 8 − 17 M [Smartt et al., 2009, Smartt, 2015], while for SESNe they are scarce [e.g. Folatelli et al., 2015, Eldridge et al., 2013, Folatelli et al., 2016]. As CCSNe progenitors are massive stars with relatively short lifetimes of ∼ 4 − 50 Myr, it is expected that at the moment of explosion the SNe are still associated with their birth place (although binaries can provide progenitors with longer time scales of 100 − 200 Myr, e.g., Zapartas et al. 2017, Eldridge et al. 2017), namely spiral arms in late time galaxies or H II regions [Bartunov et al., 1994, McMillan and Ciardullo, 1996], although there is evidence of CCSNe found in early-type galaxies with residual star formation due to gravitational interactions [Hakobyan et al., 2008]. This correlation with star-forming regions is expected to decrease toward the lower-mass progenitors range due to the longer time the star has to move from its original birthplace. Recent observations have quantified the relation between different CCSNe events and their environment and found that stripped envelope SNe are more closely associated with regions of recent star formation than their hydrogen rich counterparts, as it is expected from the increasing progenitor ZAMS mass of the sequence Ia → II → IIb/IIn → Ib/c [e.g. Anderson et al., 2012, Galbany et al., 2014]. Observations of the host, global 10.

(12) CHAPTER 1. INTRODUCTION. or local, metallicity of SNe explosions also seek to provide constraint on the different progenitors [Prieto et al., 2008, Modjaz et al., 2008, Arcavi et al., 2010, Anderson et al., 2010, Stoll et al., 2013, Galbany et al., 2014, Anderson et al., 2016, Kuncarayakti et al., 2018]. The spatial distribution of CCSNe provides yet another constraint on the nature of their progenitors as given by the comparison of the overall distribution of stellar populations [e.g. Van Dyk et al., 1999, Petrosian et al., 2005, Mikhailova et al., 2007, Kangas et al., 2017]. Anderson and James [2009] and Hakobyan et al. [2009] studied the radial distribution of SNe with different approaches and found that Type Ib/c SNe are more centrally concentrated in their host galaxies. This is expected given the usual higher metal enrichment in the center of galaxies [e.g. Henry and Worthey, 1999, Sánchez et al., 2014, Sánchez-Menguiano et al., 2016]. Hakobyan et al. [2017] studied the height distribution of SNe in edge-on disk galaxies and found that CCSNe are more concentrated (nearly twice) to the disk than thermonuclear SNe. This is consistent with the height distribution of stellar populations expected to give rise to both CCSNe and thermonuclear SNe. Some SNe are outliers to this typical behavior. One striking supernova transient is the restless SN 2009ip [e.g. Fraser et al., 2013, Mauerhan et al., 2013, Pastorello et al., 2013, Prieto et al., 2013]. This supernova was first identified as a SNe impostor in 2009 and after 3 years the progenitor star finally exploded as a Type IIn supernova [Smith et al., 2014]. SN 2009ip exploded ∼ 5 kpc away from the nearest spiral galaxy NGC 7259 center and the evidence from late time HST data rules out the presence of any significant star-forming region similar to Carina or the Orion Nebula [Smith et al., 2016]. The possibility that the progenitor was a runaway star was also rejected because the high mass of the progenitor would give it a very short time to reach the nearest star forming site located at ∼ 1.5 kpc, having velocities less than 400 km/s. Located a mere 2.4’ (∼18 kpc in projected distance) from SN 2009ip, the Type II 11.

(13) CHAPTER 1. INTRODUCTION. supernova ASASSN-14jb [Brimacombe et al., 2014, Challis, 2014, Zhang and Wang, 2014] exploded 2.5 kpc from the center and 2.1 kpc from the disk of the edge-on disk galaxy ESO 467-G051 (part of an interacting pair with NGC 7259), where no important star forming region is detected. In this work we present extensive photometric and spectroscopic observations of ASASSN-14jb and its explosion site using Integral Field Spectroscopy (IFS), which are used to constrain the SN progenitor properties. We also use the IFS observations to study its environment and discuss the implications for the explosion sites of CCSNe. This work is organized as follows. In Section 2 we describe the photometric and spectroscopic observations. Section 3 contains our analysis of the observations of the supernova and its host. In Section 3.1 we study the photometric and spectroscopic evolution of ASASSN-14jb and compare these to other Type II SNe. Our distance measurements and comparison with previous distance measurements are presented in Section 3.2. In Section 3.3 we derive physical parameters of ASASSN-14jb. In Section 3.4 we obtain and discuss the physical parameters of the host galaxy and its H II regions. We finally discuss the results and present our conclusions in Section 4.. 12.

(14) Chapter 2 Data 2.1. Photometry. Optical photometric observations of ASASSN-14jb were obtained with the ASAS-SN unit “Cassius” at CTIO in V -band and the Las Cumbres Observatory Global Telescope Network [LCOGTN; Brown et al., 2013] 1-meter telescopes at CTIO, the Siding Spring Observatory (SSO), the South African Astronomical Observatory (SAAO), and the McDonald Observatory (MDO) in the BV bands and the Sloan Digital Sky Survey (SDSS) gri filters. All the ASAS-SN images are processed in an automated pipeline using the ISIS image subtraction package [Alard and Lupton, 1998, Alard, 2000].. The fully reduced LCOGTN 1m images (bias/overscan subtracted and flat-fielded) were retrieved from the LCOGTN data archive. We solved the astrometry of each CCD frame using the astrometry.net software package [Lang et al., 2010]. We obtained calibrated BV gri magnitudes for local standard stars in the field from the APASS catalog. In Figure 4.2 we present the field of one of the LCOGT images showing the position of the SN and the position of the local standards. We also show the position of the very well-studied and nearby SN 2009ip [e.g., Fraser et al., 2013, Mauerhan 13.

(15) CHAPTER 2. DATA. et al., 2013, Pastorello et al., 2013, Prieto et al., 2013], which exploded in NGC 7259. In Table 4.1 we present the local standard star coordinates and magnitudes in the BV gri system from APASS, where the BV magnitudes are in Vega system and the gri magnitudes are the SDSS AB system. Using the IRAF apphot package, we performed aperture photometry on the ASASSN subtracted images and then calibrated the results using the AAVSO Photometric All-Sky Survey [APASS; Henden et al., 2012]. The ASAS-SN photometry is presented in Table 4.2. We extracted the photometry of ASASSN-14jb from the LCOGT images using aperture photometry directly given the isolation of the SN in its host galaxy and the high signal-to-noise ratio (SNR) of the detections. The procedure used to extract and calibrate photometry from LCOGT images is explained in detail in Appendix A. The LCOGTN photometry is presented in Table 4.3. As we lacked ASAS-SN and LCOGTN photometric observations of ASASSN-14jb obtained after the initial plateau phase of the light curve, we searched public data available for the field of the nearby SN 2009ip in the ESO data archive1 . We retrieved ESO/NTT EFOSC images in BV filters that contained ASASSN-14jb’s explosion site at three different epochs after October 2014. These images were obtained by the Public ESO Spectroscopic Survey of Transient Objects [PESSTO; Smartt et al., 2015]. The CCD frames were overscan subtracted and flat-fielded using the available calibration frames obtained the same day with standard routines in IRAF. Since the SN has lower SNR in the late-time images, we extracted the photometry of the SN using PSF fitting technique with the DAOPHOT package in IRAF, instead of aperture photometry, and we calibrated the magnitudes using local standards with photometry presented in Pastorello et al. [2013]. The ESO/NTT photometry is also reported in Table 4.3. We also supplemented our photometry with Swift U BV W 1M 2W 2 optical and near1. http://archive.eso.org/eso/eso archive main.html. 14.

(16) CHAPTER 2. DATA. UV observations of ASASSN-14jb obtained from the Swift Optical/Ultraviolet Supernova Archive [SOUSA; Brown et al., 2014]. Figure 4.3 shows the near-UV and optical light curves of ASASSN-14jb from the ASAS-SN, LCOGTN, ESO/NTT, and Swift/UVOT observations.. 2.2. Spectroscopy. A total of ten single-slit, low-resolution optical spectra of ASASSN-14jb were obtained with the FAST spectrograph [Fabricant et al., 1998] mounted on the Fred L. Whipple Observatory Tillinghast 1.5-m telescope (3500 − 7400 Å, R ∼ 1700) and with the InamoriMagellan Areal Camera and Spectrograph [IMACS; Dressler et al., 2011] mounted on the Magellan Baade 6.5m telescope at Las Campanas Observatory (3800 − 9800 Å, R ∼ 1000). The CCD images were reduced and the 1D spectra were extracted and calibrated using standard routines in the IRAF packages ccdproc and onedspec, respectively. The epochs and relevent parameters of the ten spectra obtained in the plateau phase are presented in Table 4.4. A montage of the spectra obtained is shown in Figure 4.4. We also observed the field of ASASSN-14jb as part of the All-weather MUse Supernova Integral field Nearby Galaxies (AMUSING; Galbany et al. 2016a, Prieto et al. 2016) with the Multi Unit Spectroscopic Explorer (MUSE, Bacon et al. 2010) on ESO’s Very Large Telescope UT4 (Yepun). MUSE is a state-of-the-art integral field spectrograph with a field of view of 1 arcmin2 and 0.2 spaxels. It covers the spectral range 4800 − 9300 Å with a resolving power of R ' 1800 − 3000. Our MUSE data were obtained on 2015-11-14 and consisted of two different pointings of four dithered exposures each with an integration time of 698 sec. The sky conditions were clear, and we measure a full-with-half-maximum for the stellar point-spread function of 1.08 at 6600 Å. We reduced the MUSE spectroscopy with version 1.2.1 of the pipeline provided by ESO [Weilbacher et al., 2014], which applies a bias, flat-field, illumination, field geom-. 15.

(17) CHAPTER 2. DATA. etry correction and background subtraction. The wavelength solution was tied to arc lamp frames, refined using skylines in the science data and converted into the heliocentric reference frame. Observations of the spectrophotometric standard GD108, taken immediately after the science data, provided the flux calibration. The final product of this reduction process is a data cube with two celestial coordinates sampled at 0.2 and one wavelength dimension sampled at 1.25 Å [Prieto et al., 2016]. We checked and corrected the astrometric zeropoint of the MUSE data using LCOGTN images of the SN field. Although the MUSE observations were obtained during the nebular phase of ASASSN14jb, ∼ 400 days after explosion (see Table 4.4) when the SN was quite faint (V & 22 mag), the quality and depth of the MUSE data allow us to easily detect the SN and extract a nebular phase spectrum. We used a circular aperture with a radius of 1.2 to extract the spectrum at the position of the SN with QFitsView2 . The nebular spectrum of ASASSN-14jb and the properties of its host galaxy are discussed in Sections 3.3.3 and 3.4, respectively.. 2. http://www.mpe.mpg.de/ ott/dpuser/qfitsview.html. 16.

(18) Chapter 3 Analysis 3.1. Evolution and Comparison with other Type II Supernovae. To study and put ASASSN-14jb in context, we used public data available for a set of Type II SNe that span a wide range of properties, including the luminous Type II SN 2007pk [Inserra et al., 2013], the normal Type II-P like SN 1999em [e.g. Hamuy et al., 2001], and the subluminous Type II SN 2005cs [e.g. Pastorello et al., 2009]. Our sample has also been defined according to the availability of: 1) Good coverage of the photospheric phase including gri photometry if available (8 SNe); and 2) optical nebular spectra at ≈ 400 days past explosion (6 SNe), used for comparison in Section 3.3.3. Although we do not make any distinction using the Type II-P/II-L nomenclature, our comparison sample is in the range usually applied for Type II-P like SNe, i.e., the decline slope in the V -band for our comparison sample, with the exception of SN 2007pk (for reasons that will be explained later), is . 1 mag per 100 days. Our sample is not complete in a statistical sense but spans a range of observed properties wide enough to encompass the diversity of typically normal Type II SNe and serve, thus, as a fair set for. 17.

(19) CHAPTER 3. ANALYSIS. comparison. Basic properties of the SNe in the sample and references for the data are given in Table 4.6. With this sample, we now study the photometric and spectroscopic evolution of ASASSN-14jb. We assume a distance modulus of µ = 32.0 mag to its host galaxy (see distance determination below) and a total reddening of AV = 0.05 mag (also to be justified below).. 3.1.1. Photometric Evolution. A Galactic color excess towards ASASSN-14jb of E(B − V ) = 0.0154 ± 0.001 mag (average of a 5 arcmin circle around the explosion site) was obtained using the SchlaflyFinkbeiner map for Galactic reddening [Schlafly and Finkbeiner, 2011]. The host galaxy extinction towards the SN is low, as the SN exploded at a significant distance from the edge-on disk of its host galaxy and the MUSE data shows no significant star-forming region near the explosion site. We also do not detect a narrow Na I D doublet at the host galaxy redshift, pointing to a low host extinction [Phillips et al., 2013]. In our analysis we assume a total extinction of AV = 0.05 mag, together with a standard Galactic reddening law with a total to selective extinction ratio of RV = 3.1. Figure 4.5 shows the evolution of the absolute V -band magnitude of ASASSN-14jb compared with the other Type II SNe presented in Table 4.6, as a function of the time in days since their estimated time of explosion. The peak absolute magnitude in the V -band of MV,max = −16.04 ± 0.18 mag for ASASSN-14jb is within the range of normal Type II-P like SNe, although lower than the average absolute V -magnitude −16.71 ± 1.01 mag in the sample of Type II SNe of Anderson et al. 2014a (hereafter J14). ASASSN-14jb is dimmer than the classic, well-studied Type II-SN 1999em by 0.80 ± 0.18 mag, but it has a similar behavior in the V -band plateau, including the early double peak [Hamuy et al., 2001, Leonard et al., 2002a]. Up to ∼ 80 days since explosion the photometry shows the typical behaviour of Type II-P like SNe. The plateau slope of ASASSN-14jb in the V-band, measured between 35 − 70 days 18.

(20) CHAPTER 3. ANALYSIS. after explosion, is s2,V = −0.15 ± 0.02 mag per 100 days, lower than the average V band plateau slope in the sample of A14 (1.27 ± 0.93 mag per 100 days). The negative sign indicates that the plateau brightness increases slightly as a function of time after ∼ 35 days for ASASSN-14jb. This is consistent with the correlation between the plateau slope and absolute V -band magnitude found by A14 and later works [e.g. Valenti et al., 2016]. In Figures 4.6 and 4.7 we present the extinction-corrected color evolution of ASASSN14jb compared with the color evolution of the Type II SNe presented in Table 4.6. The (B − V )0 color of ASASSN-14jb is bluer than average at 50 days after explosion, being comparable only to LSQ 13fn [Polshaw et al., 2016] and SN 2007pk [Inserra et al., 2013] at 80 days after explosion. The change in color from the onset of the plateau up to 80 days is ∆(B − V ) = 0.20 ± 0.04 mag, low compared with the reference sample. The color evolution of ASASSN-14jb in (g − r)0 , (g − i)0 and (r − i)0 also points at a bluer continuum compared to other Type II SNe, except in the (V − i)0 color. This behavior may be due to a true difference in the temperature evolution or it may be an effect of lower line-blanketing in the blue part of the spectrum, due to the low-metallicity of the progenitor. The similarity in the color evolution of ASASSN-14jb to LSQ 13fn, a Type II SN that probably comes from a low-metallicity progenitor [Polshaw et al., 2016], might point to a similar physical mechanism.. 3.1.2. Spectroscopic Evolution in the Plateau Phase. As shown in Figure 4.4, the spectral evolution of ASASSN-14jb resembles the evolution of other hydrogen rich Type II SNe. At very early times the ejecta is optically thick and the spectra show a hot blue continuum with relatively weak, but broad, Balmer lines without P-Cygni absorption throughs. In the first spectrum at ∼ 4 days after explosion, we detect broad Balmer lines in emission and the high ionization He II λ4686 line. The FWHM of the Balmer lines and the He II lines in the earliest spectrum are 19.

(21) CHAPTER 3. ANALYSIS. ∼ 11000 − 13000 km/s and ∼ 8000 km/s, respectively, and all the emission peaks are clearly blueshifted by ∼ 2000 − 4000 km/s. The strength of the He II line decreases substantially in the spectrum obtained a day later, at 5 days after explosion, and the Balmer lines start to develop P-Cygni profiles. We also detect the He I λ5875 line in the early spectra. At 6 days after explosion the He II line has weakened further and it is undetected in the spectrum obtained at 7 days after explosion. At later phases (& 30 days) the ejecta cools, the hydrogen recombination sets fully, and the P-Cygni profiles show strong absorption troughs. The blueshift of the peak of the Balmer lines decreases to ∼ 1000 km/s. This is a common feature in Type II SNe during the photospheric epochs and it has been attributed to the shallow density profile of the ejecta compared to a typical stellar wind [Dessart and Hillier, 2005b, Anderson et al., 2014a]. Transitions from metal lines such as Fe II, Sc II, and Ti II start to appear in the spectra and contribute significantly to line blanketing. The Na I D doublet lines at λλ5890, 5896 are detected at these epochs instead of He I λ5875, which is detected at earlier epochs in a similar wavelength range. In Figure 4.8 we present a comparison of photospheric phase spectra of ASASSN-14jb and other well studied Type II-P like SNe from our comparison sample, at ∼ 60 days past explosion. Compared to our sample, we observe relatively weak absorption features in the range 5300 − 6300 Å. In Figure 4.9 we present the evolution in time of Hα, Hβ and the Fe II λ5169 features. The last two spectra we obtained in the plateau phase, at 64 and 80 days after explosion, show clear absorption features on the blue side of both the Hα and Hβ main absorption troughs. These High Velocity (HV) features are at ∼ 7000 km/s, while the main absorption is at an expansion velocity of ∼ 4000 km/s. These absorption lines cannot be explained by Si II λ6355 or Ba II absorption features. These HV features in the Balmer lines are fairly common and have been observed on several Type II SNe. In a sample of 123 Type II SNe, [Gutiérrez et al., 2017b] found that 59% showed a HV feature. As Chugai 1994 explained these HV features arise as an effect of the increased opacity 20.

(22) CHAPTER 3. ANALYSIS. at a dense shell that forms through interaction of the ejecta with the circumstellar wind. The scenario of the HV feature being a Si II line requires a high temperature at that epoch. In the case of a detection in the plateau (after 30 days as in Gutiérrez et al. [2017b]) a large progenitor radius would be required to sustain the slower cooling.. 3.1.3. Expansion Velocities. The ejecta of the Type II SNe achieve near homologous expansion after a few days [Bersten et al., 2011] and at early times the hydrogen atmosphere is opaque and the line forming region is in the outer, higher velocity layers. As the temperature drops, the opacity decreases and the photosphere recedes in mass (Lagrangian) coordinate, and therefore the photosphere is at lower velocities. The photosphere is tightly bound to the opacity drop of hydrogen [Bersten et al., 2011]. In practice one usually measures the velocity of maximum absorption of a specific P-Cygni profile that can over or under estimate the true photospheric velocity [Dessart and Hillier, 2005b, Takáts and Vinkó, 2012]. Other methods used are the cross-correlation with library spectra, the comparison with detailed NLTE codes like CMFGEN or PHOENIX [Dessart and Hillier, 2005a, Baron et al., 2005] or comparison with more simplified, parametrized LTE spectral modeling (e.g. SYNOW, Takáts and Vinkó 2012). The expansion velocity measured from optical spectra is an important physical parameter directly related to the explosion energy and the plateau luminosity. These are also fundamental for distance measurements [Hamuy and Pinto, 2002]. Strong Balmer lines and Fe II lines velocities were measured in our spectra, when possible, fitting low degree polynomials to the P-Cygni profile absorptions and taking the minimum as a proxy for the expansion velocity. The spectra were normalized to a global continuum fitting a black body or a local power law. No reddening correction was applied to the spectra and we do not attempt to correct for peculiar velocities. In Figure 4.10 we present the measured expansion velocities for ASASSN-14jb and in Figure 4.11 we 21.

(23) CHAPTER 3. ANALYSIS. present the expansion velocities derived for ASASSN-14jb and the comparison sample. In the case of LSQ 13fn, due to the low signal to noise of the spectra, we only present the velocities obtained after 30 days. The values of SN 2013ab are directly taken from Bose et al. [2015]. In Table 4.5 we present the measured velocities, together with the average velocity of our comparison sample and the sample of Gutiérrez et al. [2017b]. To compare with well-known correlations [Hamuy and Pinto, 2002, Faran et al., 2014], we interpolated the velocities of our sample at 50 days after explosion using a Monte-Carlo approach. For each SN we generated a re-sampling of each velocity using the measured errors and following a non-linear least squares procedure, using the curve fit function from the NumPy library, we fit a power law model each time. For this procedure we select only velocities after 20 days, as early velocities showed do not follow the same decay rate as in the plateau. We then take the average and standard deviation of the re-samplings as the 50 days velocity and error, respectively. For all the lines measured the velocities for ASASSN-14jb are slightly under average. The Fe II λ5169 velocity interpolated at 50 days is measured to be 2774 ± 69 km/s, compared to the average of 3442 ± 977 km/s of our comparison sample. Our sample average velocity is in agreement with the value obtained with a larger sample in Gutiérrez et al. [2017a], of 3537 ± 851 km/s at 53 days after explosion. The photospheric velocities inferred are in agreement with the velocity-luminosity correlations (see section 3.2 below). As it is common in type II SNe distance measurements [Hamuy and Pinto, 2002, de Jaeger et al., 2017b], we fitted a power law to interpolate the velocities. The Fe II λ5169 logarithmic decay presents an average of −0.55 ± 0.16 in our sample, compared to the value obtained for ASASSN-14jb, of −0.62 ± 1.80. We observe a strong anti-correlation between the velocity decay slope and the V -band absolute magnitude at 50 days. This is expected as there is an internal correlation between both parameters of the power law [de Jaeger et al., 2017b] and the velocity scale correlates with the luminosity [Hamuy and Pinto, 2002]. 22.

(24) CHAPTER 3. ANALYSIS. In conclusion, ASASSN-14jb presents expansion velocities consistent, both from the scale and velocity decay, with an under average luminosity type II-P like SNe.. 3.1.4. Nebular Spectrum. The nebular spectrum of ASASSN-14jb obtained with MUSE at ∼ 394 days after the time of explosion, showed in Figure 4.17 together with a sample of Type II SNe, shows the typical nebular emission lines seen in normal Type II-P like SNe: Hα, the doublet [Ca II] λλ7291, 7324, the doublet [O I] λλ6300, 6364, the triplet Ca II λλλ8498, 8542, 8662, the doublet [Fe II] λλ7155, 7172, that presents a boxy profile, the doublet Na I λλ5890, 5896 and [O I] λ5577, in order of observed intensity [Silverman et al., 2017]. Also, a strong [C I] λ8727 line is present, which has been detected on other Type II-P like SNe such as SN 2004dj [Silverman et al., 2017]. A weak He I line at λ7065 is observed but with no He I counterpart at λ6678. The clear separation observed in different multiplets like [OI], [CaII] and Ca II + [CI] is a sign of a low explosion energy for ASASSN-14jb. The blueshift observed in the peaks of the strongest lines, of approximately ∼ 200 km/s is particularly interesting. This cannot be attributed to the same physical effect as in the plateau phase because in this optically-thin phase the outer density profile is less relevant to the emission [Anderson et al., 2014a]. Other hypotheses to explain this blueshift may be: 1) a shift of relative velocity against the host average due to dynamical effects or due to an intrinsically high-velocity for the progenitor star; 2) dust production; 3) Some lines are still optically thick at 400 days past explosion; or 4) an asymmetry in the explosion itself. Notable is the double peak observed in Hα which is not observed in other lines, although this is also seen in the nebular spectra of other Type II-P like SNe [Elmhamdi et al., 2003, Chugai et al., 2005, Utrobin and Chugai, 2009].. 23.

(25) CHAPTER 3. ANALYSIS. 3.2. Distance. Type II SNe distance measurements can be categorized in theory based methods, such as the Expanding Photosphere Method [EPM, Kirshner and Kwan, 1974, Schmidt et al., 1992], and the spectra-fitting expanding atmosphere method [SEAM, Baron et al., 2004], and the empirical ones such as the Standard Candle Method [SCM , Hamuy and Pinto, 2002]. Great improvement in Type II SN distance determination has been achieved recently. Rodriguez et al. 2018 (in prep. henceforth R18) obtained a Hubble diagram with a 0.12 mag dispersion using near-infrared observations, improving the optical calibrations presented in Rodrı́guez et al. [2014] (from hereby R14) using the Photospheric Magnitude Method (PMM). de Jaeger et al. [2017b] (henceforth DJ17) built a high redshift (z < 0.5) Hubble diagram using CSP-SDSS-SNLS data applying their recent method called the Photometric Color Method (PCM), that avoids using spectral velocities, showing the potential for future photometric-only surveys that will use Type II SNe for Cosmology. In this section we outline and make use of the most recent empirical methods to measure the distance to ASASSN-14jb. We do not apply a K-correction in the following analysis due to the low redshift of the source, which for the V-band corresponds to less than 0.01 mag.. 3.2.1. Photospheric Magnitude Method. The PMM is a method to measure distances to Type II SNe that relies on a timevariable standardization of the photospheric magnitude evolution. This method can be viewed as a generalization for various epochs of the SCM (Hamuy and Pinto 2002) in the sense that both rely on the physical connection between the expansion velocity (a measure of the explosion energy) and the plateau luminosity, but the PMM has a different approach in the calibration. For a distance to be measured we need a good 24.

(26) CHAPTER 3. ANALYSIS. constraint on the shock-breakout epoch, magnitudes corrected from both host and local extinction and corrected by the K-correction, and expansion velocities inferred from the minima of the P-Cygni profiles (usually from the Fe II lines) in the photospheric phase. Explicitly we have, − Mλ − R µλ = mcorr λ R = −5log (Rph ) + 5 = −5log (vph (t − t0 )) + 5. (3.1). is the rest-frame and extinction corrected magnitude at the wavelength λ, where mcorr λ Mλ is the photospheric magnitude, R is the factor that considers the surface area of the photosphere and depends on the photospheric radius, which for homologous expansion is simply the product of photospheric velocity vph and the time since the beginning of the expansion t−t0 (see R14 for details). Although we have named t0 the “explosion epoch”, this reference time is actually associated with the shock breakout. As our constraint on t0 is of the order of the shock propagation time to reach a RSG surface after core-collapse [Arnett, 1996] we do not make an explicit distinction. We have measured the velocities of ASASSN-14jb from the Fe II λ5169 line (see Figure 4.10) and the extinction corrected magnitudes in the V -band without applying a K-correction. Averaging the distances obtained at the three epochs with Fe II λ5169 velocities, we obtain an average distance modulus of hµiP M M = 31.94 ± 0.06 mag. To make a direct comparison of the distance measurements of other methods we need to take in consideration that the PMM was calibrated using local distance ladder measurements, so in the method there is an implicit value of the Hubble constant. As the value reported from R14 is H0 = 68 km/s/Mpc for the V -band calibration, we will take this value as reference in the following analysis. We also use now the CMB frame redshift of zCM B = 0.00505 for the ASASSN-14jb host, obtained from NED.. 25.

(27) CHAPTER 3. ANALYSIS. 3.2.2. Standard Candle Method. The most common method to measure Type II-P like distances is the SCM which relies on the tight correlation between brightness and the expansion velocity at a given epoch in the plateau phase. Similar to the Type Ia SNe calibrations, the SCM parameters are fitted directly in the Hubble diagram following the model [Hamuy and Pinto, 2002], Mλ1 = −αlog (vph ) + β(Mλ2 − Mλ1 ). (3.2). where Mλk is the absolute magnitude in the k-band, vph is the photospheric velocity, usually at 50 days, and (α, β) are the parameters to fit using a statistically significant sample. The distance modulus in the DJ17 prescription takes the form, v(Hβ) µλ1 = mλ1 + αlog − β(mλ2 − mλ1 ) − 5 log(H0 ) − MSCM + 25 < v(Hβ) >. (3.3). where the brackets denote the sample average. We use the parameters obtained it DJ17 for the SDSS i, (r − i) color combination1 . With the interpolated velocity in Hβ at 40 days, using a power law model, of v(Hβ) = 4181 ± 146 km/s and the color at 45 days (r − i)0 = 0.03 ± 0.13mag, the distance modulus is µSCM = 32.17 ± 0.20 mag.. 3.2.3. Photometric Colour Method. The PCM exploits the correlation between the light curve decline rate at the plateau, s2 , and the luminosity [Anderson et al., 2014a]. In this method the distance modulus is obtained via,. µP CM = i − αP CM s2,i + βP CM (r − i)0 + MP CM + 5log(H0 ) − 25. (3.4). where the values of the parameters taken from DJ17 are αP CM = 0.39 ± 0.08, βP CM = 0.8 ± 0.48, MP CM = −0.84 ± 0.08. Measuring the slope in the i-band light curve we 1. αSCM = 3.18 ± 0.41, βSCM = 0.97 ± 0.25, MSCM = −1.13 ± 0.04. 26.

(28) CHAPTER 3. ANALYSIS. find s2,i = −0.365 ± 0.187. With the corresponding color (r − i)0 = 0.027 ± 0.16 we obtain µP CM = 32.4 ± 0.2 mag.. 3.2.4. Adopted Distance. As we show in Figure 4.12, all the distances obtained through different methods are consistent with each other if we consider each method’s intrinsic dispersion. They are also consistent with the distance obtained from the Virgo-inflow corrected recession velocity of the host galaxy and the Hubble law, assuming H0 = 68 km/s/Mpc. Given this, we simply take as the final distance the weighted average, considering the statistical errors only, of µ = 32.00 ± 0.18 mag (D = 25.1 Mpc). In Figure 4.12 we also show the distance obtained for SN 2009ip from Potashov et al. 2013, which represent a distance difference of ≈ 4.8 Mpc between the host galaxies of these SNe.. 3.3. Physical parameters. Progenitors of type II-P like SNe have been constrained by pre-explosion images to be RSGs in the range of initial masses of ≈ 8 − 17M. [Smartt et al., 2009, Smartt,. 2015]. The discrepancy between these initial mass constraints and the predictions from evolutionary codes [Heger et al., 2003, Ekström et al., 2012] has been called the “RSG problem”. In this section, we use our observations for ASASSN-14jb to constrain the basic physical parameters of the progenitor and compare thems with the expected values for a type II-P like progenitor.. 3.3.1. Temperature Evolution and the Progenitor Radius. The early emission, a few days after explosion, for a Type II SN comes from the adiabatic-cooling of the outer envelope and depends mainly on the pre-supernova radius. 27.

(29) CHAPTER 3. ANALYSIS. [Rabinak and Waxman, 2011, Sapir and Waxman, 2017]. Observationally, the presupernova radius has been constrained from early UV and optical photometry, using semi-analitic modeling [Rabinak and Waxman, 2011, Nakar and Sari, 2010], on individual Type II SNe [Gezari et al., 2010, Valenti et al., 2014, Bose et al., 2015, Huang et al., 2016, 2018, Tartaglia et al., 2018], and on larger samples to make statistic studies [González-Gaitán et al., 2015, Gall et al., 2015, Rubin et al., 2016]. In this subsection, we build the spectral energy distribution (SED) for ASASSN-14jb at early times, from ∼4 to 26 days after explosion and with the available analytic models we derive an estimate of the progenitor radius. Using the Swift/UVOT near-UV and optical photometry, we build the SED of ASASSN-14jb at each epoch, after correcting for extinction assuming a standard Galactic extinction law with total to selective extinction ratio RV = 3.1 [Cardelli et al., 1989]. We fit a blackbody function to the SED and obtain the photospheric temperature and radius evolution, assuming the distance to the host galaxy obtained in Section 3.2. The results are presented in Figure 4.14. The observed photospheric temperatures are in the range 5-21×104 K and the radii are in the range 2-18×103 R . We observe a change in the evolution after ∼ 10 days after explosion. The observed temperature logarithmic decay increases from −0.726 ± 0.055 to −0.850 ± 0.019 and the photospheric radius growth rate increases from 0.890 ± 0.012 to 1.45 ± 0.06 at later times. Rabinak and Waxman 2011, now RW11, developed analytic expressions for the early evolution of the photospheric radius Rph and temperature Tph , which are expected to be valid before hydrogen recombination is significant. These expressions depend on the progenitor properties before explosion such as radius R∗ , opacity κ, and density profile. They also depend on the explosion energy E and ejecta mass Mej . The equations are, −0.31 0.093 0.81 0.41 Rph = 3.3fρ−0.062 E51 Mej κ0.34 t5 × 1014 [cm] 1/4. −0.054 0.0027 −0.28 −0.45 Tph = 1.6fρ−0.037 E51 Mej R∗,13 κ0.34 t5 [eV ]. 28. (3.5) (3.6).

(30) CHAPTER 3. ANALYSIS. where 0.079 < fρ < 0.13 for RSGs, depends on the progenitor density profile, Mej is the ejected mass on solar units, E = E51 × 1051 erg is the explosion energy , κ34 = 0.34 × κ is the opacity, R∗ = R∗,13 × 1013 cm is the progenitor radius and t = t5 × 105 s is the time since shock breakout. It is well known that the emission from Type II-P SNe differs from a pure blackbody [Eastman et al., 1996, Dessart and Hillier, 2005b]. When the atmosphere is highly ionized, the electron scattering opacity dominates and the photon thermalization layer differs from the photosphere, the former being underneath at higher temperatures. Because of this the observed temperature (or color temperature Tc , as it is usually called) can differ significantly from the photospheric temperature Tph . This effect, together with the increasing line opacity at later times (line blanketing) and departures from the plane parallel atmosphere approximation, can alter significantly the observed SED. Typically this effect can be corrected using the dillution factor ζ 2 , so that the emergent flux is defined as, 2. Fλ := θ πBλ (Tc ) =. ζλ2. . Rph dL. 2 πBλ (Tc ). (3.7). where the dilution factor ζλ2 is in general wavelength dependent, dL is the luminosity distance and Bλ (Tc ) is the Planck function. The dilution factor directly affects the angular diameter distance estimate to a supernova, but, as we take the luminosity distance given by empirical methods, the dilution factor only affects the photospheric radius estimation and not the temperature. Note, however, that in the RW11 scheme ζ := 1/fT2 ≈ 0.694, which translates in a factor of 1/ζ ≈ 1.44 increase in the photospheric radius (see equation 38 in RW11). In our case we choose a representative value of ζλ = 0.5 for this high temperature regime, as obtained in Dessart and Hillier [2005b], Pejcha and Prieto [2015a], that is, we scale all our Rph values by a factor of two. To test the effect of the deviations from a blackbody in the SED we used a semiempirical approach. We first selected the 15 M. progenitor mass supernova explosion. models presented in Dessart et al. [2013] that best fits the observed near-NUV colors. 29.

(31) CHAPTER 3. ANALYSIS. The best model in this case was the m15z8m8 (see Dessart et al. 2013 for the model details). Then, at each epoch we fitted the SED following the following modified model, Fλdata. 2. . w. = θ (πBλ (Tc )) ×. Fλmodel max(Fλmodel ). (1−w) (3.8). which is a weighted geometric mean of a blackbody and the selected normalized explosion model. The normalization ensures that deviations in the model’s intrinsic brightness do not affect the fitted parameters. In principle, the weighting parameter w is unconstrained but to ensure a smooth correction we fixed its value between 0.6 − 0.9 and selected the model with the minimum residuals. Figure 4.14 show our results for both approaches. We observe that by including the models, both the temperature and radius evolution become smoother. We think this is because we are correctly accounting for line blanketing using our hybrid approach. We note that all the models used here start from ≈ 10 days past explosion so all the SEDs before that time were fitted with a single blackbody in all cases. The derived slope in the temperature changes to −0.703 ± 0.017 after 10 days. The photospheric radius follows roughly a linear expansion with a logarithmic slope 1.167 ± 0.068 in the same epochs. Now we study the RW11 models with our results using the modified blackbody. We fitted the temperature evolution assuming the fiducial values of E51 = 1,κ34 = 1,fρ = 0.1. We obtain a progenitor radius with a value of 504 ± 11 R and we can change its value up to only ≈ 20% for a reasonable range of values of the fρ and κ parameters. As discussed in RW11, the models are not expected to be reliable after recombination starts to be important and they mention that a good reference is that the temperature cannot be . 1 eV. If we redo the fit with the points that fulfill this criteria we obtain a larger radius of 985 ± 49 R . If we assume that these estimates are not accurate enough because the SBO date is inaccurate, we can shift the chosen SBO epoch and what we found is that the radius is decreased. Where the fit is optimal, according to χ2 minimum, the radius is 579 ± 28 R for fit at early epochs, with a shift in the SBO of −1.75 days, and a reduced χ2 = 3.62. Another way 30.

(32) CHAPTER 3. ANALYSIS. to see the variations in the derived progenitor radius more easily is inverting Equation 6 (a corrected version of Equation 33 in RW11). The original equation 33 of RW11 is 12 R∗ = 0.7×fρ0.1 [Tcol /(fT /1.2)]4 t1.9 5 ×10 [cm], but this is inconsistent with Equation 13 of. the same paper (our Equation 5), where fT = Tcol /Tph is the time dependent ratio of the color temperature Tcol and true photospheric temperatures Tph . RW11 uses fT = 1.2 as a good approximation at early times and we assume the same value for our calculations. The corrected equation for the progenitor radius can then be written as, 4 Tcol (t) 0.148 R∗ ≈ 1.52 × fρ t51.8 × 1012 [cm]. (fT /1.2). (3.9). The values obtained using Equation 9 at early times are consistent with R∗ ≈ 400 − 1000 R and, as expected, the radius decreases as 1) the actual decay of the photospheric temperatures is faster than the model, and 2) the correction between the color and photospheric temperature changes due to the changing opacity of the ejecta. Given that RW11 seems to be unable to explain the data for ASASSN-14jb (see Figure 4.14), we fit the photospheric temperature with the decay rate as a free parameter. In this case the radius becomes extremely degenerate with the chosen explosion epoch and no strong constrain can be obtained. With this analysis, we choose as our best estimate of the progenitor radius, the RW11 fit with an SBO offset of −1.75 days, which gives an estimate of R∗ = 579 ± 28 R . This value is consistent with the color evolution of the mildly sub-solar (Z = 0.4 Z ) and compact (R∗ = 611 R ) RSG progenitor (m15z8m3) used in Dessart et al. [2013]. As the photospheric radius growth is strongly dependent on the explosion energy to ejecta mass ratio (equation 9), with our fits, fixing κ34 = 1,fρ = 0.1 as before, we can obtain the ejecta mass as a function of the explosion energy. With this we obtain that 1.32 Mej ≈ 36.0 × Eexp which will be useful in the following.. 31.

(33) CHAPTER 3. ANALYSIS. 3.3.2. Bolometric Light Curve and the. 56. Ni Mass. To construct the bolometric light curve of ASASSN-14jb we used the bolometric corrections of the semi-analytic model presented by Pejcha and Prieto [2015a] on the extinction corrected B − V colors. In Figure 4.13 we present the bolometric light curve for ASASSN-14jb, against the comparison sample of SNe II, which covers ≈ 1.5 orders of magnitude in brightness at the plateau. We measure the bolometric luminosity at ∼ 50 days past explosion to be log (L/L ) = 8.4 ± 0.18 for ASASSN-14jb. This is consistent to be under-luminous compared to the average Type II-P like SN 1999em, SN 2012aw or SN 2013ab. From the late-time photometry obtained with EFOSC imaging and the bolometric light curve, we estimate a 56 Ni mass of 0.0210 ± 0.0025 M , which is slightly lower than the median in the distribution of normal Type II SNe of 0.031 M [Müller et al., 2017]. For estimating the. 56. Ni mass we assume a model of partial trapping of gamma rays. [Clocchiatti and Wheeler, 1997], where the intensity of the pure Cobalt decay is scaled by the transparency factor I(t) = (1 − exp (−(T0 /t)2 )). The late-time light curve and model fit are presented in Figure 4.15. In this model, assuming a fixed density slope and gamma-ray opacity (η = 10, κγ = 0.06 and an electron fraction per nucleon Ye = 0.03), T0 depends on the ejecta mass and explosion energy such that T0 = C(η, κγ )M E −0.5 . We obtain T0 = 319 ± 33 days from the best-fit model, which would correspond to an ejecta mass of 12 ± 1 M for an explosion energy of 1 foe = 1051 erg. Although the faster than Cobalt decay is expected in more diluted ejecta, as in the case of stripped enveloped SNe and some faster declining Type II SNe such as SN 2014G, SN 2013hj and ASASSN-15nx [Bose et al., 2016, Terreran et al., 2016, Bose et al., 2018], the estimated ejected mass is consistent with a low mass range Type II-P progenitor, given the low energetics shown in, for example, the expansion velocities (Eexp < 1 foe). The ejected mass derived from this analysis can be compared with the one obtained with early light curve analysis 0.5 done in the previous section. The scaling in this case is Mej ≈ 12.0 × Eexp M which. 32.

(34) CHAPTER 3. ANALYSIS. implies an “agreement” pair (Mej , Eexp ) at (6 M ,0.25 foe). Anderson et al. 2014a discusses the observation of higher deviation from the expected full trapping decay for the brighest type II SNe in their sample (in the V band) and they suggest that this would be explained by the expected more diluted (i.e. less massive ejecta or more extended) hydrogen associated with faster declining SNe. In Figure 4.16 the (Mej , Eexp ) curves estimated in our analysis are presented, together with the curve with semi-analitycal models obtained in Pejcha and Prieto 2015b. In our case the ejecta mass is slightly below what would be needed for a relatively low energy explosion of ≈ 0.25 foe to have a high gamma ray opacity (τγ = (T0 /t)2 ≥ 1) at ≈ 400 days. We also used the correlation between the Hα FWHM and the 56 Ni mass [Maguire et al., 2012] as an independent probe, and with the value of FWHM ' 35 Å we obtain a. 56. Ni. mass of 0.0118 ± 0.0012 M , which is lower but consistent with the estimate obtained from the late-time bolometric light curve. The low nickel mass agrees roughly with the explosion parameters according to the classic explosion models for type II SNe [Popov, 1993, Litvinova and Nadezhin, 1985].. 3.3.3. Nebular Spectra and the Progenitor Mass. After the optically thick phase we observe the inner layers of the exploding star which show the elements synthesized in the explosion. The emission due to the radioactive and non-thermal heating of electrons provides information on the elements physical conditions, together with their spatial distribution and possible asymmetries [Maguire et al., 2012, Jerkstrand, 2017]. In the nebular phase typical spectra of Type II-P like SNe show prominent Hα and O I lines, Ca II and Fe II (including forbidden) emission lines. The [O I] λλ6300, 6364 has been used to estimate the Oxygen core mass and therefore the main sequence mass of the star [e.g. Jerkstrand et al., 2012, Elmhamdi et al., 2003, Fransson and Chevalier, 1989, Uomoto, 1986]. We compare our spectra with available models for nebular optical spectra from 33.

(35) CHAPTER 3. ANALYSIS. Dessart et al. [2013], Jerkstrand et al. [2014] and Jerkstrand et al. [2018]. Dessart et al. [2013] (henceforth D13) presented models for Type II SNe radiation using a grid of stellar models from MESA STAR of a 15 M progenitor star with different parameters controlling the evolution such as the mixing length, metallicity, and rotation. These models were artificially exploded with a piston to produce different initial explosion energies. After this, the radiative transfer problem was solved with CMFGEN to generate synthetic spectra at different epochs and light curves. In this comparison we choose the nebular spectra in their “m15” model set that has epochs which are up to 10 days to 400 days past explosion, which are mlt1, mlt3, z2m2 and z8m3 (see Table 1 in D13 for details). Jerkstrand et al. [2014] (now J14) calculated models for 12, 15, 19 and 25 M stars evolved using the KEPLER code, assuming Solar metallicity and no rotation. The models were exploded with a fixed energy of 1.2 foe using an artificial piston. We also included the recent model of 9 M from Jerkstrand et al. [2018] (hereby J18). In Figure 4.18 and 4.19 the ASASSN-14jb nebular spectrum together with the models, at the [O I] λλ6300, 6364 and Hα wavelength regions, are presented. In the case of D13 models we see that the models z2m2 and z8m3 are the best fit, which correspond to a progenitor metallicity of Z = 0.02, 0.008, oxygen mass of MO = 0.325, 0.507 M and M (56 N i) = 0.05, 0.081 M , respectively (for details see Table 1 in D13). In the case of the J14 and J18 models we see the preference between the 9 M and 12 M models, considering that the models in general over-predicts the Hα emission. We must add that this comparison is not trivial given that all models assume higher. 56. Ni masses. than measured for ASASSN-14jb. This is easily realized when looking at the width of the lines in both models, which is affected by the. 56. Ni mass and explosion energy. To. take this into account, we normalized the spectra to the peak of each of the Hα lines. After doing this, in the J14 models we obtain better agreement with the 15 M. model. but more closely to the 9 M . In the D13 models we see a high degeneracy given the differences in the details of stellar evolution. 34.

(36) CHAPTER 3. ANALYSIS Other approach we can use to alleviate the 56 Ni mass discrepancy between ASASSN14jb and the models, is to obtain the flux in a given line and normalize to the 56 Ni decay power. Doing this, we obtain the surprising result that the ratio of ≈ 0.04, using the [O I] λλ6300, 6364, is consistent with the higher mass range 15 − 19 M according to J14 models but, again, the low mass model of 9 M. is closer. As explained in the Appendix. H of Jerkstrand et al. [2018] this is explained as the oxygen shell in the 9 M. model. lies closer to the 56Ni yields and also significant Fe I contribution to the [O I] λ6364 line is attributed. As a final estimate we applied the analytical formula used in J14 to give an upper limit to the emitting oxygen mass, L[O I] /β[O I] max(MO I ) = × exp 9.7 × 1041 where L[O. I]. . 22720 T. . is the total luminosity of the λλ6300, 6364 doublet in erg/s, β[O. (3.10) I]. is the. Sobolev escape probability of the same transition, and T is the equilibrium temperature in K. The temperature was estimated using the ratio of the [O I] λλ6300, 6364 forbidden doublet and the [O I] λ5577 forbidden line, as in J14, obtaining values in the range ≈ 3900−4300 2 . All the fluxes were estimated fitting a skewed Gaussian to the line profiles. The resulting values for the maximum oxygen mass are in the range ≈ 0.09 − 0.18 M . This is above the 9 M. of J18 by 20 − 40% and a factor of 1.5 − 3 lower that the. 14 M model in J14. Considering the nucleosynthesis yields of the models W18 and N20 presented in Sukhbold et al. [2016] for a 12.5 M. progenitor, the oxygen yields. ≈ 0.35 M help push ASASSN-14jb to a less massive progenitor than this models. of. predict. We conclude that the nebular spectra confidently points to a low mass progenitor, in the range 10 − 12 M . 2. Assuming. β6300,6364 β5577. = 0.5 − 1 as in J14.. 35.

(37) CHAPTER 3. ANALYSIS. Warm Dust Detection in Spitzer Imaging In order to constrain dust formation at late times, as has been detected in other normal Type II SNe [e.g., Szalai et al., 2018], we searched the Spitzer data archive3 for mid-infrared imaging data of ASASSN-14jb. The field of ASASSN-14jb and its host galaxy have been observed 9 times since 2010 with the Spitzer IRAC instrument [Fazio et al., 2004] in the 3.6 µm and 4.5 µm bands. Two of these epochs are observations obtained after the supernova went off, on 2015-09-13 (program ID 11053), ≈ 330 days past explosion, and on 2016-08-17 (program ID 12099), ≈ 670 days past explosion. We retrieved these images and also images from 2014-09-05 (program ID 10139) to be used as templates for image subtraction. We used the HOTPANTS [Becker, 2015] package for difference imaging analysis on the Spitzer IRAC images and run aperture photometry on the difference images. ASASSN-14jb is clearly detected in the 2015-09-13 images with m3.6 = 17.85 ± 0.03 mag and m4.5 = 16.38 ± 0.03 mag (magnitudes in the AB system), but undetected on 2016-08-17. The Spitzer mid-infrared absolute magnitudes (M3.6 ' −14.1, M4.5 ∼ −15.6) and red color (m3.6 − m4.5 ∼ 1.5 mag) of ASASSN-14jb at ≈ 333 days after explosion are consistent with warm dust formation in other Type II SNe [e.g., Prieto et al., 2012, Szalai et al., 2018]. We used a black-body function with dust emissivity Qλ ∝ λ−1 to fit the mid-infrared SED and obtained the following best-fit parameters: L ∼ 1.1 × 106 L , T ∼ 444 K, and R ∼ 839 AU. Using Equation 1 in Prieto et al. [2009] (from Dwek et al. 1983), we obtain a total dust mass of Md ∼ 10−4 M . These estimates are consistent with models of newly formed dust in the SN ejecta and observations for other Type II SNe at a similar epoch after explosion [e.g. Sarangi and Cherchneff, 2015]. 3. http://sha.ipac.caltech.edu/applications/Spitzer/SHA/. 36.

(38) CHAPTER 3. ANALYSIS. 3.3.4. Progenitor Metallicity from the Photospheric Spectral Lines. We calculate the pseudo-equivalent width (pEW) of the Fe II λ5018 line and compare with the explosion models from Dessart et al. [2013]. These models cover four progenitor metallicities (2 Z , 1 Z , 0.4 Z and 0.1 Z ) and a couple more solar compositions with different mixing length prescriptions, which alters the progenitor radii. More recently, Anderson et al. [2016] compared the measured pEWs with H II region metallicities derived from oxygen abundances. In Figure 4.20 we present the measured values for ASASSN-14jb together with the expected evolution of the models of D14. We see that the estimated pEWs of the Fe II line are more consistent with the curve for 0.4 Z , but it should be clear that changing the progenitor radius can also give a slightly subsolar value. If we linearly interpolate the models and, assuming the dispersion given by different radius at solar metallicity as a proxy for the error, the metallicity obtained for ASASSN14jb is Z = 0.3 ± 0.1Z . A more complete set of models is needed to give a more definitive estimate from the photospheric spectral line diagnostics.. 3.4. The Host Galaxy of ASASSN-14jb with MUSE. Through the analysis of the vicinity or unresolved stellar neighborhood of the SNe explosions, it is possible to statistically constrain the progenitor properties, such as the mass, from the age of the nearby stellar population and metallicity, from several line diagnostics, of the different types of exploding stars [e.g. Stoll et al., 2013, Anderson et al., 2015, Galbany et al., 2016c]. Recent observations with integral field spectrographs (IFS) has helped push new limits on the studies of SNe environments [Galbany et al., 2016a, 2018]. H II regions are low density regions of gas ionized by massive stars [Osterbrock,. 37.

(39) CHAPTER 3. ANALYSIS. 1989]. The massive OB stars in these regions, from which the majority of ionizing photons come from, have lifetimes of a few to a few tens of Myr [e.g. Pols et al., 1998]. When the most massive stars explode as supernovae the ionizing flux decays rapidly and the H II region fades fast. Because of this, the strength of recombination lines such as Hα trace very confidently the ongoing star formation. As giant H II regions have typical physical sizes of about a hundred to a few hundreds parsecs [Sánchez et al., 2012], the MUSE IFS spatial resolution allows us to resolve individual regions in nearby host galaxies. In our MUSE observations of ASASSN-14jb the seeing was 1.00 08 in the R-band, which gives a resolution of ∼ 130 pc2 at the distance of the SN. In the following analysis, we make use of the 2D emission line maps derived from the MUSE datacube, following the procedures explained in detail in Galbany et al. 2016a. In short, we fit the spectrum of each spaxel using a combination of the stellar continuum, modeled using the STARLIGHT stellar population synthesis code [Cid Fernandes et al., 2009], and Gaussian profiles at the central wavelengths of the main nebular emission lines. The emission line fluxes with > 3σ detections are extinction corrected using the Galactic reddening map [Schlafly and Finkbeiner, 2011] and an estimation of the internal extinction of the gas using the Balmer decrement. We also use the HII-Explorer code [Sánchez et al., 2012] to detect and analyze H II regions in the datacube.. 3.4.1. Overall Properties of ESO 467-G051. The host galaxy of ASASSN-14jb, ESO 467-G051, is an edge-on, late-type disk galaxy [morphological classification Scd in the RC3 catalog, de Vaucouleurs et al., 1991]. We measured the total BV gri magnitudes of the galaxy from coadds of the best-seeing LCOGT images using a large elliptical aperture in ds9. We retrieved a near-infrared Ks -band image of the field from the ESO data archive obtained by PESSTO (program ID 191.D-0935) with SOFI on the ESO/NTT telescope and measured the magnitude of the host using an elliptical aperture calibrated to 2MASS [Cutri et al., 2003]. 38.

(40) CHAPTER 3. ANALYSIS. We also obtained archival ultraviolet magnitudes from the Galaxy Evolution Explorer (GALEX) archive Morrissey et al. [2007] and Spitzer mid-infrared magnitudes from the S4G survey catalog [Sheth et al., 2010]. The integrated apparent magnitudes are presented in Table 4.7. The absolute magnitudes of ESO 467-G051 are MB ≈ −17.6 mag, MV ≈ −18.0 mag, and MKs ≈ −19.1 mag, after correcting for Galactic extinction. Figure 4.21 shows a histogram with the distribution of host galaxy MKs magnitudes of Type II SNe discovered by ASAS-SN in 2013-2017 from Holoien et al. [2017c,a,b] and Holoien et al. (2018, in prep). The host galaxy of ASASSN-14jb is in the lower Ks absolute magnitude (lower stellar mass) range of the Type II SNe discovered by ASAS-SN. Using the F U V , N U V , BV gri, Ks , 3.6 µm, and 4.5 µm total magnitudes for the host galaxy, we fit stellar population synthesis (SPS) models using the Fitting and Assessment of Synthetic Templates [FAST; Kriek et al., 2009] code. For the SPS fits with FAST, we assume a Galactic extinction law, a Salpeter stellar initial mass function, an exponentially declining star formation law (SFR ∝ e−t/τ , with τ = 1 Gyr), and the Bruzual and Charlot [2003] SPS models. We obtain a best-fit SPS model with the following parameters (with 1σ error bars): M? ' 1×109 M , age ' 3.2 Gyr, and SFR ' 0.07 M /yr. We also integrated the Hα extinction-corrected fluxes (see below) obtained from the MUSE datacube of the host galaxy to get an estimated SFR(Hα ) ∼ 0.07 M /yr [Kennicutt, 1998], which is fully consistent with the host galaxy SED SPS fit. Neutral hydrogen (HI 21 cm) observations suggest that this galaxy is undergoing a direct encounter with ESO 467-G050 or NGC 7259 (∆V = 92.6 ± 5.6km/s. Nordgren et al. 1997), the host galaxy of SN 2009ip. From these observations the total HI flux obtained is SHI = 23.1 ± 0.8 Jy km/s and the derived HI mass reported is MHI = 3.1 ± 0.8 [109 M ] 4 , which, using our best distance estimate from Section 3.2 of 25 ± 1 Mpc, the HI mass scales to MHI = 3.5 ± 0.9 [109 M ]. This gives a gas fraction of 4. MHI = 2.356 × 105 × D2. R. Sν dν M. , D = 23.5 Mpc. 39.

(41) CHAPTER 3. ANALYSIS Mgas /(Mgas + M∗ ) ≈ 0.77. From the Hyperleda database5 the maximal HI rotational velocity is 60.2 ± 1 km/s. In our Hα velocity map we observe this velocity limit more clearly on the East side. Also from the derived velocity dispersion map we observe a mean value of 50 km/s, which seems high given the low SFR but consistent with a high gas fraction [Green et al., 2014].. 3.4.2. Host H II Regions Properties. Using the Hα map derived from the MUSE datacube, the H II regions were selected with the HII-Explorer package as in Sánchez et al. [2012]. With the H II region catalog, we get the integrated spectra for each of them and we can obtain the nebular emission line intensities. We calculated the O3 = log([O III]λ5007/Hβ), N 2 = log([N II]λ6584/Hα) and S2 = log([S II]λ6584/Hα) line ratios and with these we plot all the regions in the Baldwin-Phillips-Terlevich (BPT) diagrams [Baldwin et al., 1981], presented in Figures 4.24 and 4.25. Each H II region in the diagram is colored by the projected distance to the ASASSN-14jb explosion site and the size of the circle marking the position of each region is a decreasing function of this distance to make the visualization easier. The majority of the H II regions lie in a narrow range for N2 values, between −1.5 and −0.8. For the O3 values the regions cover from 0.8 to −0.8, while for S2 the regions are in the interval from −1.0 to −0.3. All the regions are consistent with star-forming regions in the BPT diagram using the definitions in Kewley et al. [2001]. The low values of the N2 ratio are evidence for the relatively low oxygen abundance (metallicity) of the regions, which agrees with the mean of all the used gas-phase oxygen abundance diagnostics as we show next. 5. http://leda.univ-lyon1.fr/. 40.

(42) CHAPTER 3. ANALYSIS. 3.4.3. Hα and Oxygen Abundance Maps. In Figure 4.22 we present a map of the Hα emission line fluxes (dereddened) derived from the MUSE datacube, which includes H II regions, diffuse emission in the host, and the nebular SN emission at the position of ASASSN-14jb. We mark the positions of the detected H II regions with the HII-Explorer code. The nearest H II region to the SN explosion site is at a distance of ≈ 1.4 kpc of the SN position and at a vertical distance of ≈ 2.1 kpc from the disk major axis. We measure a gas-phase oxygen abundance from the O3N2 and N2 strong line methods [Marino et al., 2013] for the H II regions obtaining a median, with 99% confidence +0.10 intervals, of 12 + log(O/H) = 8.27+0.16 −0.20 and 8.19−0.24 , respectively. If we use the new. calibration by Dopita et al. 2016, hereafter D16, which uses [S II], [N II], and Hα lines, we obtain lower values with a median of 12 + log(O/H) = 7.77+0.10 −0.24 . The O3N2 and D16 maps are presented in Figure 4.23. In Figure 4.26 we show the distributions of the oxygen abundances of the H II regions with the N2, O3N2 and D16 nebular emission line diagnostics. All the mean values are sub-Solar (using as reference 12 + log(O/H). = 8.69, from Asplund et al. 2009),. and significantly below the average of the the PMAS/PPAK Integral-Field Supernova Host Compilation (PISCO) sample mean for Type II SNe nearby H II regions, of 12 + log(O/H) = 8.54 [Galbany et al., 2018]. Our host galaxy B-band absolute magnitude, of MB = −17.6 mag, compares well with the mean B band magnitude of MBhost = −17.66 mag in the recent sample of Type II SNe in low-metallicity, dwarf galaxies [Gutiérrez et al., 2018]. Also, as showed in Gutiérrez et al. [2018], lower luminosity galaxies hosts lower metallicity Type II SNe, filling the space between the 0.1 and 0.4 Z. metallicity models. This is expected given the strong correlation between stellar. mass (or galaxy luminosity) and metallicity. It is worthwhile to mention that the O3N2 map shows lower oxygen abundances in the cores of the H II regions. This could be a consequence of the unaccounted internal gradient of the ionization parameter of each H 41.