Thermonuclear explosions of rapidly differentially rotating white dwarfs: candidates for superluminous Type Ia supernovae?

The observed sub-class of “superluminous” Type Ia supernovae lacks a convincing theoretical explanation. If the emission of such objects were powered exclusively by radioactive decay of 56Ni formed in the explosion, a progenitor mass close to or even above the Chandrasekhar limit for a non-rotating white dwarf star would be required. Masses significantly exceeding this limit can be supported by differential rotation. In an upcoming paper in Astronomy & Astrophysics, we explore explosions and predict observables for various scenarios resulting from differentially rotating carbon-oxygen white dwarfs close to their respective limit of stability. Specifically, we have investigated a prompt detonation model, detonations following an initial deflagration phase (“delayed detonation” models), and a pure deflagration model.

Snapshot of a thermonuclear explosion of a rapidly rotating WD as a delayed detonation. Deflagration ashes are shown in pink, detonation ashes in light blue. The dark blue surface indicates the outer boundary of the star.

In postprocessing steps, we performed nucleosynthesis and three-dimensional radiative transfer calculations, that allow us, for the first time, to consistently derive synthetic observables from our models. We find that all explosion scenarios involving detonations produce very bright events. The observables predicted for them, however, are inconsistent with any known subclass of Type Ia supernovae. Pure deflagrations resemble 2002cx-like supernovae and may contribute to this class. We discuss implications of our findings for the explosion mechanism and for the existence of differentially rotating white dwarfs as supernova progenitors.

Paper accepted for publication in Astronomy & Astrophysics, full text available at

OGLE14-073 – a promising pair-instability supernova candidate

The recently discovered bright type II supernova OGLE14-073 evolved very slowly. The light curve rose to maximum for 90 days from discovery and then declined at a rate compatible with the radioactive decay of 56Co. In a paper led by A. Kozyreva (Tel Aviv University), we show that a pair-instability supernova is a plausible mechanism for this event. We calculate explosion models and light curves with the radiation hydrodynamics code STELLA starting from two progenitor stars with an initial mass of 150 solar masses and low metallicity (Z=0.001). We obtain satisfactory fits to OGLE14-073 broadband light curves by including additional 56Ni in the centre of the models and mixing hydrogen down into the inner layers of the ejecta to a radial mass coordinate of 10 solar masses. The extra 56Ni required points to a slightly more massive progenitor star. The mixing of hydrogen could be due to large-scale mixing during the explosion. We also present synthetic spectra for our models simulated with the Monte Carlo radiative transfer code ARTIS. The synthetic spectra reproduce the main features of the observed spectra of OGLE14-073. We conclude that OGLE14-073 is one of the most promising candidates for a pair-instability explosion.

Synthetic spectrum for one of the models of Kozyreva et al. 2018 (P150-bf, shown in red) 7 days before peak brightness. For comparison the observed spectrum of OGLE14-073 is shown in black. The labels indicate the species that cause the most prominent spectral features.

Paper published in Monthly Notices of the Royal Astronomical Society, Volume 479, Issue 3, Pages 3106–3114, full text available at

660 WE-Heraeus seminar “Supernovae – From Simulations to Observations and Nucleosynthetic Fingerprints”

Thanks to new transient surveys and advances in high-performance computing our knowledge on supernova explosion has tremendously increased in the past decade. To consolidate progress, it is important to make specialists in different aspects of supernova research aware of progress and challenges on other fronts. Together with Anders Jerkstrand (Max Planck Institute for Astrophysics Garching) and Bernhard Mueller (Monash University), I have co-organized the 660. Wilhelm und Else Heraeus Seminar “Supernovae – From Simulations to Observations and Nucleosynthetic Fingerprints”, which was held at Physikzentrum Bad Honnef from January 21 to 24, 2018. The seminar, generously funded by the Wilhelm und Else Heraeus Stiftung, brought together world-leading experts, early career researchers, and students. With over 60 participants from all over the world we discussed the physical modelling of stellar explosions, their nucleosynthesis, radiation transfer, and supernova observations to identify future opportunities at the intersection of these four different research fields.

Early light curves for Type Ia supernova explosion models

Upcoming high-cadence transient survey programmes will produce a wealth of observational data for Type Ia supernovae. These data sets will contain numerous events detected very early in their evolution, shortly after explosion. Together with a team led by Dr. Ulrich Nöbauer from the Max Planck Institute for Astrophysics,  we used numerical simulations to calculate synthetic observables for a number of different explosion models, specifically focusing on the first few days after explosion.

Synthetic light curves for various supernova explosion models in the Bessell U (upper left), B (upper right), V (lower left) and R (lower right) wavelength bands during the first 10 days after explosion. The inset shows the same curves on a logarithmic timescale. Credit: U. Nöbauer/MPA

Our results show that overall the early light curve evolution is similar for most of the investigated models. Characteristic imprints are induced by radioactive material located close to the surface. However, these  are very similar to the signatures expected from interactions between the SN ejecta and circumstellar material or a companion star. Apart from pure deflagration explosion models, none of the synthetic light curves exhibit the commonly assumed power-law rise. This can lead to substantial errors in the determination of the time of explosion. In summary, we illustrate with our calculations that even with very early data an identification of specific explosion scenarios is challenging, if only photometric observations are available. For more information see the research highlight at the Max Planck Institute for Astrophysics.

Paper published in Monthly Notices of the Royal Astronomical Society, Volume 472, Issue 3, Pages 2787–2799, full text available at

Kosmischen Leuchtfeuern auf der Spur

Am 24. Mai 2017 werde ich beim Arbeitskreis Astronomie im Studium Generale der Universität Stuttgart einen öffentlichen Abendvortrag über Typ Ia Supernovae halten. Der Vortrag beginnt um 19:30 im Hörsaal V 31.01 auf dem Campus Vaihingen.

Kurzzusammenfassung: In den Endphasen ihrer Entwicklung durchlaufen Sterne dramatische Prozesse die zu gewaltigen Explosionen führen. Diese sogenannten Supernovae spielen eine wichtige Rolle für die chemische Entwicklung des Universums und lassen sich auf Grund ihrer immensen Helligkeit auch noch in entfernten Galaxien beobachten. Eine bestimmte Klasse von Supernovae hat in den letzten zwei Jahrzehnten besondere Aufmerksamkeit erfahren. Eine empirische Relation zwischen der Helligkeit und zeitlichen Entwicklung dieser Typ Ia Supernovae macht sie zu kosmischen Leuchtfeuern, mit denen sich grosse Entfernungen im Universum vermessen lassen. Ende der 1990er Jahre hat die Anwendung dieses Verfahrens zu einer Revolution in unserem Verständnis vom Universum geführt. Bei der Vermessung weit entfernter Typ Ia Supernovae stellte sich heraus, dass diese dunkler waren als im Rahmen des damaligen kosmologischen Standardmodells erwartet wurde. Dies lässt sich nur dadurch erklären, dass unser Universum beschleunigt expandiert. Trotz dieser enormen Bedeutung für die moderne Astrophysik ist der Explosionsmechanismus von Typ Ia Supernovae und die Frage nach ihren Vorläufersternen immer noch nicht vollständig geklärt. In diesem Vortrag werde ich aktuelle Beobachtungsergebnisse und neue Computersimulationen zu Typ Ia Supernovae vorstellen, die dazu beitragen unser Verständnis dieser kosmischen Leuchtfeuer zu verbessern.


Invited review talk at conference on the AGB-Supernovae mass transition

On March 27-31 an international astrophysics conference on the AGB-Supernovae mass transition was held at the Observatory of Rome in Frascati, Italy. About 100 researchers from all over the world discussed the evolution of stars that start their life with a mass of about 7 to 11 solar masses and reach the super-asymptotic giant branch (AGB) phase during their later evolution. At least three different types of supernovae are connected with the limits of this mass range: isolated stars with an initial mass larger than 11 solar masses end their lives as core collapse supernovae (Type II SN). An initial stellar mass of 7 solar masses is the upper limit for the formation of carbon-oxygen white dwarf stars, whose subsequent evolution through mass transfer in a binary system may lead to thermonuclear supernova explosions (Type Ia SN). Finally, single stars in the super-AGB upper mass range may be the site of electron-capture supernovae. I was invited to provide a review about recent progress in the modelling of Type Ia supernovae on this conference.

Predicting polarization signatures for double-detonation and delayed-detonation models of Type Ia supernovae

Calculations of synthetic spectropolarimetry are one means to test multi-dimensional explosion models for Type Ia supernovae. In a recent paper, we demonstrated that the violent merger of a 1.1 and 0.9 M⊙ white dwarf binary system is too asymmetric to explain the low polarization levels commonly observed in normal Type Ia supernovae. In this paper, we present polarization simulations for two alternative scenarios: the sub-Chandrasekhar mass double-detonation and the Chandrasekhar mass delayed- detonation model. Specifically, we study a two-dimensional double-detonation model and a three-dimensional delayed-detonation model, and calculate polarization spectra for multiple observer orientations in both cases. We find modest polarization levels (< 1 per cent) for both explosion models. Polarization in the continuum peaks at ∼ 0.1−0.3 per cent and decreases after maximum light, in excellent agreement with spectropolarimetric data of normal Type Ia supernovae. Higher degrees of polarization are found across individual spectral lines.

In particular, the synthetic Si II λ6355 profiles are polarized at levels that match remarkably well the values observed in normal Type Ia supernovae, while the low degrees of polarization predicted across the O I λ7774 region are consistent with the non-detection of this feature in current data. We conclude that our models can reproduce many of the characteristics of both flux and polarization spectra for well-studied Type Ia supernovae, such as SN 2001el and SN 2012fr. However, the two models considered here cannot account for the unusually high level of polarization observed in extreme cases such as SN 2004dt.

Paper published in Monthly Notices of the Royal Astronomical Society, Volume 462, Issue 1, p.1039-1056,  full text available at

Relocation to Heidelberg

After three exciting and stimulating years as an Oskar Klein Fellow at Stockholm University’s Oskar Klein Centre, I am moving to Heidelberg on October 1st to take up a new position as Gliese Fellow at the Zentrum für Astronomie der Universität Heidelberg (ZAH). In close collaboration with scientists from the PSO group at the Heidelberg Institute for Theoretical Studies (HITS) I will use theoretical models to shed light on the progenitors and physical mechanisms of stellar explosions.

Three-dimensional simulations of gravitationally confined detonations compared to observations of SN 1991T

Ejecta structure of the GCD model.

Ejecta structure of our GCD model at 60s past explosion. Burning products from the off-centre ignited detonation (indicated by the blue surface) are engulfed by deflagration ashes (shown in red and rendered semi-transparently to allow to see the detonation ashes).

The gravitationally confined detonation (GCD) model has been proposed as a possible explosion mechanism for Type Ia supernovae (SNe) in the single-degenerate evolution channel. Driven by buoyancy, a deflagration flame rises in a narrow cone towards the surface of the exploding white dwarf star. For the most part, the flow of the expanding ashes remains radial, but upon reaching the outer, low-pressure layers of the white dwarf, an additional lateral component develops. This makes the deflagration ashes converge again at the opposite side, where the compression heats fuel and a detonation may be launched. To test the GCD model, we perform a 3D explosion simulation for a model with an ignition spot offset near the upper limit of what is still justifiable, 200 km. This simulation meets our deliberately optimistic detonation criteria and we initiate a detonation. The detonation burns through the white dwarf and leads to its complete disruption. We determine nucleosynthetic yields by post-processing a million tracer particles with a 384 nuclide reaction network and we present multi-band light curves and time-dependent optical spectra. We find that our synthetic observables show a prominent viewing-angle sensitivity in UV and blue bands, which is in tension with observed SNe Ia.


Synthetic spectra of our GCD model at around maximum light. The asymmetry of the ejecta imposes a strong viewing-angle sensitivity at at short wavelengths.

The strong dependence on viewing-angle is caused by the asymmetric distribution of the deflagration ashes in the outer ejecta layers. Finally, we perform a comparison of our model to SN 1991T. The overall flux-level of the model is slightly too low and the model predicts pre-maximum light spectral features due to Ca, S, and Si that are too strong compared SN 1991T. We therefore find it unlikely that the GCD model can explain SN 1991T or similar SNe.

Paper accepted for publication in Astronomy and Astrophysics, full text available at

The peculiar Type Ia supernova iPTF14atg: Chandrasekhar-mass explosion or violent merger?

Synthetic spectral time series and iPTF14atg

Synthetic spectral time series of our merger model (dark red line) and the observed spectral evolution of iPTF14atg (black line, epochs denote the time since explosion). The light red area indicates the spread due to different viewing angles from the asymmetric merger ejecta.

iPTF14atg, a subluminous peculiar Type Ia supernova (SN Ia) similar to SN 2002es, is the first SN Ia for which a strong UV flash was observed in the early-time light curves. This has been interpreted as evidence for a single-degenerate (SD) progenitor system where such a signal is expected from interactions between the SN ejecta and the non-degenerate companion star. In this paper, we compare synthetic observables of multi-dimensional state-of-the-art explosion models for different progenitor scenarios to the light curves and spectra of iPTF14atg. From our models, we have difficulties explaining the spectral evolution of iPTF14atg within the SD progenitor channel. In contrast, we find that a violent merger of two carbon-oxygen white dwarfs with 0.9 and 0.76 solar masses, respectively, provides an excellent match to the spectral evolution of iPTF14atg from 10d before to several weeks after maximum light. Our merger model does not naturally explain the initial UV flash of iPTF14atg. We discuss several possibilities like interactions of the SN ejecta with the circus-stellar medium and surface radioactivity from a He ignited merger that may be able to account for the early UV emission in violent merger models.

Paper published in Monthly Notices of the Royal Astronomical Society, full text available at

Observing for PESSTO at La Silla Observatory

Although a theorist by training, I am also involved in observational projects. During my postdoctoral fellowship at MPA, I was part of the Nearby Supernova Factory and performed remote observations with the UH88 telescope on Mauna Kea at Hawaii. Currently, I am involved in the Public ESO Spectroscopic Survey for Transient Objects (PESSTO). PESSTO is granted 90 nights per year on ESO‘s New Technology Telescope (NTT) at La Silla Observatory in the Chilean Andes to follow-up astronomical transients.


Credit: ESO/H. Dahle

Together with two colleagues from ESO and MPA, tonight we are starting a three week observation campaign for PESSTO at La Silla. We will use the EFOSC2 and SOFI spectrographs to obtain optical and near-infrared spectra of recent supernovae that are investigated by the collaboration. In addition, we will also classify new transient candidates, which we obtain from wide-field transient surveys like e.g. the La Silla QUEST survey or photometric alerts from ESA’s Gaia satellite.

Type Ia supernovae from violent mergers of carbon–oxygen white dwarfs: polarization signatures

The violent merger of two carbon–oxygen white dwarfs has been proposed as a viable progenitor for some Type Ia supernovae. However, it has been argued that the strong ejecta asymmetries produced by this model might be inconsistent with the low degree of polarization typically observed in Type Ia supernova explosions. In this paper, we test this claim by carrying out a spectropolarimetric analysis for the model proposed by Pakmor et al. for an explosion triggered during the merger of a 1.1 and 0.9 M carbon–oxygen white dwarf binary system. Owing to the asymmetries of the ejecta, the polarization signal varies significantly with viewing angle. We find that polarization levels for observers in the equatorial plane are modest (≲1 per cent) and show clear evidence for a dominant axis, as a consequence of the ejecta symmetry about the orbital plane. In contrast, orientations out of the plane are associated with higher degrees of polarization and departures from a dominant axis.

merger_polWhile the particular model studied here gives a good match to highly polarized events such as SN 2004dt, it has difficulties in reproducing the low polarization levels commonly observed in normal Type Ia supernovae. Specifically, we find that significant asymmetries in the element distribution result in a wealth of strong polarization features that are not observed in the majority of currently available spectropolarimetric data of Type Ia supernovae.

Paper published in Monthly Notices of the Royal Astronomical Society, full text available at

Special session on SNe Ia at EWASS 2015

EWASS2015_logoOn June 25 we hosted a special session “Hunting down the elusive progenitors and explosion mechanisms of Type Ia supernovae” as part of the annual European Week of Astronomy and Space Science which took place on Tenerife (Spain) from June 22-26 .

Our special session aimed at bringing together theorists and observers to give an overview about the current observational state of the art, present recent theoretical progress and discuss the open problems in SNe Ia. The programme consisted of a total of 3 invited talks, 15 contributed talks and 9 posters with participants coming from Australia, Chile, Denmark, Germany, Japan, Netherlands, Russia, Slovakia, Spain, Sweden, Turkey and the United Kingdom. With about 50 scientists attending, the session was quite a success.

The programme was divided in three blocks covering multi-wavelength observations (invited speaker: Assaf Sternberg), binary population synthesis, rates and progenitor searches (invited speaker: Silvia Toonen), and theoretical explosion models and their comparison to observational data (invited speaker: Pablo Loren Aguilar). Highlights of the meeting were presentations on the first gamma-ray detection of a SN Ia, radio constraints on the progenitor of SN 2014J, the discovery of a potential SN Ia progenitor system in a planetary nebula and first synthetic spectropolarimetry for 3D merger models.

Deflagrations in hybrid CONe white dwarfs: a route to explain the faint Type Iax supernova 2008ha

Stellar evolution models predict the existence of hybrid white dwarfs (WDs) with a carbon-oxygen core surrounded by an oxygen-neon mantle. Being born with masses ~1.1 Msun, hybrid WDs in a binary system may easily approach the Chandrasekhar mass (MCh) by accretion and give rise to a thermonuclear explosion. In this paper, we investigate an off-centre deflagration in a near-MCh hybrid WD under the assumption that nuclear burning only occurs in carbon-rich material. Performing hydrodynamics simulations of the explosion and detailed nucleosynthesis post-processing calculations, we find that only 0.014 Msun of material is ejected while the remainder of the mass stays bound. The ejecta consist predominantly of iron-group elements, O, C, Si and S.


We also calculate synthetic observables for our model and find reasonable agreement with the faint Type Iax SN 2008ha. This shows for the first time that deflagrations in near-MCh WDs can in principle explain the observed diversity of Type Iax supernovae. Leaving behind a near-MCh bound remnant opens the possibility for recurrent explosions or a subsequent accretion-induced collapse in faint Type Iax SNe, if further accretion episodes occur.

Paper published in Monthly Notices of the Royal Astronomical Society, Volume 450, Issue 3, p.3045-3053, full text available at