PI: Fionntán Callan
Fionntán Callan, Stuart Sim, Luke Shingles, Christine Collins, Joshua Pollin
Astrophysics Research Centre, Queen’s University Belfast, Northern Ireland
Type Ia supernovae are the extremely luminous (approximately 5 billion solar luminosities), explosive end points of low mass stars. They are a key source of heavy elements (producing >50% of the Universe’s iron) and around 70% of them (so called “normal” Type Ia supernovae) can be used to measure cosmic distances, leading to the discovery of dark energy (2011 Nobel Prize). While known to arise from thermonuclear explosions of white dwarf stars in close binaries, the evolutionary channels and explosion mechanisms that produce Type Ia supernovae remain a mystery. Distinguishing between the many competing explosion models for Type Ia supernovae is key to our understanding of these cosmic explosions. A leading theoretical model for Type Ia supernovae is the “double detonation” scenario in which the explosion of a white dwarf is triggered by the ignition of a detonation in a surface layer of helium (He-shell). A defining property predicted by hydrodynamic explosion simulations of these models is the presence of burning products of the He-shell, such as calcium and titanium, in the outer explosion ejecta. Previous simulations predict spectral imprints of these burning products are present for the first few weeks after explosion. Accurately modelling these signatures of the He-shell burning products so they can be compared to Type Ia supernova observations requires radiative transfer simulations that capture both the multidimensional ejecta structure of the explosion (Sim et al. 2013) and the detailed non local thermodynamic equilibrium (NLTE) physics that shapes the spectrum of light emitted by the supernova (Dessart et al. 2014).
We have used DiRAC to carry out simulations of 2D double detonation explosion models from the sequence of Gronow et al. (2021) utilising our Monte Carlo radiative transfer code artis (https://github.com/artis-mcrt/artis). Originally artis adopted an approximate NLTE treatment for the plasma conditions. While intended to capture some NLTE effects this treatment is still based on the assumptions of local thermodynamic equilibrium. Shingles et al. (2020) carried out substantial developments to artis adding a full NLTE solution of the plasma state. In this project, we carried out the first early phase simulation that utilised this detailed NLTE treatment while also accounting for the multidimensional structure of the explosion ejecta. Comparing to a 1D NLTE artis simulation and a 2D artis simulation utilsing our approximate treatment of NLTE physics for the same explosion model we find substantial differences between the predicted spectra (see Fig. 1). A particularly striking difference appears in the ionisation states of calcium predicted by the different simulations, especially in the high-velocity calcium produced in the He-shell detonation which is a key signature of this scenario (see again Fig. 1). This work demonstrates the importance of radiative transfer simulations that account for the multidimensional ejecta structure of realistic explosion models and also include a detailed treatment of NLTE physics for accurately predicting the observable signatures of different Type Ia supernovae explosion models and for robustly accessing their viability through comparisons with observations.

Fig. 1: Top panel: Ion density of singly ionised calcium (Ca ii) against velocity at 30 days post explosion for our three double detonation radiative transfer simulations. The 1D model is based on the equatorial direction of the 2D model. The 2D NLTE simulation predicts a substantially different calcium ionisation state to both other simulations, in particular showing more Ca ii at higher velocities. Bottom panel: Optical spectra at 30 days post explosion for the 1D simulation and the equatorial viewing angle from the 2D simulations compared to the observed SN 2011fe. The 2D NLTE simulation shows clear spectroscopic differences compared to the other simulations. This is illustrated by the key Ca ii near-infrared absorption feature which is highlighted. The 1D and 2D approximate NLTE simulations predict this feature appears at lower velocities (~8200Å), consistent with SN 2011fe. However, the 2D NLTE simulation predicts a much higher velocity feature (~7900Å) demonstrating the importance of both multidimensional and full NLTE radiative transfer simulations when evaluating the success of explosion models relative to observations.
References:
[1] Sim et al. 2013, MNRAS, 436, 333
[2] Dessart et al. 2014, MNRAS, 441, 3249
[3] Gronow et al. 2021, A&A, 649, A155
[4] Shingles et al. 2020, MNRAS, 492, 2029