Supernova spectral time series contain a wealth of information about the progenitor and explosion process of these energetic events. The modeling of these data requires the exploration of very high dimensional posterior probabilities with expensive radiative transfer codes. Even modest parametrizations of supernovae contain more than ten parameters and a detailed exploration demands at least several million function evaluations. Physically realistic models require at least tens of CPU minutes per evaluation putting a detailed reconstruction of the explosion out of reach of traditional methodology. The advent of widely available libraries for the training of neural networks combined with their ability to approximate almost arbitrary functions with high precision allows for a new approach to this problem. Instead of evaluating the radiative transfer model itself, one can build a neural network proxy trained on the simulations but evaluating orders of magnitude faster. Such a framework is called an emulator or surrogate model. In this work, we present an emulator for the TARDIS supernova radiative transfer code applied to Type Ia supernova spectra. We show that we can train an emulator for this problem given a modest training set of a hundred thousand spectra (easily calculable on modern supercomputers). The results show an accuracy on the percent level (that are dominated by the Monte Carlo nature of TARDIS and not the emulator) with a speedup of several orders of magnitude. This method has a much broader set of applications and is not limited to the presented problem.

This work presents the observations and analysis of ATLAS19dqr/SN 2019bkc, an extraordinary rapidly evolving transient event located in an isolated environment, tens of kiloparsecs from any likely host. Its light curves rise to maximum light in $5-6$ d and then display a decline of $\Delta m_{15} \sim5$ mag. With such a pronounced decay, it has one of the most rapidly evolving light curves known for a stellar explosion. The early spectra show similarities to normal and `ultra-stripped' type Ic SNe, but the early nebular phase spectra, which were reached just over two weeks after explosion, display prominent calcium lines, marking SN 2019bkc as a Ca-rich transient. The Ca emission lines at this phase show an unprecedented and unexplained blueshift of 10,000 -- 12,000 km/s. Modelling of the light curve and the early spectra suggests that the transient had a low ejecta mass of $0.2 - 0.4$ M$_\odot$ and a low kinetic energy of $ (2-4)\times 10^{50}$ erg, giving a specific kinetic energy $\sim1$ [$10^{51}$ erg]/M$_\odot$. The origin of this event cannot be unambiguously defined. While the abundance distribution used to model the spectra marginally favours a progenitor of white dwarf origin through the tentative identification of \ArII, the specific kinetic energy, which is defined by the explosion mechanism, is found to be more similar to an ultra-stripped core-collapse events. SN 2019bkc adds to the diverse range of physical properties shown by Ca-rich events.

There are now hundreds of publicly available supernova spectral time series. Radiative transfer modeling of this data gives insights into the physical properties of these explosions such as the composition, the density structure, or the intrinsic luminosity---this is invaluable for understanding the supernova progenitors, the explosion mechanism, or for constraining the supernova distance. However, a detailed parameter study of the available data has been out of reach due to the high dimensionality of the problem coupled with the still significant computational expense. We tackle this issue through the use of machine-learning emulators, which are algorithms for high-dimensional interpolation. These use a pre-calculated training dataset to mimic the output of a complex code but with run times orders of magnitude shorter. We present the application of such an emulator to synthetic type II supernova spectra generated with the TARDIS radiative transfer code. The results show that with a relatively small training set of 780 spectra we can generate emulated spectra with interpolation uncertainties of less than one percent. We demonstrate the utility of this method by automatic spectral fitting of two well-known type IIP supernovae; as an exemplary application, we determine the supernova distances from the spectral fits using the tailored-expanding-photosphere method. We compare our results to previous studies and find good agreement. This suggests that emulation of TARDIS spectra can likely be used to perform automatic and detailed analysis of many transient classes putting the analysis of large data repositories within reach.

As a candidate 'super-Chandrasekhar' or 09dc-like Type Ia supernova (SN Ia), SN 2012dn shares many characteristics with other members of this remarkable class of objects but lacks their extraordinary luminosity. Here, we present and discuss the most comprehensive optical data set of this SN to date, comprised of a densely sampled series of early-time spectra obtained within the Nearby Supernova Factory project, plus photometry and spectroscopy obtained at the VLT about 1 yr after the explosion. The light curves, colour curves, spectral time series and ejecta velocities of SN 2012dn are compared with those of other 09dc-like and normal SNe Ia, the overall variety within the class of 09dc-like SNe Ia is discussed, and new criteria for 09dc-likeness are proposed. Particular attention is directed to additional insight that the late-phase data provide. The nebular spectra show forbidden lines of oxygen and calcium, elements that are usually not seen in late-time spectra of SNe Ia, while the ionisation state of the emitting iron plasma is low, pointing to low ejecta temperatures and high densities. The optical light curves are characterised by an enhanced fading starting ~60 d after maximum and very low luminosities in the nebular phase, which is most readily explained by unusually early formation of clumpy dust in the ejecta. Taken together, these effects suggest a strongly perturbed ejecta density profile, which might lend support to the idea that 09dc-like characteristics arise from a brief episode of interaction with a hydrogen-deficient envelope during the first hours or days after the explosion.

We present the results of our photometric and spectroscopic follow-up of the intermediate-luminosity optical transient AT 2017jfs. At peak, the object reaches an absolute magnitude of Mg=-15.46+-0.15 mag and a bolometric luminosity of 5.5x10^41 erg/s. Its light curve has the double-peak shape typical of Luminous Red Novae (LRNe), with a narrow first peak bright in the blue bands, while the second peak is longer lasting and more luminous in the red and near-infrared (NIR) bands. During the first peak, the spectrum shows a blue continuum with narrow emission lines of H and Fe II. During the second peak, the spectrum becomes cooler, resembling that of a K-type star, and the emission lines are replaced by a forest of narrow lines in absorption. About 5 months later, while the optical light curves are characterized by a fast linear decline, the NIR ones show a moderate rebrightening, observed until the transient disappeared in solar conjunction. At these late epochs, the spectrum becomes reminiscent of that of M-type stars, with prominent molecular absorption bands. The late-time properties suggest the formation of some dust in the expanding common envelope or an IR echo from foreground pre-existing dust. We propose that the object is a common-envelope transient, possibly the outcome of a merging event in a massive binary, similar to NGC4490-2011OT1.

We present substantial extensions to the Monte Carlo radiative transfer code TARDIS to perform spectral synthesis for type II supernovae. By incorporating a non-LTE ionization and excitation treatment for hydrogen, a full account of free-free and bound-free processes, a self-consistent determination of the thermal state and by improving the handling of relativistic effects, the improved code version includes the necessary physics to perform spectral synthesis for type II supernovae to high precision as required for the reliable inference of supernova properties. We demonstrate the capabilities of the extended version of TARDIS by calculating synthetic spectra for the prototypical type II supernova SN1999em and by deriving a new and independent set of dilution factors for the expanding photosphere method. We have investigated in detail the dependence of the dilution factors on photospheric properties and, for the first time, on changes in metallicity. We also compare our results with two previously published sets of dilution factors by Eastman et al. (1996) and by Dessart & Hillier (2005), and discuss the potential sources of the discrepancies between studies.