51�������

Abstract:

There is a general consensus that Type Ia supernovae (SNe Ia) comprise the thermonuclear explosions of a C/O white dwarf (WD) in a binary system. However, the nature of their progenitors and explosion mechanism still remains elusive. Their early light curves can offer strong constraints on these parameters, as they trace the outermost layers and provide evidence of their initial conditions. In this work, we study the Zwicky Transient Facility (ZTF) data release 2 sample of SNe Ia in search for objects with early flux excess (EEx SNe Ia). Two main approaches were used: i) finding deviations between early-time observations and models without early excess and ii) finding an agreement between observations and models with early excess. For the former approach, different methods, such as power-law and SALT2 fits, were used to identify EEx SN Ia candidates. For the second approach, we used double-detonation (DD) and companion-interaction (CI) models, finding 145 and 199 candidates, respectively. In general, there is a disagreement between the methods in the exact objects identified as EEx SNe Ia as well as in their number. By combining multiple methods, we found 17 strong EEx SN Ia candidates, 6 of which are of the overluminous subtype (five 91T-like and one 03 fg-like). The most prominent excesses reach ∼10–17% of peak flux (up to ∼−17 mag). We found that the best EEx SN Ia candidates have higher stretch on average compared to the rest of the SNe Ia. They also seem to prefer lower mass and bluer host galaxies, potentially linked to younger stellar populations. Additionally, we estimated the fraction of EEx SNe Ia to be ≲25% of all SNe Ia, in agreement with previous works, with overluminous subtypes having a higher relative fraction compared to normal SNe Ia. When studying the parameter distribution of the DD models for the EEx SN Ia candidates, we see a preference for intermediate WD core mass (1 M ⊙ ), a lower fraction of shell burning (20%), and heavier He-burning products ( 56 Ni). For the CI models, we see no trends with any of the parameters. We are unable to draw strong conclusions on whether one model is preferred over the other as a viable description of SN Ia explosions Finally, we highlight the challenge of identifying EEx SNe Ia in real time, making the triggering of early-time spectroscopic observations for these objects a difficult task.