56Ni dredge-up in Supernova 1987A

Abstract

We use early-time observations of He I 10830 Åto measure the extent of upward mixing of radioactive material in SN 1987A. This work develops and extends the work of Graham (1988), and places constraints on actual explosion models. The presence of the He I 10830 Å(2s^{3}S--2p^{3}P) line at $\geq 10$ days post-explosion implies re-ionisation by $γ$-rays from upwardly-mixed radioactive material produced during the explosion. Using the unmixed explosion model 10H (Woosley 1988) as well as mixed versions of it, we estimated the $γ$-ray energy deposition by applying a purely absorptive radiative transfer calculation. The deposition energy was used to find the ionisation balance as a function of radius, and hence the 2s$^{3}$S population density profile. This was then applied to a spectral synthesis model and the synthetic spectra were compared with the observations. Neither model 10H nor the mixed version, 10HMM, succeeded in reproducing the observed He I 10830 Åline. The discrepancy with the data found for 10HMM is particularly significant, as this model has successfully reproduced the X-ray and $γ$-ray observations and the UVOIR light curve. We find that a match to the He I line profile is achieved by reducing the extent of mixing in 10HMM. Our reduced-mixing models also reproduce the observed $γ$-ray line light curves and the iron-group velocities deduced from late-time infrared line profiles. We suggest that the He I line method provides a more sensitive measure of the extent of mixing in a type II supernova explosion.