2015MNRAS.451.2212G

We investigate the early-time light curves of a large sample of 223 Type II supernovae (SNe II) from the Sloan Digital Sky Survey and the Supernova Legacy Survey. Having a cadence of a few days and sufficient non-detections prior to explosion, we constrain rise-times, i.e. the durations from estimated first to maximum light, as a function of effective wavelength. At rest-frame g' band (λeff = 4722 Å), we find a distribution of fast rise-times with median of (7.5±0.3) d. Comparing these durations with analytical shock models of Rabinak & Waxman and Nakar & Sari, and hydrodynamical models of Tominaga et al., which are mostly sensitive to progenitor radius at these epochs, we find a median characteristic radius of less than 400 solar radii. The inferred radii are on average much smaller than the radii obtained for observed red supergiants (RSG). Investigating the post-maximum slopes as a function of effective wavelength in the light of theoretical models, we find that massive hydrogen envelopes are still needed to explain the plateaus of SNe II. We therefore argue that the SN II rise-times we observe are either (a) the shock cooling resulting from the core collapse of RSG with small and dense envelopes, or (b) the delayed and prolonged shock breakout of the collapse of an RSG with an extended atmosphere or embedded within pre-SN circumstellar material.