On the association between core-collapse supernovae and H II regions

Previous studies of the location of core-collapse supernovae (ccSNe) in their host galaxies have variously claimed an association with H II regions; no association or an association only with hydrogen-deficient ccSNe. Here, we examine the immediate environments of 39 ccSNe whose positions are well known in nearby (≤15 Mpc), low-inclination (≤65°) hosts using mostly archival, continuum-subtracted Hα ground-based imaging. We find that 11 out of 29 hydrogen-rich ccSNe are spatially associated with H II regions (38 ± 11 per cent), versus 7 out of 10 hydrogen-poor ccSNe (70 ± 26 per cent). Similar results from Anderson et al. led to an interpretation that the progenitors of Type Ib/c ccSNe are more massive than those of Type II ccSNe. Here, we quantify the luminosities of H II region either coincident with or nearby to the ccSNe. Characteristic nebulae are long-lived (∼20 Myr) giant H II regions rather than short-lived (∼4 Myr) isolated, compact H II regions. Therefore, the absence of an H II region from most Type II ccSNe merely reflects the longer lifetime of stars with ⪅12 M⊙ than giant H II regions. Conversely, the association of an H II region with most Type Ib/c ccSNe is due to the shorter lifetime of stars with >12 M⊙ stars than the duty cycle of giant H II regions. Therefore, we conclude that the observed association between certain ccSNe and H II provides only weak constraints upon their progenitor masses. Nevertheless, we do favour lower mass progenitors for two Type Ib/c ccSNe that lack associated nebular emission, a host cluster or a nearby giant H II region. Finally, we also reconsider the association between long gamma-ray bursts and the peak continuum light from their (mostly) dwarf hosts, and conclude that this is suggestive of very high mass progenitors, in common with previous studies.