Using Hα emission as a tracer of ongoing (<16 Myr old) and near-ultraviolet (UV) emission as a tracer of recent (16–100 Myr old) star formation, we present constraints on the properties of core-collapse (CC) supernova (SN) progenitors through the association of their explosion sites with star-forming regions. Amalgamating previous results with those gained from new data, we present statistics of a large sample of SNe; 163.5 Type II (58 IIP, 13 IIL, 13.5 IIb, 19 IIn and 12 ‘impostors’, plus 48 with no sub-type classification) and 96.5 Type Ib/c (39.5 Ib and 52 Ic, plus five with no sub-type classification). Using pixel statistics we build distributions of associations of different SN types with host galaxy star formation. Our main findings and conclusions are as follows.
- 1An increasing progenitor mass sequence is observed, implied from an increasing association of SNe to host galaxy Hα emission. This commences with the Type Ia showing the weakest association, followed by the Type II, then the Ib, with the Type Ic showing the strongest correlation to star-forming regions. Thus, our progenitor mass sequence runs Ia–II–Ib–Ic.
- 2Overall, the Type Ibc SNe are found to occur nearer to bright H ii regions than SNe of Type II. This implies that the former have shorter stellar lifetimes, thus arising from more massive progenitor stars.
- 3While Type IIP SNe do not closely follow the ongoing star formation, they accurately trace the recent formation. This implies that their progenitors arise from stars at the low end of the CC SN mass sequence, consistent with direct detections of progenitors in pre-explosion imaging.
- 4Similarly, the Type IIn SNe trace recent but not the ongoing star formation. This implies that, contrary to the general consensus, the majority of these SN do not arise from the most massive stars.
Results and suggestive constraints are also presented for the less numerous SNe of Types IIL and IIb, and SN ‘impostors’. Finally, we present an analysis of possible biases in the data, the results of which argue strongly against any selection effects that could explain the relative excess of Type Ibc SNe within bright H ii regions. Thus, intrinsic progenitor differences in the sense of the mass sequence we propose remain the most plausible explanation of our findings.