The Magnetar Model for Type I Superluminous Supernovae. I. Bayesian Analysis of the Full Multicolor Light-curve Sample with MOSFiT

Abstract We use the new Modular Open Source Fitter for Transients to model 38 hydrogen-poor superluminous supernovae (SLSNe). We fit their multicolor light curves with a magnetar spin-down model and present posterior distributions of magnetar and ejecta parameters. The color evolution can be fit with a simple absorbed blackbody. The medians (1 σ ranges) for key parameters are spin period 2.4 ms (1.2–4 ms), magnetic field G ( G), ejecta mass 4.8 (2.2–12.9 ), and kinetic energy erg ( erg). This significantly narrows the parameter space compared to our uninformed priors, showing that although the magnetar model is flexible, the parameter space relevant to SLSNe is well constrained by existing data. The requirement that the instantaneous engine power is ∼10 44 erg at the light-curve peak necessitates either large rotational energy ( ms), or more commonly that the spin-down and diffusion timescales be well matched. We find no evidence for separate populations of fast- and slow-declining SLSNe, which instead form a continuum in light-curve widths and inferred parameters. Variations in the spectra are explained through differences in spin-down power and photospheric radii at maximum light. We find no significant correlations between model parameters and host galaxy properties. Comparing our posteriors to stellar evolution models, we show that SLSNe require rapidly rotating (fastest 10%) massive stars ( ), which is consistent with their observed rate. High mass, low metallicity, and likely binary interaction all serve to maintain rapid rotation essential for magnetar formation. By reproducing the full set of light curves, our posteriors can inform photometric searches for SLSNe in future surveys.