2022ApJ...936..114M

A growing number of core-collapse supernovae (SNe) that show evidence for interaction with dense circumstellar medium (CSM) are accompanied by “precursor” optical emission rising weeks to months prior to the explosion. The precursor luminosities greatly exceed the Eddington limit of the progenitor star, implying that they are accompanied by substantial mass loss. Here, we present a semi-analytic model for SN precursor light curves, which we apply to constrain the properties and mechanisms of the pre-explosion mass loss. We explore two limiting mass-loss scenarios: (1) an “eruption” arising from shock breakout following impulsive energy deposition below the stellar surface; and (2) a steady “wind,” due to sustained heating of the progenitor envelope. The eruption model, which resembles a scaled-down version of Type IIP SNe, can explain the luminosities and timescales of well-sampled precursors, for ejecta masses ∼ 0.1-1 M_⊙ and velocities ∼ 100-1000 km s^–1. By contrast, the steady wind scenario cannot explain the highest precursor luminosities ≳ 10⁴¹ erg s^–1, under the constraint that the total ejecta mass does not exceed the entire progenitor mass (though the less luminous SN 2020tlf precursor can be explained by a mass-loss rate ∼ 1 M_⊙ yr^–1). However, shock interaction between the wind and pre-existing (earlier ejected) CSM may boost its radiative efficiency and mitigate this constraint. In both the eruption and wind scenarios, the precursor ejecta forms compact (≲10¹⁵ cm) optically thick CSM at the time of core collapse; though only directly observable via rapid post-explosion spectroscopy (≲ a few days before being overtaken by the SN ejecta), this material can boost the SN luminosity via shock interaction.