Supernova 1987A was a "failed supernova" twenty thousand years before its jet-driven explosion

I show that the progenitor of supernova (SN) 1987A faded substantially and became red for hundreds of years to observers in and near the plane of the equatorial ring, because the equatorial ring, which was formed by a binary interaction between the SN 1987A progenitor and a main-sequence companion that ejected material about 20,000 years before the explosion, obscured the progenitor. For a few hundred years, the luminosity for such hypothetical observers was mainly light scattered by the two outer rings, which amounted to 5% of the progenitor luminosity. In present terms, this event would have been classified as a "failed supernova" by these observers, although there was no core collapse, no black hole formation, and nothing failed. Rather, this event was a type II ILOT (intermediate-luminosity optical transient), in which equatorial ejecta temporarily obscured the central source. After a few hundred years, the outer rings became transparent, the inner ring became more transparent, and the SN 1987A progenitor brightened over a few thousand years to its normal luminosity for equatorial observers. This finding strengthens the claim that "failed supernova" candidates are likely to be type II ILOTs. Although the binary interaction also spun up the core, the bipolar morphology of SN 1987A (its Keyhole) is misaligned with the triple-ring system. I use the jittering jets explosion mechanism (JJEM) to speculate on a scenario that might explain this misalignment. This study adds small but unique support to the claim that the JJEM is the primary explosion mechanism of CCSNe.