Old cold dust heated by supernova 1987A

The mid-infrared emission (at wavelengths near 10 μm) from SN1987A, which had been fading steadily since reaching a peak about 120 days after outburst, began to increase again on about day 450¹. This increase has continued up to at least day 578, and is probably due to heating of dust grains by light emitted at the optical maximum. Here we present additional observations and discuss the mass, location and nature of the emitting dust. The dust grains cannot have formed in the ejecta of the supernova, nor do they appear to be symmetrically distributed about it; rather, they are located at a distance of about 1 light year from, and mostly behind, the supernova.