Can photometry alone classify supernovae without spectroscopy?

Astronomers developed an unsupervised classification approach that treats supernovae as a mixture of distinct populations. By fitting light curves to radioactive-decay models and using a statistical mixture model, they separated Type Ia from Type Ibc supernovae with 90% accuracy—without any labeled training data. This matters because next-generation surveys like Rubin will collect millions of transient observations, making complete spectroscopic follow-up impossible; photometry alone could now suffice.