This document pre-registers an empirical test that will be run on upcoming Rubin Observatory supernova data. The test addresses a small but reproducible signal in Type Ia supernovae — the standard candles astronomers use to measure cosmic distances — in which the apparent brightness of each supernova varies slightly with the amount and type of foreground starlight along its line of sight. Characterized on the existing Pantheon+ supernova compilation, the signal accounts for roughly 1.6% of the residual variance in inferred distances and follows a specific signed pattern between blue light (from young stars) and near-infrared light (from old stars) in the foreground galaxies. Before Rubin Year-1 data are released, this document locks in the exact statistical thresholds that determine whether the signal will be confirmed or falsified on the independent Rubin sample. The locked content is: three subsample tests of the headline correlation, eighteen specific predictions on spatial and spectral features, and eight cosmic-structure sign predictions. All thresholds are calibrated on the prior Pantheon+ characterization; reproduction on independent data is the test. No claim is made about dark energy, the Hubble tension, or any specific cosmological model. The claim is empirical: a specific correlation, locked in advance, ready for independent verification.
Duaa Mustafa (Wed,) studied this question.