Approximation of Multivariate Time Series in Astronomy Using Vector Gaussian Processes

In astronomy, it is often necessary to approximate sparse vector functions, such as multiband light curves of transient objects. Traditional methods typically treat each photometric band independently, ignoring interband correlations and thus failing to fully exploit the available multiband information in sparsely sampled datasets. We propose a vector Gaussian process (GP) framework that models multi-band light curves jointly as a single vector function, capturing cross-band correlations and improving interpolation quality. Applied to the Open Supernova Catalog, our method yields more accurate bolometric light curves of superluminous supernovae and enables anomaly detection by identifying photometrically unusual events. Compared to univariate baselines, it demonstrates higher stability and predictive power. The approach is implemented in the open-source gp-multistate-kernel Python package and is well suited for time-domain data from current and future surveys such as the Vera C. Rubin Observatory’s Legacy Survey of Space and Time (LSST).