For each target object, we provide information about coincident or nearby objects that have been cataloged by external surveys including SDSS, DESI Legacy Imaging Surveys, Pan-STARRS1, [Neo]WISE, and Gaia. This cross-matching can provide valuable contextual information about the host galaxy and/or known variable stars at the target location, which can be used in object classifications or statistical analysis of populations. The provided information includes basic information about the cross-matched object as well as a link to the database entry of the external catalog. For objects with multiple cross-matches, we provide a ranking of the cross-matched objects by how likely they are to actually be associated with the target object (as opposed to simply being coincident on the sky). The ranking is be integrated with our classification schemes to provide coherent interpretations of the data (for example, an object observed by LSST cannot simultaneously have high likelihoods of both being a nova and residing in a host galaxy at z>0.1).
In addition, we provide direct access to these external catalogs via public datasets hosted in the Google cloud, independent of the Pitt-Google Broker. Some of these were preexisting, and we are making the rest available as we come online.
We provide object classifications for a range of object types including supernovae (and subtypes), novae, and many classes of variable stars. We use open source, publicly available template fitting or machine learning classi cation codes when available and suciently accurate (e.g., for supernovae and variable stars). Our methods include modeling of selection efficiency and purity suitable for robust scientific analysis.