A Morphological Classification Model to Identify Unresolved PanSTARRS1 Sources II: Update to the PS1 Point Source Catalog

TL;DR

This paper updates the PS1 Point Source Catalog by adding approximately 144 million new classifications using a modified machine learning approach, increasing the catalog size by 10% and aiding transient detection.

Contribution

The paper introduces an updated classification method applying PS1 forced photometry, expanding the catalog significantly with minimal performance loss.

Findings

Added ~144 million classifications, increasing catalog size by 10%.

Performance degraded by 6-8% compared to original method.

Enhanced catalog improves transient detection efficiency.

Abstract

We present an update to the PanSTARRS-1 Point Source Catalog (PS1 PSC), which provides morphological classifications of PS1 sources. The original PS1 PSC adopted stringent detection criteria that excluded hundreds of millions of PS1 sources from the PSC. Here, we adapt the supervised machine learning methods used to create the PS1 PSC and apply them to different photometric measurements that are more widely available, allowing us to add $\sim$144 million new classifications while expanding the the total number of sources in PS1 PSC by $\sim$10%. We find that the new methodology, which utilizes PS1 forced photometry, performs $\sim$6-8% worse than the original method. This slight degradation in performance is offset by the overall increase in the size of the catalog. The PS1 PSC is used by time-domain surveys to filter transient alert streams by removing candidates coincident with point sources that are likely to be Galactic in origin. The addition of $\sim$144 million new classifications to the PS1 PSC will improve the efficiency with which transients are discovered.