Identifying Brightest Cluster Galaxies in the SPT-3G Cluster Sample with Machine Learning

Brightest Cluster Galaxies (BCGs) are the brightest, most massive galaxies found in galaxy clusters and are often spatially correlated with the largest dark matter halo in galaxy clusters. The mass content of galaxy clusters is dominated by dark matter, therefore knowing approximately where the largest dark matter halo sits in a galaxy cluster is vital information for modeling galaxy clusters to study their evolution and lensed objects behind them. However, current tools for detecting BCGs in galaxy clusters can be inaccurate, with recent studies of optical red-sequence selection finding misidentification rates of 15-25%. In this poster, we present a tool "Image Marker," a tool for marking and logging coordinates of marks in a large sample of images. We will be using Image Marker to produce a human-curated training set of identified BCGs using the upcoming galaxy cluster sample from the SPT-3G experiment. We also examine the method of detecting BCGs with machine learning that we will begin applying soon.