A Deep HST Study of the Globular Cluster NGC 6397: Reduction Methods

TL;DR

This paper presents novel reduction methods for analyzing extremely faint stars in HST data of NGC 6397, emphasizing techniques to distinguish stars from galaxies and optimize detection in deep, multi-exposure images.

Contribution

The paper introduces new data reduction techniques for faint star detection in crowded fields, utilizing artificial-star tests to improve completeness and measurement accuracy.

Findings

Effective methods for separating stars from galaxies in deep HST images

Optimized algorithms for combining multiple exposures to detect faint stars

Artificial-star tests enhance detection completeness and measurement precision

Abstract

We describe here the reduction methods that we developed to study the faintest red dwarfs and white dwarfs in an outer field of NGC6397, which was observed by \hst for 126 orbits in 2005. The particular challenge of this data set is that the faintest stars are not readily visible in individual exposures, so special care must be taken to combine the information in all the exposures in order to identify and measure them. Unfortunately, it is hard to find the faintest stars without also finding a large number of faint galaxies, so we developed specialized tools to distinguish between the point-like stars and the barely resolved galaxies. We found that artificial-star tests, while obviously necessary for completeness determination, can also play an important role in helping us optimize our finding and measuring algorithms. Although this paper focuses on this data set specifically, many of the techniques are new and might find application in other work, particularly when a large number of images is available for a single field.