RJ-plots: An improved method to classify structures objectively

TL;DR

RJ-plots is an enhanced automated classification method for interstellar structures, improving accuracy over previous techniques and revealing correlations between structure concentration and star formation metrics.

Contribution

The paper introduces RJ-plots, an improved automated classification technique for interstellar structures, with better distinction of morphologies and multi-scale application capabilities.

Findings

RJ-plots improves classification accuracy for ring-like and concentrated clouds.

Strong correlation between central concentration and star formation efficiency.

Spherical structures become more common at smaller scales but are never dominant.

Abstract

The interstellar medium is highly structured, presenting a range of morphologies across spatial scales. The large data sets resulting from observational surveys and state-of-the-art simulations studying these hierarchical structures means that identification and classification must be done in an automated fashion to be efficient. Here we present RJ-plots, an improved version of the automated morphological classification technique J-plots developed by Jaffa et al. (2018). This method allows clear distinctions between quasi-circular/elongated structures and centrally over/under-dense structures. We use the recent morphological SEDIGISM catalogue of Neralwar et al. (2022) to show the improvement in classification resulting from RJ-plots, especially for ring-like and concentrated cloud types. We also find a strong correlation between the central concentration of a structure and its star formation efficiency and dense gas fraction, as well as a lack of correlation with elongation. Furthermore, we use the accreting filament simulations of Clarke et al. (2020) to highlight a multi-scale application of RJ-plots, finding that while spherical structures become more common at smaller scales they are never the dominant structure down to $r\sim0.03$ pc.