Cloud formation in colliding flows: influence of the choice of cooling function

TL;DR

This study compares simple and detailed cooling functions in simulations of molecular cloud formation, revealing that while some properties are insensitive, cloud morphology and velocity distribution are significantly affected by the cooling approach.

Contribution

It provides a direct comparison between parametrized and non-equilibrium chemistry cooling functions, highlighting their impact on cloud morphology and dynamics in simulations.

Findings

Cloud properties like mass and cold gas fractions are insensitive to cooling function choice.

Cloud morphology and velocity distribution depend strongly on the cooling approach.

Proper non-equilibrium treatment of ionization and recombination is essential for accurate high-temperature cooling modeling.

Abstract

We study the influence of the choice of cooling function on the formation of molecular clouds in high-resolution three-dimensional simulations of converging flows. We directly compare the results obtained using the simple, parametrized cooling function introduced by Koyama & Inutsuka (2002) and used by a number of converging flow studies with the results of the detailed calculation of the non-equilibrium chemistry and thermal balance of the gas. We find that a number of the cloud properties, such as the mass and volume filling fractions of cold gas, are relatively insensitive to the choice of cooling function. On the other hand, the cloud morphology and the large-scale velocity distribution of the gas do strongly depend on the cooling function. We show that the differences that we see can largely be explained by differences in the way that Lyman-alpha cooling is treated in the two complementary approaches, and that a proper non-equilibrium treatment of the ionisation and recombination of the gas is necessary in order to model the high-temperature cooling correctly. We also investigate the properties of the dense clumps formed within the cloud. In agreement with previous models, we find that the majority of these clumps are not self-gravitating, suggesting that some form of large-scale collapse of the cloud may be required in order to produce gravitationally unstable clumps and hence stars. Overall, the physical properties of the dense clumps are similar in both simulations, suggesting that they do not depend strongly on the choice of cooling function. However, we do find a systematic difference of around 10K in the mean temperatures of the clumps produced by the two models.