Star Formation Laws: the Effects of Gas Cloud Sampling

TL;DR

This paper models how sampling effects of molecular clouds influence observed star formation laws at different spatial scales, revealing scale-dependent slopes and scatter that reconcile various observational results.

Contribution

It introduces a simple stochastic model demonstrating the impact of cloud sampling on the star formation-molecular gas relation, highlighting the scale at which the relation becomes fully sampled.

Findings

The slope of the SFR-molecular gas relation approaches unity at scales >1-2 kpc.

Sampling effects cause the observed scatter and slope variations at small scales.

A steep (>1) power-law index between SFR and molecular gas is supported by galaxy data.

Abstract

Recent observational results indicate that the functional shape of the spatially-resolved star formation-molecular gas density relation depends on the spatial scale considered. These results may indicate a fundamental role of sampling effects on scales that are typically only a few times larger than those of the largest molecular clouds. To investigate the impact of this effect, we construct simple models for the distribution of molecular clouds in a typical star-forming spiral galaxy, and, assuming a power-law relation between SFR and cloud mass, explore a range of input parameters. We confirm that the slope and the scatter of the simulated SFR-molecular gas surface density relation depend on the size of the sub-galactic region considered, due to stochastic sampling of the molecular cloud mass function, and the effect is larger for steeper relations between SFR and molecular gas. There is a general trend for all slope values to tend to ~unity for region sizes larger than 1-2 kpc, irrespective of the input SFR-cloud relation. The region size of 1-2 kpc corresponds to the area where the cloud mass function becomes fully sampled. We quantify the effects of selection biases in data tracing the SFR, either as thresholds (i.e., clouds smaller than a given mass value do not form stars) or backgrounds (e.g., diffuse emission unrelated to current star formation is counted towards the SFR). Apparently discordant observational results are brought into agreement via this simple model, and the comparison of our simulations with data for a few galaxies supports a steep (>1) power law index between SFR and molecular gas.