Transient protostellar cores in high mass star forming regions revealed by time-resolved synthetic imaging of dust emission

TL;DR

This study uses time-resolved synthetic 1.3 mm ALMA-like imaging of a high-mass, globally collapsing clump to investigate how cores identified by dendrogram analyses relate to protostars. By combining high-resolution hydrodynamic simulations, radiative-transfer post-processing, and CASA-based interferometric observations, the authors show that most dendrogram-identified cores are transient and often devoid of protostars, with cores along feeder filaments changing boundaries over time. Core temperatures vary widely, and deriving masses with per-core temperatures yields CMFs that differ substantially from those assuming a single temperature, illustrating biases in common observational approaches. The work demonstrates that only global mm-flux and core-count trends correlate robustly with time, highlighting the dynamic, filamentary nature of massive star formation and cautioning against assuming direct, persistent core–star correspondence in such regions.

Abstract

The connection between dense gas cores and their infant protostars is key to understanding how stars form in molecular clouds. In this paper we investigate the properties, persistence, and protostellar content of cores that would be identified by a dendrogram analysis of 1.3 mm ALMA images. We use a time series of synthetic images produced by post-processing a simulation of star formation in a massive globally collapsing clump, with polaris to calculate dust radiative transfer and CASA to generate synthetic ALMA data. Identifying sinks in the simulation with protostars, we find that most dendrogram-identified cores do not contain any protostars, with many cores being transient features associated with clumpy flow along feeder filaments. Cores with protostars generally host <4, and protostellar mass is not strongly correlated with the mass of the parent cores due to their transience and shifting boundaries. Calculating observationally-relevant intensity-weighted average temperatures for all cores, we find that even at early times the core temperature distribution spans tens of Kelvin, and its width increases with time. The 1.3 mm peak and integrated intensity of the brightest mm core do not increase monotonically as the most massive associated protostar grows, indicating it cannot be assumed that brighter mm sources host more massive protostars. Leveraging the time domain, we test observational properties that have been proposed as potential evolutionary indicators and find that only the total 1.3 mm flux density of the region, the total 1.3 mm flux density in cores, and the number of cores show strong, statistically significant correlation with time.

Figures (17)

• Figure 1: The column density of the cloud at $T=4.135$ Myr, centred on the location of the most massive sink in the snapshot. The left panel shows the large scale filamentary structure of the cloud that creates overdense regions where sink particles form. The right panel shows a zoom view; the field of view shown corresponds to the yellow rectangle in the left panel. In the right panel, positions of sink particles with masses $\geq 0.02 M_{\odot}$ at the end of the simulation are overlaid, colour-coded by their current total mass.
• Figure 2: Final synthetic ALMA 1.3 mm image of the snapshot shown in the right panel of Fig. \ref{['fig:NH']}, after post-processing with POLARIS and CASA. The $1.3$ mm dust emission is shown in colourscale, overlaid with markers showing the current masses of sink particles above $1 M_{\odot}$ at the end of the simulation; the contours show cores identified by astrodendro (see Section \ref{['sec:core_ident']}).
• Figure 3: Synthetic ALMA 1.3 mm continuum images for each snapshot, centred on the position of the most massive sink and with the core boundaries identified with astrodendro (Section \ref{['sec:core_ident']}, Table \ref{['tab:properties']}) overlaid in blue. The images shown have been corrected for the primary beam response; the colourscale and field of view are the same for all images and the scalebars assume D=2 kpc (Section \ref{['sec:RT']}). In each panel, the synthesised beam is shown at lower left and the time of the snapshot in Myr and number of identified cores are given at upper left.
• Figure 4: Synthetic ALMA 1.3 mm peak intensity (left panel) and integrated flux density (right panel) of the brightest mm core in each timestep (the core with the highest peak intensity), plotted against the equivalent stellar mass of the most massive sink in this core. The colour of each point indicates the total number of cores identified in the synthetic image of that snapshot (Table \ref{['tab:snapshot_stats']}, see also Figure \ref{['fig:24paneldendro']}). Once the equivalent stellar mass of the most massive sink reaches a few M$_{\odot}$, it has little correlation with the mm peak intensity.
• Figure 5: Number of cores identified with astrodendro within the 30% mosaic response (see Section \ref{['sec:core_ident']}) plotted against snapshot time. Points are colour coded by the equivalent stellar mass of the most massive sink within the brightest mm core. Note that this generally corresponds to the most massive sink in the simulation, with the notable exception of T=4.135 Myr, when the most massive sink in the simulation is not within the brightest mm core.