Observational Bias and Young Massive Cluster Characterisation II. Can Gaia accurately observe young clusters and associations?

TL;DR

This study evaluates Gaia DR3's ability to observe and characterize young massive clusters by simulating observations, analyzing detection accuracy, and comparing observed properties with original models across various distances.

Contribution

It introduces a method to assess Gaia's observational biases and accuracy in characterizing young clusters through synthetic simulations and real data comparison.

Findings

Qualitative conclusions are reliable despite high incompleteness.

Quantitative measurements of cluster properties are often inaccurate at larger distances.

Massive clusters like Westerlund2 are better characterized than smaller ones.

Abstract

Observations of clusters suffer from issues such as completeness, projection effects, resolving individual stars and extinction. As such, how accurate are measurements and conclusions are likely to be? Here, we take cluster simulations (Westerlund2- and Orion- type), synthetically observe them to obtain luminosities, accounting for extinction and the inherent limits of Gaia, then place them within the real Gaia DR3 catalogue. We then attempt to rediscover the clusters at distances of between 500pc and 4300pc. We show the spatial and kinematic criteria which are best able to pick out the simulated clusters, maximising completeness and minimising contamination. We then compare the properties of the 'observed' clusters with the original simulations. We looked at the degree of clustering, the identification of clusters and subclusters within the datasets, and whether the clusters are expanding or contracting. Even with a high level of incompleteness (e.g. $<2\%$ stellar members identified), similar qualitative conclusions tend to be reached compared to the original dataset, but most quantitative conclusions are likely to be inaccurate. Accurate determination of the number, stellar membership and kinematic properties of subclusters, are the most problematic to correctly determine, particularly at larger distances due to the disappearance of cluster substructure as the data become more incomplete, but also at smaller distances where the misidentification of asterisms as true structure can be problematic. Unsurprisingly, we tend to obtain better quantitative agreement of properties for our more massive Westerlund2-type cluster. We also make optical style images of the clusters over our range of distances.