Context. The physical processes driving the formation of Galactic spiral arms are still under debate. Studies using open clusters favour the description of the Milky Way spiral arms as long-lived structures following the classical density wave theory. Current studies comparing the Gaia DR2 field stars kinematic information of the solar neighbourhood to simulations, find a better agreement with short-lived arms with a transient behaviour.
Aims. Our aim is to provide an observational, data-driven view of the Milky Way spiral structure and its dynamics using open clusters as the main tracers, and to contrast it with simulation-based approaches. We used the most complete catalogue of Milky Way open clusters, with astrometric Gaia EDR3 updated parameters, estimated astrophysical information, and radial velocities, to revisit the nature of the spiral pattern of the Galaxy.
Methods. We used a Gaussian mixture model to detect overdensities of open clusters younger than 30 Myr that correspond to the Perseus, Local, Sagittarius, and Scutum spiral Arms, respectively. We used the birthplaces of the open cluster population younger than 80 Myr to trace the evolution of the different spiral arms and compute their pattern speed. We analysed the age-distribution of the open clusters across the spiral arms to explore the differences in the rotational velocity of stars and spiral arms.
Results. We are able to increase the range in Galactic azimuth where present-day spiral arms are described, better estimating its parameters by adding 264 young open clusters to the 84 high-mass star-forming regions used so far, thus increasing the number of tracers by 314%. We used the evolution of the open clusters from their birth positions to find that spiral arms nearly co-rotate with field stars at any given radius, discarding a common spiral pattern speed for the spiral arms explored.
Conclusions. The derivation of different spiral pattern speeds for the different spiral arms disfavours classical density waves as the main drivers for the formation of the Milky Way spiral structure, and it is in better agreement with simulation-based approaches that tend to favour transient spirals. The increase in the number of known open clusters, as well as in their derived properties, allows us to use them as effective spiral structure tracers and homogenise the view from open clusters and field stars on the nature of the Galactic spiral arms.