The phase spiral in Gaia DR3

Aims. We aim to study the phase spiral in the Milky Way (MW) disc with data from the third data release of Gaia (DR3) and use it as an inference tool to decipher the late-time evolution of the Galaxy. Methods. We used an edge-detection algorithm to find the border of the phase spiral, allowing us to robustly quantify its shape at different positions and for different selections. We calculated the time of onset of the phase-mixing by determining the different turns of the phase spiral and using the vertical frequencies from commonly used models of the gravitational potential of the MW. Results. We find that the phase spiral extends down to -1.2 kpc in height below the plane (about 3-5 scale heights of the thin disc) and beyond +/- 50 km s(-1) in V-Z. We see a secondary branch mostly at positive vertical velocities when coloured by azimuthal velocity and in the counts projection. We also find complex variations of the phase spirals with angular momentum and azimuth. All these findings are possible evidence of multiple perturbations (from different times or from different perturbers) and/or of the complexity of the phase-mixing process. We detect the phase spiral from 6 to 11 kpc from the Galactic centre and find signatures of vertical asymmetries 1-2 kpc beyond this range. We measure small but clear variations with azimuth. When we determine the phase mixing times from the phase spiral at different angular momenta and using the different spiral turns (at different Z), we obtain inconsistent times with systematic differences (times increasing with |L-Z| and with |Z|). Our determinations are mostly in the range of [0.3-0.9] Gyr, with an average of 0.5 Gyr. The inconsistencies do not change when using different commonly used potential models for the MW, different stellar distances, or frequencies for different kinetic temperatures; they could stem from the inconsistency of the assumed gravitational potentials with the true MW, and from oversimplification of the modelling, in particular where self-gravity is neglected or where multiple perturbations and/or interference with other processes are not considered. Conclusions. The wealth of information provided by the new Gaia DR3 data should encourage us to make progress in crucial modelling aspects of the disc dynamics, such as non-equilibrium, self-gravity, propagation of different types of bending waves, and interactions between different mechanisms. Such advancements could finally enable us to establish the origin of the phase spiral and its relation to the Sagittarius dwarf galaxy.