Context. Recent statistical deconvolution methods have produced extended kinematical maps in a range of heliocentric distances that are a factor of two to three larger than those analysed in Gaia Collaboration (2018, A&A, 616, A11) based on the same data. Aims. In this paper, we use such maps to derive the rotation curve both in the Galactic plane and in off-plane regions and to analyse the density distribution. Methods. By assuming stationary equilibrium and axisymmetry, we used the Jeans equation to derive the rotation curve. Then we fit it with density models that include both dark matter and predictions of the MOND (Modified Newtonian dynamics) theory. Since the Milky Way exhibits deviations from axisymmetry and equilibrium, we also considered corrections to the Jeans equation. To compute such corrections, we ran N-body experiments of mock disk galaxies where the departure from equilibrium becomes larger as a function of the distance from the centre. Results. The rotation curve in the outer disk of the Milky Way that is constructed with the Jeans equation exhibits very low dependence on R and z and it is well-fitted both by dark matter halo and MOND models. The application of the Jeans equation for deriving the rotation curve, in the case of the systems that deviate from equilibrium and axisymmetry, introduces systematic errors that grow as a function of the amplitude of the average radial velocity. In the case of the Milky Way, we can observe that the amplitude of the radial velocity reaches ∼10% that of the azimuthal one at R ≈ 20 kpc. Based on this condition, using the rotation curve obtained from the Jeans equation to calculate the mass may overestimate its measurement.

Context. In our Paper I, by using statistical deconvolution methods, extended kinematics maps of Gaia-DR2 data have been produced in a range of heliocentric distances that are a factor of two to three larger than those analyzed previously by the Gaia Collaboration with the same data. It added the range of Galactocentric distances between 13 kpc and 20 kpc to the previous maps. Aims. Here, we investigate the dynamical effects produced by different mechanisms that can explain the radial and vertical components of these extended kinematic maps, including a decomposition of bending and breathing of the vertical components. This paper as a whole tries to be a compendium of different dynamical mechanisms whose predictions can be compared to the kinematic maps. Methods. Using analytical methods or simulations, we are able to predict the main dynamical factors and compare them to the predictions of the extended kinematic maps of Gaia-DR2. Results. The gravitational influence of Galactic components that are different from the disk, such as the long bar or bulge, the spiral arms, or a tidal interaction with Sagittarius dwarf galaxy, may explain some features of the velocity maps, especially in the inner parts of the disk. However, they are not sufficient in explaining the most conspicuous gradients in the outer disk. Vertical motions might be dominated by external perturbations or mergers, although a minor component may be due to a warp whose amplitude evolves with time. Here, we show with two different methods, which analyze the dispersion of velocities, that the mass distribution of the disk is flared. Despite these partial explanations, the main observed features can only be explained in terms of out-of-equilibrium models, which are either due to external perturbers or to the fact that the disk has not had time to reach equilibrium since its formation.