It is believed that the Galactic disk forms inside-out, in which the stellar population at increasing radii is younger and more metal poor. This picture is consistent with most Galactic Chemical Evolution(GCE) models which also predict a tight correlation between the metallicity and age of stars of a given radius. However, it is only a result of "steady state" without dynamical evolution of stars. In the simple GCE model, the system is only open for gas (allowing for infall, outflow, and radial inflows), stars are assumed to be remained where they were born. Sellwood & Binney (2002) argued that a star near the corotation resonance of a spiral could gain or lose enough angular momentum for its guiding radius to change by up to 2 kpc, leaving the star at its new mean galacto-centric distance with no increase in its epicyclic amplitude. The resulting radial migration or radial mixing is described as "churning" wich leads to a more efficient mixing for older stars so that a flatter radial metallicity gradient forms. In addition, the orbits of disk stars generally become more eccentric over time, which "blurs" measurements of the metallicity gradient when the galacto-centric distance of older stars are used. In fact, the metallicity distribution of all but the youngest stars shows a broad distribution, which persists even after correcting for epicyclic blurring. The broad metallicity distribution contradicts with the prediction of chemical evolution models.

　 YU Jincheng et al. from Shanghai Astronomical observatory of CAS have selected two stellar samples from the Geneva Copenhagen survey and the Radial Velocity Experiment, each sample contains about 10,000 local thin-disk, main-sequence stars. They use guiding radius which is determined by the conservation of z-direction angular momentum, to eliminate the blurring effects. And they also use the effective temperature of the main sequence stars as a proxy of stellar age. They show that the metallicity gradient flattens as the age increases. Combining with the M-dwarf data, they also find the tread of decreasing metallicity gradient with increasing age breaks to a flat tread around 10 Gyr, roughly the age of the thin disk, which is also the upper time limit of the radial mixing effect. The work will be published in Astrophysical Journal of August 2012.

　 Although the results are consistent with the prediction of radial migration theory, it is different from most classical GCE models. The most serious uncertainty comes from the age estimation of stars. We need larger sample stars with more accurate age, metallicity, radial velocity and distances. The Chinese LAMOST survey is expected to play important role in this area. We are going to acquire millions of stellar spectra in the Milky Way disk, which will provide much larger sample for testing the theory. (Reported by YU Jincheng of SHAO, June 6, 2012)