Disk evolution in the solar neighborhood I. Disk frequencies from 1 to 100 Myr

Aims. We study the evolution of circumstellar disks in 22 young (1 to 100 Myr) nearby (within 500 pc) associations over the entire mass spectrum using photometry covering from the optical to the mid-infrared. Methods. We compiled a catalog of 2340 spectroscopically-confirmed members of these nearby associations. We analyzed their spectral energy distributions and searched for excess related to the presence of protoplanetary disks. The dataset has been analyzed in a homogeneous and consistent way, allowing for meaningful inter-comparison of results obtained for individual regions. Special attention was given to the sensitivity limits and spatial completeness of the observations. Results. We derive disk fractions as probed by mid-infrared excess in the 22 regions. The unprecedented size of our sample allows us to confirm the timescale of disk decay reported in the literature and to find new trends. The fraction of excess sources increases systematically if measured at longer wavelengths. Disk percentages derived using different wavelength ranges should therefore be compared with caution. The dust probed at 22-24 mu m evolves slower than that probed at shorter wavelengths (3.4-12 mu m). Assuming an exponential decay, we derive a timescale tau = 4 : 2 5 : 8 Myr at 22-24 mu m for primordial disks, compared to 2 similar to 3 Myr at shorter wavelengths (3.4-12 mu m). Primordial disks disappear around 10 similar to 20 Myr. Their decline matches in time a brief increase of the number of evolved disks (defined here as including transitional and debris disks). There is more dispersion in the fraction of excess sources with age when measured at 22-24 mu m in comparison to shorter wavelengths. Conclusions. The increase in timescale of excess decay at longer wavelength is compatible with inside-out disk clearing scenarios. The increased timescale of decay and larger dispersion in the distribution of disk fractions at 22-24 mu m suggest that the inner (terrestrial-planet forming) and outer (giant-planet forming) zones evolve differently, the latter potentially following a variety of evolutionary paths. The drop of primordial disks and the coincident rise of evolved disks at 10 Myr are compatible with planet formation theories suggesting that the disappearance of the gas is immediately followed by the dynamical stirring of the disk.