The Herschel/PACS view of the Cep OB2 region: Global protoplanetary disk evolution and clumpy star formation

Context. The Cep OB2 region, with its two intermediate-aged clusters Tr 37 and NGC7160, is a paradigm of sequential star formation and an ideal site for studies of protoplanetary disk evolution. Aims. We use Herschel data to study the protoplanetary disks and the star formation history of the region. Methods. Herschel/PACS observations at 70 and 160 mu m probe the disk properties (mass, dust sizes, structure) and the evolutionary state of a large number of young stars. Far-IR data also trace the remnant cloud material and small-scale cloud structure. Results. We detect 95 protoplanetary disks at 70 mu m, 41 at 160 mu m, and obtain upper limits for more than 130 objects. The detection fraction at 70 mu m depends on the spectral type (88% for K4 or earlier stars, 17% for M3 or later stars) and on the disk type (similar to 50% for full and pre-transitional disks, similar to 35% for transitional disks, no low-excess/depleted disks detected). Non-accreting disks are not detected, suggesting significantly lower masses. Accreting transition and pre-transition disks have systematically higher 70 mu m excesses than full disks, suggestive of more massive, flared and/or thicker disks. Herschel data also reveal several mini-clusters in Tr 37, which are small, compact structures containing a few young stars surrounded by nebulosity. Conclusions. Far-IR data are an excellent probe of the evolution of disks that are too faint for sub-millimetre observations. We find a strong link between far-IR emission and accretion, and between the inner and outer disk structure. Herschel confirms the dichotomy between accreting and non-accreting transition disks. Accretion is a powerful measure of global disk evolution: substantial mass depletion and global evolution need to occur to shut down accretion in a protoplanetary disk, even if the disk has inner holes. Disks likely follow different evolutionary paths: low disk masses do not imply opening inner holes, and having inner holes does not require low disk masses. The mini-clusters reveal multi-episodic star formation in Tr 37. The long survival of mini-clusters suggest that they formed from the fragmentation of the same core. Their various morphologies favour different formation/triggering mechanisms acting within the same cluster. The beads-on-a-string structure in one mini-cluster is consistent with gravitational fragmentation or gravitational focusing, acting on very small scales (solar-mass stars in similar to 0.5 pc filaments). Multi-episodic star formation could also produce evolutionary variations between disks in the same region. Finally, Herschel also unveils what could be the first heavy mass loss episode of the O6.5 star HD206267 in Tr 37.