Syn‐convergent extension observed using the RETREAT GPS network, northern Apennines, Italy

We present crustal deformation results from a geodetic experiment (Retreating‐Trench, Extension, and Accretion Tectonics (RETREAT)) focused on the northern Apennines orogen in Italy. The experiment centers on 33 benchmarks measured with GPS annually or more frequently between 2003 and 2007, supplemented by data from an additional older set of 6 campaign observations from stations in northern Croatia, and 187 continuous GPS stations within and around northern Italy. In an attempt to achieve the best possible estimates for rates and their uncertainties, we estimate and filter common mode signals and noise components using the continuous stations and apply these corrections to the entire data set, including the more temporally limited campaign time series. The filtered coordinate time series data are used to estimate site velocity. We also estimate spatially variable seasonal site motions for stations with sufficient data. The RMS scatter of residual time series are generally near 1 mm and 4 mm, horizontal and vertical, respectively, for continuous and most of the new campaign stations, but scatter is slightly higher for some of the older campaign data. Velocity uncertainties are below 1 mm/yr for all but one of the stations. Maximum rates of site motion within the orogen exceed 3 mm/yr (directed NE) relative to stable Eurasia. This motion is accommodated by extension within the southwestern and central portions of the orogen, and shortening across the foreland thrust belt to the northeast of the range. The data set is consistent with contemporaneous extension and shortening at nearly equal rates. The northern Apennines block moves northeast faster than the Northern Adria microplate. Convergence between the Northern Apennines block and the Northern Adria microplate is accommodated across a narrow zone that coincides with the northeastern Apennines range front. Extension occurs directly above an intact vertically dipping slab inferred by previous authors from seismic tomography. The observed crustal deformation is consistent with a buried dislocation model for crustal faulting, but associations between crustal motion and seismically imaged mantle structure may also provide new insights on mantle dynamics.