Tilting and Horizontal Movement at and across the Northern Border of the Adria Plate

We report on the analysis of the seven permanent GPS stations of the recently installed Friuli Regional Deformation Network (FReDNet) of NE-Italy for the time period that goes from the moment of installation (between June 2002 and July 2003, according to the station) to 31 December 2004. We first make a comparison of two available solutions, calculated by the Agenzia Spaziale Italiana (ASI, Centro di Geodesia Spaziale "G. Colombo") and by the Centro Ricerche Sismologiche (CRS, Istituto Nazionale di Oceanografia e Geofisica Sperimentale). The rms (root mean square) difference between the two solutions results to be between 4 and 8 mm, depending on the component. The greatest difference is found for the vertical component, the best agreement is found for the NS component. We then proceed with the analysis of the time series, using the solutions of CRS, as this agency has longer time series available. Our attention goes to three classes of signals, which is the high frequency component (several days to weeks), the seasonal and annual signal, and the interpolated linear trend, representative of the average velocity of the monuments. The mutual correlation coefficients of the stations for the same component are between 0.3 and 0.7, with the highest values found for the vertical component, which demonstrates that there is a common signal, which is in phase. The correlation coefficient does not show a distance relation of the stations, which indicates that the origin of the common signal could be due to pressure loading or an incomplete atmospheric correction. In order to determine reliable velocities, we determine the sinusoidal annual and semi-annual oscillations, which are present in all stations. The amplitudes of the yearly oscillation vary between 0.3 mm and 1.4 mm for the horizontal components, and are definitely larger for the vertical component (between 2 and 5 mm). Those of the semi-annual period range between 13%1 and 95% of the annual amplitude. At last we determine the relative velocities of all stations with respect to the Trieste station, which can be considered tectonically stable, as it is located on the aseismic Karst. We find that the pedemountain stations have a relative southward motion of about 2-3 mm/year and an uplift rate of 1-2 mm/year. The stations on the Friuli plane have nearly no horizontal movement relative to Trieste, and show a relative subsidence. The observations are in agreement with the expectations from neotectonics, which give subsidence of the Friuli plane and uplift of the pedemountain area. They also agree with north-westward tilting observed with the long-base tiltmeters of the Grotta Gigante station (Trieste Karst). The southward movement of the pedemountain stations fits the picture of the relative movement of the Eurasian and Adria plates, the Trieste and Friuli-plane stations belonging to the Adria plate, the remainder stations belonging to the Eurasian plate.