GNGTS 2023 - Atti del 41° Convegno Nazionale

Session 1.1 GNGTS 2023 Fig. 2: Temporal evolution (V1) and spatial response (U1) of the independent component (IC1) corresponding to the hydrologically-induced deformation of the study site. Seismic observations are centred on the monitoring of the perturbations of the crust mechanical properties in the form of relative seismic-velocity variations ( δ v/v). For the seismic observations we apply coda-wave analysis to crosscorrelation results obtained from ambient-noise recordings (Shapiro et al. 2005), filtered at the [0.1 - 1] Hz frequency-band in order to restrict the noise signal to oceanic microseisms. We produce hour-long crosscorrelation functions and carry out stacks of 30 days to produce daily lapse-time samples. The seismic-coda analysis is carried out by Moving Window Cross-Spectrum analysis (Poupinet et al. 1984, Clarke et al. 2011) and extends over 46s of coda. For the hydrological observations we make use of a mass-balance analysis to estimate the Total Water Storage (TWS) variation corresponding to our control area (upper Tiber basin, see Fig.1). The TWS estimate considers a lumped- hydrological model based on local temperature, potential evapotranspiration and precipitation observations, rainfall-runoff modelling and river flow measurements (Pintori et al. 2021). Fig. 3 shows the comparison of the three time series: geodetic, seismological and hydrological. With a dominant seasonal behaviour, the agreement of the TWS variation with the geodetic time series validates the corresponding independent component to be the isolated source of ground deformation due to water content variations in the area of study. The seismic time series ( δ v/v), features the same seasonal behaviour observed in the former two results, which proves that the strong effect due to hydrological changes in the area extends along the first 3-4 km of crust. Since the δ v/v is displayed with reversed abscissa, this result reveals anticorrelation with the TWS, thus exposing the inverse relation of the seismic-velocity changes with the variation of water content in the ground. Besides, the seismic-velocity perturbation due to strong seismic events away from the study area such as the 2016 seismic sequence in Central Italy, is detectable (green-dashed lines), which makes us presume that the δ v/v signal related to local seismic swarms at the ATF of moderate amplitude are certainly hindered by the dominant hydrological seismic effects.