Politecnico di Milano

. In this work we analysed the time series of daily solutions of 4 Italian GPS permanent stations with the aim of investigating the presence of temporal correlations and their impact on the estimation of weekly solution and velocity field precisions. We found that precisions are remarkably lower when temporal correlations are considered; in particular, the mean horizontal precisions of weekly solutions are up to 5 times lower and the horizontal velocity precisions are about 1.5-2 times lower. This topic has 2 relevant applications: the assessment of the quality of a reference system maintenance by GPS permanent stations and the coordinate differences significance test for geodynamical applications.

This work aims at identifying and modelling statistical dependencies between empirical amplification functions of sites in central Italy and the main geological and geophysical characteristics of the region, within a geostatistical analysis framework. The empirical functions, named δS2S, are estimated by decomposing the re siduals of the median predictions of a non-ergodic ground motion model of elastic acceleration response spectra developed for the reference region. To select the model that best describes the spatial variability of the data, the performance of stationary and non-stationary spatial models is compared, the latter being able to constrain the prediction of the empirical functions to physical quantities available in the region and descriptive of the geology, topography and geographical location of the site. Finally, we obtain optimal models of δS2S, for each spectral ordinate, parameterised as a function of geographical coordinates and an input map of shear wave velocity in the upper 30 m (Vs30) constructed ad hoc by combining information gathered from two high-resolution maps available for the region. The methodology allows the development of a new practice-oriented framework for the empirical estimation of site amplification, which can be adopted for the gen eration of shaking scenarios in the context of regional hazard and seismic risk assessment.

In this article, we implement a new approach to calibrate ground‐motion models (GMMs) characterized by spatially varying coefficients, using the calibration dataset of an existing GMM for crustal events in Italy. The model is developed in the methodological framework of the multisource geographically weighted regression (MS‐GWR, Caramenti et al., 2020), which extends the theory of multiple linear regression to the case with model coefficients that are spatially varying, thus allowing for capturing the multiple sources of nonstationarity in ground motion related to event and station locations. In this way, we reach the aim of regionalizing the ground motion in Italy by specializing the model in a nonergodic framework. Such an attempt at regionalization also addresses the purpose of capturing the regional effects in the modeling, which is needed for the Italian country, where ground‐motion properties vary significantly across space. Because the proposed model relies on the italian GMM (ITA18) (Lanzano et al., 2019) dataset and functional form, it could be considered the ITA18 nonstationary version, thus allowing one to predict peak ground acceleration and velocity, as well as 36 ordinates of the 5%‐damped acceleration response spectra in the period interval T=0.01–10s⁠. The resulting MS‐GWR model shows an improved ability to predict the ground motion locally, compared with stationary ITA18, leading to a significant reduction of the total variability at all periods of about 15%–20%. The article also provides scenario‐dependent uncertainties associated with the median predictions to be used as a part of the epistemic uncertainty in the context of probabilistic seismic hazard analyses. Results show that the approach is promising for improving the model predictions, especially on densely sampled areas, although further studies are necessary to resolve the observed trade‐off inherent to site and path effects, which limits their physical interpretation.

In 2014, the Jeddah Municipality made a call for an estimate of a centimetric precision geoid model to be used for engineering and surveying applications, because the regional geoid model available at that time did not reach a sufficient precision. A project was set up to this end and dedicated sets of gravity and Global Positioning System (GPS)/levelling data were acquired in the framework of this project. In this paper, a thorough analysis of these newly acquired data and of the last available Global Gravity Field Models (GGMs) has been done in order to obtain a geoid undulation estimate with the prescribed precision. In the framework of the Remove–Compute–Restore (RCR) approach, the collocation method was used to obtain the height anomaly estimation that was then converted to geoid undulation. The remove and restore steps of the RCR approach were based on GGMs, derived from the Gravity Field and Steady-State Ocean Circulation Explorer (GOCE) and Gravity Recovery and Climate Experiment (GRACE) dedicated gravity satellite missions, which were used to improve the long wavelength components of the Earth’s gravity field. Furthermore, two different quasi-geoid collocation estimates were computed, based on gravity data only and on gravity plus GPS/levelling data (the so-called hybrid estimate). The best solutions were obtained with the hybrid geoid estimate. This was tested by comparison with an independent set of GPS/levelling geoid undulations that were not included in the computed solutions. By these tests, the precision of the hybrid geoid is estimated to be 3.7 cm. This precision proved to be better, by a factor of two, than the corresponding one estimated from the pure gravimetric geoid. This project has been also useful to verify the importance and reliability of GGMs developed from the last satellite gravity missions (GOCE and GRACE) that have significantly improved our knowledge of the long wavelength components of the Earth’s gravity field, especially in areas with poor coverage of terrestrial gravity data. In fact, the geoid models based on satellite-only GGMs proved to have a better performance, despite the lower spatial resolution with respect to high-resolution models (i.e., Earth Gravitational Model 2008 (EGM2008))

Motivated by the crucial implications of Ground Motion Models in terms of seismic hazard analysis and civil protection planning, this work extends a scalar Ground Motion Model for Italy to the framework of Functional Data Analysis. The inherent characteristic of seismic data to be incomplete over the observation domainofoscillation periods entails embedding the analysis in the context of partially observed functional data and performing data reconstruction. This work proposes a novel methodology that accounts for the fact that parts of the curves are directly observed and other parts are reconstructed, thus, characterized by greateruncertainty.Themethoddefinesobservation-specificfunctionalweights,whichentertheestimation process to reduce the impact that the less reliable portions of the curves have on the final estimates. The classical methods of smoothing and concurrent functional regression are extended to include weights. The advantages of the proposed methodology are assessed on synthetic data. Eventually, the weighted func tional analysis performed on seismological data is shown to provide a natural smoothing and stabilization of the spectral estimates of the Ground Motion Model considered. Supplementary materials for this article are available online.

The thermomechanic evolution of the lithosphere–upper mantle system during Calabrian sub- duction is analysed using a 2-D finite element approach, in which the lithosphere is composi- tionally stratified into crust and mantle. Gravity and topography predictions are cross-checked with observed gravity and topography patterns of the Calabrian region. Modelling results indi- cate that the gravity pattern in the arc-trench region is shaped by the sinking of light material, belonging to both the overriding and subduction plates. The sinking of light crustal material, up to depths of the order of 100–150 km is the ultimate responsible for the peculiar gravity signature of subduction, characterized by a minimum of gravity anomaly located at the trench, bounded by two highs located on the overriding and subducting plates, with a variation in magnitude of the order of 200 mGal along a wavelength of 200 km, in agreement with the isostatically compensated component of gravity anomaly observed along a transect crossing the Calabrian Arc, from the Tyrrhenian to the Ionian Seas. The striking agreement between the geodetic retrieved profiles and the modelled ones in the trench region confirms the crucial role of compositional stratification of the lithosphere in the subduction process and the cor- rectness of the kinematic hypotheses considered in our modelling, that the present-day config- uration of crust–mantle system below the Calabrian arc results from trench’s retreat at a rate of about 3 cm yr −1 , followed by gravitational sinking of the subducted slab in the last 5 Myr.

We describe a success story at the junction between South-Eastern Alps and external Dinarides that has led to an early deployment of GPS stations prior to the predicted July 12th 2004 moderate size Slovenia Krn Mountain earthquake. The success story consisted in a straightforward integration between a long-lasting lithosphere-scale rock mechanics experiment, along with GPS monitoring, leading to a physical model of stress evolution and tested earthquake prediction experiment using M8S, CN and RTP algorithms to point out the area of the impending earthquake. Within the alarmed area by the prediction algorithms, the lithosphere-scale rock mechanics experiment revealed that the location of the 2004 event falls within an area of stress shadow due to the recent 1998 Bovec earthquake, but is also very close to an area of increased stress due to the long-lasting effect of the 1511 event. The pre and post 2004 earthquake GPS data provided the following results: 1- the Krn Mountain earthquake magnitude has to be increased from M W 5.2 to 5.5, therefore doubling the fault slip in order to provide a better fit to the near-field displacements. Accordingly the RTP 2004 alarm in Northern Dinarides can be considered a successful prediction now that the magnitude is inside the prediction range. 2- the existence of an important amount of aseismic deformation related to such a moderate size earthquake and the feasibility of monitoring these transients; 3- the evidence of a resolved acceleration of the strain rates one year prior to the earthquake; 4- the robustness of the Bayesian approach in detecting discontinuities in the times series, their magnitude and statistical significance. The discontinuities or jumps in the time series can correspond to coseismic deformation or time-dependent deformation such as creeping, slow motion, strain acceleration and transients in general; 5- when integrated with tested earthquake prediction algorithms, the capability to forecast earthquakes can be extended to the scale of the active fault systems.

Collocation approach has been applied to get a global Moho model in spherical approximation based on a GOCE geopotential model. A simple single layer model, with known density contrast, has been considered and a linearized relationship between the spherical harmonic coefficients of the anomalous potential and those of the Moho depth has been derived. This allows the covariance propagation from gravity to Moho depth. The derived covariance functions are then used in the collocation estimate of the global Moho depth. In order to be as close as possible to the considered model, reductions for the gravity signal related to topography/bathymetry have been applied. Simulated and real data tests have been performed and the obtained global solution has been compared with Moho estimates available in literature. The obtained global Moho has been then used as a starting solution for a regional refinement assuming planar approximation. In this second step the computation has been performed in the Central Mediterranean area, based on collocation, local gravity and topography/bathymetry data.

Recent tomographic investigations performed down to ~300 km depth in the Calabrian Arc region gave insight in favor of the hypothesis that the Ionian subducting slab is continuous in depth beneath the central part of the Arc, while detachment of the deep portion of the subducting structure may have already taken place beneath the edges of the Arc itself. In the present study, we perform new geophysical analyses to further explore the structure of the subduction system and the structure and kinematics of the crustal units in the study area for a more comprehensive view of the local geodynamic scenario. Local earthquake tomography that we address to the exploration of the upper 40 km in the whole region of southern Italy furnishes P-wave velocity domains, suggesting southeast-ward long-term drifting of the southern Tyrrhenian unit with an advancement front matching well with the segment of Calabrian Arc where the subducting slab was found continuous and trench retreat can be presumed to have been active in the most recent times. This scenario of retreating subduction trench inducing drifting of the lithospheric unit overriding the subducting slab is further supported by the analysis of gravity anomalies, allowing us to better constrain the transitional zones between different subduction modes (continuous vs. detached slab) along the Arc. Also, the relocation of recent crustal seismicity, associated with geostructural data taken from the literature, provides evidence for NW-trending seismogenic structures in northeastern Sicily and northern Calabria that we interpret as Subduction-Transform Edge Propagator (STEP) faults guiding the southeast-ward drifting process of the southern Tyrrhenian unit. Crustal earthquake relocations show also seismolineaments in southern Calabria corresponding to the NE-trending longitudinal structures of the Arc where the great shallow earthquakes of 28 December 1908, and 5 and 7 February 1783 occurred. Seismicity and the extensional stress regime detected in these structures find also reasonable location in the proposed scenario, being interpretable in terms of shallow response of the central segment of the Arc to slab rollback and trench retreat.

The analysis of Global Positioning System (GPS) coordinates time series is a valuable tool in quantifying crustal deformations. The longer continuous GPS time series allow estimation of nonlinear signatures. As a matter of fact, besides the linear and periodic behaviors, other relevant signals are present in such time series as the so-called transient deformations. They can be related to, e.g., slow slip events, which play a crucial role in studying fault mechanisms. To give reliable estimates of these signals, an appropriate and rigorous approach for defining the deterministic and the stochastic models of the data is needed. We prove that the theory of the second order stationary random process (SOSRP) can be used to describe the stochastic behavior of the daily GPS time series. In particular, the second order stationarity condition has to be verified for the daily GPS coordinate time series to be described as a SOSRP. This method has been already used for modeling the gravity field of the earth and in predicting/filtering problems, and this work shows that it can also be useful for characterizing the colored noise in the GPS time series.