Dipartimento di Geoscienze, Università di Padova, Italy.

To investigate the transfer of elastic energy between a regional stress field and a set of localized faults, we project the stress rate tensor inferred from the Italian GNSS (Global Navigation Satellite Systems) velocity field onto faults selected from the Database of Individual Seismogenic Sources (DISS 3.2.0). For given Lamé constants and friction coefficient, we compute the loading rate on each fault in terms of the Coulomb failure function (CFF) rate. By varying the strike, dip and rake angles around the nominal DISS values, we also estimate the geometry of planes that are optimally oriented for maximal CFF rate. Out of 86 Individual Seismogenic Sources (ISSs), all well covered by GNSS data, 78–81 (depending on the assumed friction coefficient) load energy at a rate of 0–4 kPa yr−1. The faults displaying larger CFF rates (4–6 ± 1 kPa yr−1) are located in the central Apennines and are all characterized by a significant strike-slip component. We also find that the loading rate of 75% of the examined sources is less than 1 kPa yr−1 lower than that of optimally oriented faults. We also analysed 2016 August 24 and October 30 central Apennines earthquakes (Mw 6.0–6.5, respectively). The strike of their causative faults based on seismological and tectonic data and the geodetically inferred strike differ by <30°. Some sources exhibit a strike oblique to the direction of maximum strain rate, suggesting that in some instances the present-day stress acts on inherited faults. The choice of the friction coefficient only marginally affects this result.

We develop new approaches to calculating 30-year probabilities for occurrence of moderate-to-large earthquakes in Italy. Geodetic techniques and finite-element modelling, aimed to reproduce a large amount of neotectonic data using thin-shell finite element, are used to separately calculate the expected seismicity rates inside seismogenic areas (polygons containing mapped faults and/or suspected or modelled faults). Thirty-year earthquake probabilities obtained from the two approaches show similarities in most of Italy: the largest probabilities are found in the southern Apennines, where they reach values between 10% and 20% for earthquakes of M W ≥ 6.0, and lower than 10% for events with an M W ≥ 6.5.

Bouguer gravity anomalies in the region of Western Himalayas, Karakoram and Tien Shan show large negative values, but classical isostatic models are insufficient to account for the detailed pattern of the observed anomalies. In the past years the gravimetric surveys in the Karakoram done by Marussi, Caputo and others in 1954 have been extended and intensified. The full body of available gravimetric data, including the pendulum observations by De Filippi and Hedin at the beginning of this century, have been re-analyzed. Terrain corrections have been computed systematically for all available data using a unique algorithm and Digital Terrain Model. The isostatic anomalies along a profile from the Indo-Gangetic foredeep, across the Karakoram range and terminating in the Tarim basin show the oscillating values already noted by Marussi. It is here proposed that this oscillatory pattern can be explained by a model in which the convergent boundaries of the Indian and Tarim plates deform by elastic flexure, besides isostasy. The gravity data constrain the numerical values of the model parameters, particularly the flexural rigidity of the plates. For the Indian plate the best fitting value of the flexural rigidity is D = 5 1024 N m, a value very similar to those reported in Central Himalaya. The flexural rigidity of the Tarim plate turns out to be considerably larger D = 7 1025 N m, which makes the Tarim more rigid than the neighboring Central Tibet. Both plates are loaded by an estimated shear stress of 7 1012 N m-1 located in a region corresponding to the Nanga Parbat Haramosh syntaxis. It is concluded that the Indo-Asian continental collision in the Western Himalaya and Karakoram resulted in the development of flexural basins on both sides, unlike the Central Himalaya where the collision produced a flexural basin, the Ganga basin, to the south and, to the north, the indentation of an isostatically supported Tibetan block with possible rheological layering and eastward lateral extrusion.

Two critical items in the energetic budget of a seismic province are the strain rate, which is measured geodetically on the Earth's surface, and the yearly number of earthquakes exceeding a given magnitude. Our study is based on one of the most complete and recent seismic catalogs of Italian earthquakes and on the strain rate map implied by a multiyear velocity solution for permanent GPS stations. For each of 36 homogeneous seismic zones we use the appropriate Gutenberg-Richter relation, which is based on the seismicity catalog, to estimate a seismic strain rate, which is the strain rate associated with the mechanical work due to a coseismic displacement. We show that for each seismic zone, the volume storing most of the elastic energy associated with the long-term deformation, and hence the seismic strain rate, is inversely proportional to the static stress drop. The GPS-derived strain rate for each seismic zone limits the corresponding seismic strain rate, and an upper bound for the average stress drop is estimated. We show that the implied regional static stress drop varies from 0.1 to 5.7 MPa for catalog earthquakes in the moment magnitude range [4.5–7.3]. The stress drop results are independent of the regional a and b parameters and heat flow but are very sensitive to the assumed maximum magnitude of a seismic province. The data do not rule out the hypothesis that the stress drop positively correlates with the time elapsed after the largest earthquake recorded in each seismic zone.