Seismic characterisation
- EinsteinArray and GIANT network — A network of 12 broadband stations installed in boreholes around the candidate area, integrated with meteorological and infrasonic sensors and with the strengthening of the regional Sardinian network.
- Spiral seismic array (Mamone) — A measurement campaign specifically designed to characterise seismic noise in the 1–10 Hz band, the most critical for ET.
- Seismic velocity variations — Analysis of ambient seismic noise cross-correlations to monitor the elastic properties of the shallow crust over time.
- Seismicity catalogue and source discrimination — Construction of a complete catalogue of Sardinian seismicity using machine learning techniques, with identification of anthropogenic events and estimation of moment magnitude through spectral inversions.
Approfondisci
EinsteinArray and strengthening of the Sardinian network (GIANT)
EinsteinArray is a network of 12 broadband seismic stations installed in boreholes at approximately 10 metres depth, with a spacing of 10–15 km around the candidate area. The network allows the study of seismic noise in the microseism band, detection of local seismicity — including quarry blasts — and characterisation of the spatial and seasonal variation of background noise. The array integrates with the four permanent stations already active in the area and with the future regional GIANT network, which will adopt Very Broad Band (VBB) sensors for optimal recording of low-frequency signals such as teleseisms and ocean microseisms. To complete the infrastructure, 6 meteorological stations and 16 infrasonic sensors are planned for a multiparametric environmental characterisation of the site.
Spiral seismic array for the 1–10 Hz band
The frequency range between 1 and 10 Hz, crucial for assessing site suitability against ET requirements, still lacks adequate characterisation. A 16-element array in spiral configuration has therefore been designed for installation in the area of the "Mamone" penal colony (Province of Nuoro). The spiral geometry provides excellent resolution in the direction of arrival and propagation velocity of seismic waves with a reduced number of sensors. Installation is planned for March–April 2026, with an acquisition period of approximately two months.
Geometry of the seismic array installed at the Mamone Penal Colony
Researchers working on the installation of a seismic station on the array
Monitoring of seismic velocity variations
Through an interferometric technique applied to ambient seismic noise cross-correlations, it is possible to construct time series of seismic wave velocity variations in the shallow crust. These variations reflect perturbations of the stress state, variations in the water content of the subsurface or effects of extreme meteorological events, and can be integrated with GNSS deformation measurements to estimate expected movements at the depths of the Sos Enattos mine tunnels. The analysis is conducted with the open source software MSNoise (http://www.msnoise.org/), ensuring full reproducibility and comparability with other ET candidate sites.
Seismicity catalogue and source discrimination
Between 2022 and 2025, the integration of 8 stations from the AdriaArray experiment with the INGV permanent network (IV) and the MedNet network (MN) allowed for the first time a homogeneous coverage of the entire Sardinia with inter-station distances below 30 km. These data feed an automated machine learning-based workflow for the construction of a complete catalogue of local seismicity comprising phase identification with PhaseNet, association of arrivals into events with Gamma, probabilistic location with NonLinLoc and calculation of moment magnitude Mw using a probabilistic spectral inversion method.
Given the very low level of tectonic seismicity in Sardinia, a significant proportion of the events in the catalogue are of anthropogenic origin — in particular quarry explosions. Machine learning techniques applied directly to waveforms allow these sources to be separated from natural ones, exploiting the reduced surface wave content typical of shallow explosions.
Map of the events (earthquakes and quarry explosions) that have been located and for which the Moment Magnitude Mw has been calculated, represented by the color of the circles
Electromagnetic characterisation
- CARMA-ET — Magnetotelluric campaign for the spatial and spectral characterisation of ambient magnetic noise and estimation of induced telluric currents.
- Audiomagnetotelluric surveys (AMT) — Reconstruction of the electrical resistivity structure of the subsurface down to 2 km depth.
- Airborne electromagnetic prospecting (EinstAEM) — 3D modelling of subsurface resistivity and fracturing down to 500 m depth over approximately 400 km².
- Continuous magnetic monitoring — Permanent magnetic station for continuous monitoring of the geomagnetic field.
Approfondisci
CARMA-ET: characterisation of ambient magnetic noise
IThe CARMA-ET project is dedicated to measuring and understanding magnetic noise at the Sos Enattos site. The area of interest extends over approximately 25x25 km² and includes several potential sources of magnetic disturbance, both anthropogenic — solar panels, wind farm, industrial areas, dam — and natural, related to geomagnetic field variability and solar activity. These sources can produce direct disturbances and generate telluric currents induced in the subsurface, which represent one of the possible sources of interference for ET infrastructure and measurement systems.
The project involves a magnetotelluric measurement campaign with Long Period stations, for the study of slow and large-scale variations, and Broad Band stations, dedicated to the characterisation of high-frequency disturbances, at four extreme points of the ET area plus a remote reference station. Results will allow mapping of the spatial and spectral distribution of magnetic noise, estimation of induced telluric currents, and provision of fundamental indications for infrastructure layout optimisation, noise mitigation strategies and environmental monitoring systems. Estimated total duration: 3–4 months.
Field activities for ground electromagnetic measurements: sensor installation, electrode deployment, and setup of solar-powered acquisition equipment.
Audiomagnetotelluric surveys (AMT)
The AMT technique uses the measurement of natural variations in electric and magnetic fields to reconstruct the electrical resistivity structure of the subsurface down to approximately 2 km depth. In the Sos Enattos area, the resistivity distribution allows identification of structural discontinuities, faults, fractured zones and preferential pathways for fluid circulation — fundamental elements for the design of ET's underground infrastructure. Measurements are organised along 2D profiles in the directions of the planned infrastructure, with orthogonal transects to intercept lateral variations.
Airborne electromagnetic prospecting (EinstAEM)
The EinstAEM project involves airborne electromagnetic prospecting over approximately 450 km², with 2,800 km of data acquired from a helicopter-borne platform. The objective is to reconstruct three-dimensional models of subsurface resistivity and chargeability down to 500 m depth, with a lateral resolution of 30–50 m. The data will allow identification of fracture zones, areas with preferential fluid circulation and lithological contacts relevant to underground works. Activities officially commenced in February 2026.
The airborne electromagnetic (AEM) system antenna suspended beneath a helicopter during a geophysical survey. The rigid loop carries the transmitter and receiver coils used to remotely sense the subsurface.
AEM system take-off: a technician guides the operation as the helicopter lifts the antenna from the ground, ready to begin the survey.
The AEM antenna resting on a paddock during an operational break. The helicopter is parked nearby while grazing cattle provide an unusual backdrop to the geophysical equipment.Continuous magnetic monitoring
To complete the electromagnetic field analysis, a permanent magnetic station is planned that will form part of the National Magnetic Network managed by INGV. A low-noise triaxial fluxgate magnetometer will continuously monitor the vector components of the geomagnetic field, alongside a scalar sensor. The data will allow characterisation of the subsurface response to external magnetic field variations, estimation of induced telluric currents and construction of long-period time series for statistical estimation of magnetic noise levels, in synergy with CARMA-ET and AMT results.
Ground magnetometer installation for measuring the Earth's magnetic field in the study area (Mamone, Sardinia)
Geodesy and ground deformation
- GNSS densification network — 8–10 new semi-permanent stations for high-precision measurement of ground deformations.
- Multi-temporal SAR interferometry (MT-InSAR) — Generation of sub-centimetre ground deformation maps from COSMO-SkyMed and Sentinel-1 satellite imagery.
Approfondisci
GNSS densification network
Although Sardinia is tectonically stable, the extraordinary sensitivity of ET requires detailed knowledge of any ground deformation, even of hydrological or atmospheric origin. The project involves the installation of 8–10 new semi-permanent GNSS stations around the ET area, with multifrequency and multi-constellation receivers powered by solar energy. The network will measure deformations with daily strain resolution down to 10-7, distinguishing hydrological loading from tectonic movements and providing calibration for InSAR maps.
Tripod supporting the GNSS antenna of the SOSE station.
Installation of the SOSE GNSS station, part of the RING network (http://webring.gm.ingv.it) at Sos Enattos
Complete SOSE GNSS station with solar power supply
Semi-permanent low-impact GNSS station with solar power supply.Multi-temporal SAR interferometry (MT-InSAR)
The MT-InSAR technique measures millimetric displacements of the Earth's surface using historical series of satellite images acquired with Synthetic Aperture Radar (SAR) sensors. For Sos Enattos, approximately 130 images acquired by the COSMO-SkyMed satellite of the Italian Space Agency (X-band SAR sensors, resolution ~3x3 m² on the ground) are processed, covering January 2022 to December 2025, in both ascending and descending orbits. 88 Sentinel-1 images (ESA, C-band) were also processed for September 2022–July 2025 using the Small Baseline Subset (SBAS) technique, with a ground resolution of 100 m. Final products are the average ground velocity map (mm/year) and displacement time series (mm); COSMO-SkyMed and Sentinel-1 results will be cross-validated.
The validated final products — average deformation velocity maps and displacement time series with precision of 1–2 mm/year — will allow identification of geological criticalities such as landslides and subsidence, and will be integrated with GNSS data to reconstruct the continuous displacement field of the area.
Ground deformation velocity map derived from MT-InSAR (Multi-Temporal Interferometric SAR) analysis, draped over a 3D digital terrain model. Each coloured point represents a stable radar reflector (building, outcrop) whose elevation is tracked over time by the satellite. Green tones indicate stability, while warm colours (yellow, orange, red) highlight subsidence — a slow sinking of the ground surface.
Artistic rendering of the COSMO-SkyMed radar satellite in orbit around the Earth. Developed by the Italian Space Agency (ASI), this system acquires high-resolution radar imagery in any weather and lighting condition, and is used among other applications for ground deformation monitoring.
Ground displacement time series for a single measurement point (lat. 37.98°N, lon. 23.60°E), derived from MT-InSAR analysis. Each black dot represents a satellite acquisition; the red line shows the average trend, corresponding to a subsidence rate of approximately −6.9 mm/year. The plot reveals a steady, progressive sinking of the ground surface over approximately five years.
Structural geology and 3D modelling
3D geological model and Discrete Fracture Network (DFN).
Approfondisci
3D geological model and Discrete Fracture Network (DFN)
The objective of this line of activity, conducted by the SISMOLab3D of INGV, is the construction of an integrated 3D geological-geophysical model of the crystalline basement of the candidate area, defining structural geometries and rock volumes and providing the framework necessary for all site geophysical analyses. This is complemented by the construction of Discrete Fracture Network (DFN) models, which explicitly represent the fracturing of the medium in order to analyse its hydraulic behaviour.
The work consists of systematic structural surveys at surface and in the accessible tunnels of the Sos Enattos mine, statistical analysis of fracture systems, 3D modelling and numerical simulations of fluid migration. The connectivity and permeability parameters obtained will be integrated into the regional 3D model, calibrated with AEM resistivity data, and made available for the design of ET's underground works.
Preliminary 3D geological model of the study area. The topographic surface is integrated with local geological mapping. Cylinders represent borehole locations and depths, while blue lines indicate one of the possible layouts for the Einstein Telescope (ET) infrastructure. The red surface highlights the main fault, representing the dominant structural element within the investigated crustal volume.
