Obtaining reliable estimates of the subsurface state distributions is essential to determine the location of e.g. potential nuclear waste disposal sites. However, providing these is challenging since it requires solving the problem numerous times yielding high computational cost. To overcome this, we use a physics-based machine learning method to construct surrogate models. We demonstrate how it produces physics-preserving predictions, which differentiates it from purely data-driven approaches.

In this manuscript, we investigated the kinematics and dynamics of strike-slip damage zones using laboratory mechanical tests and analogue modelling techniques. The results underline the importance of a multi-scale approach (from crustal to outcrop-scale) to improve the understanding of such deformation processes, deriving fundamental correlations with the physical and mechanical properties of the model materials applied in the experiments.

West Norwegian 'diamond facies' eclogite contains tiny mineral inclusions of quartz and amphibole lamellae that are not stable in the diamond field. Low trace amounts of water in the lamellae-bearing host minerals suggest that the inclusion microstructure was not formed by fluid infiltration but by dehydration during early exhumation of these rocks. Some samples with higher water content argue that a late fluid overprint was spatially restricted and obliterated evidence of extreme metamorphism.

At any scale, the limited size of a study area introduces a bias in the interpretation of linear features, defined as right-censoring bias. We show the effects of not considering such bias and apply survival analysis techniques to obtain unbiased estimates of multiple parametrical distributions in three censored length datasets. Finally, we propose a novel approach to select the most representative model from a sensible candidate pool using the Probability Integral Transform technique.

Our study investigates the statistical relationship between geological faults and earthquakes in the Southwestern Swiss Alps. We analyze how the fault size and earthquake rupture are related and find differences in how faults at different depths rupture seismically. While shallow faults tend to rupture only partially, deeper faults are more likely to rupture along their entire length, potentially resulting in larger earthquakes.

We assessed the poro-perm relations of both host rocks and fault rocks of Mesozoic carbonate rocks, by integrating a laboratory petrophysical studyn with a digital image analysis. Three different protocols were employed to compute permeability: i) Effective Medium Theory on laboratory data, ii) constant crack aperture and iii) crack density values from 2D images. Carbonate host rocks did not show a clear poro-perm trend due to the presence of stiff, sub-rounded pores and of small vugs.

We investigate geophysical properties and the distribution of vertical stress, which is defined by the weight of the rock column above a certain location in the subsurface, in the upper 5 km of the North Alpine Foreland Basin in Germany. Our results help to understand the present-day geological configuration and to improve safety for subsurface use, such as deep geothermal energy production in the study area.

In this study, we present an innovative method to study the inner structure of ore deposits using seismic waves produced by earthquakes and production blasts. Results of numerical simulations and field tests show that the proposed method can effectively detect mineralization zones inside orebodies.

For years, mankind has heated sapphire to increase its blue color and value. Gemologists have struggled for decades to detect heat-treated sapphire. Each test costs money. High-tech instruments like FTIR are costly. As a result, luminescence under SWUV and LWUV light provides a cheaper and more practical technique for identifying heat-treated sapphire. This work highlights that blue luminescence under SWUV light could indicate heated sapphire, whereas purplish red may also be helpful.

We performed a paleoseismological study on the Mt. Morrone active fault, which constitutes one of the most problematic structures of the central Apennines in terms of large-earthquake probability. Information on historical seismicity is relatively few and the paleoseismological record is limited only to the northern portion of the fault. We investigated four trenches and we recovered the occurrence of three events over the last 5–6 kyrs, suggesting an average recurrence interval of 1800 years.

Several crosshole seismic surveys in the undisturbed Rotondo granite are used to analyze the seismic anisotropy in the BedrettoLab in the Swiss alps. The P- and S1-waves show a clear trend of faster velocities in NE-SW direction and slower velocities perpendicular to it. This pattern describes a tilted transverse isotropic velocity model. The symmetry plane is mostly aligned with the direction of maximum stress but also the orientation of fractures are expected to influence the wave velocities.