Main content

Use the navigation on the left to browse through our publications of the past few years.

Most recent publications

Accordion. Press Tab to navigate to entries, then Enter to open or collapse content.


Transpression within contractional fault steps

Figure 7 of Nabavi et al. (2017)  
Minimum compressive stress (σ3) distribution for overlapping fault steps under a convergence angle of 60° with the minimum values at the fault tips. Low-σ3 zones develop in the tensional quadrants outside the contractional fault step.

Analysis of transpression within contractional fault steps using finite-element method

Nabavi S.T., Alavi S.A., Mohammadi S., Ghassemi M.R. and Frehner M.

Journal of Structural Geology 96, 1–20

Go to article


Two-dimensional finite-element modelling of elastic Newtonian rheology is used to compute stress distribution and strain localization patterns in a transpression zone between two pre-existing right-stepping, left-lateral strike-slip fault segments. Three representative fault segment interactions are modelled: underlapping, neutral, and overlapping. The numerical results indicate that at the onset of deformation, displacement vectors are oblique to the regional compression direction (20–90°). The orientations of the local σ1 (the maximum compressive stress) and σ3 (the minimum compressive stress) directions strongly depend on the structural position within the transpression zone. For neutral and overlapping fault steps, there is a contractional linking damage zone between the fault segments. For overlapping faults, the σ1 trajectories within the transpression zone deflects significantly forming a sigmoidal pattern, which is created by two rotational flow patterns close to the fault tips. These flow patterns are related to friction effects and different shear deformation, from pure shear outside of the fault steps toward simple shear along the fault segments. Interaction between the two fault segments perturbs the stress field and reflects the heterogeneous nature of deformation. A lozenge- (for underlapping steps), rhomboidal- (for neutral steps), and sigmoidal-shaped (for overlapping steps) transpression zone developed between the two segments. The modelled mean stress pattern shows a similar pattern to that of the contractional steps, and decrease and increase in underlapping and overlapping fault steps, respectively. Comparison of the Kuh-e-Hori transpression zone, between the Esmail-abad and West Neh left-stepping right-lateral strike-slip fault segments in SE Iran, with the modelling results shows strong similarities with the neutral step configuration.

3D model of the Panixer Pass Transverse Zone

Figure 7 of von Däniken & Frehner (in press)  
Birds-eye view of the 3D structural model from NW (a) and SE (b), as well as view from the top (c) and bottom NW (d). In c and d, some units are removed to have an unobstructed view to some details of the model.

3D structural model and kinematic interpretation of the Panixer Pass Transverse Zone (Infrahelvetic Complex, eastern Switzerland)

von Däniken P. and Frehner M.

Swiss Journal of Geosciences 110 (2017), 653–675

Go to article


The Panixer Pass Transverse Zone in the eastern Swiss Alps is oriented perpendicular to most alpine structures in the area. Its main element is the SSE-trending Crena-Martin Fold, a downward facing fold with Permian Verrucano in its core, which is cut by the Glarus Thrust. Hence Verrucano can be found below the Glarus Thrust in the Infrahelvetic Complex. Across the Panixer Pass Transverse Zone the structural buildup of the Infrahelvetic Complex changes considerably. Multiple published theories of the structural evolution are not satisfying particularly because traditional 2D geological cross-sections are insufficient to understand the 3D complexity. The main result and product of our study is a 3D structural model of the Panixer Pass Transverse Zone providing insight into its geometry. As modeling input, we produced a lithostratigraphic map and collected structural orientation data. The 3D structural model honors the observed surface geology and the expected 3D subsurface geometry. Our field data indicates that the shearing and transport direction was continuously NNW-directed, except for a phase of north-directed shearing during the early movement along the Glarus Thrust (late Calanda Phase) and related foliation development in the Helvetic Nappes. The Panixer Pass Transverse Zone developed prior to the penetrative foliation during a thrust-dominated deformation phase (Cavistrau Phase), for which we created a kinematic block model. According to this model, the Crena-Martin Fold is the result of multiple lateral ramps and related lateral fault-bend folds that all developed in a similar positon. In particular, we do not propose ENE-WSW-directed shortening to form the Crena-Martin Fold. The latter was finally cut at low angle by a dextral strike-slip fault to create the final geometry of the Panixer Pass Transverse Zone. Our kinematic model reproduces the main features of the 3D structural model and embeds well into previously proposed sequences of deformation phases.

Frequency scaling of seismic attenuation

Figure 1 of Chapman et al. (2017)  
(a) Frequency-dependent attenuation for wave-induced gas exsolution dissolution (WIGED), for mesoscopic wave-induced fluid flow (WIFF), and for a standard-linear solid (SLS) model. At frequencies beyond the attenuation peak, the scaling characteristics are different for the different models. (b) Schematic illustrations of fluid distribution for periodically alternating fluid layers (WIFF) and for homogeneous distribution of pore scale gas bubbles (WIGED).

Frequency scaling of seismic attenuation in rocks saturated with two fluid phases

Chapman S., Quintal B., Tisato N. and Holliger K.

Geophysical Journal International 208 (2017), 221–225

Go to article


Seismic wave attenuation is frequency dependent in rocks saturated by two fluid phases and the corresponding scaling behaviour is controlled primarily by the spatial fluid distribution. We experimentally investigate the frequency scaling of seismic attenuation in Berea sandstone saturated with two fluid phases: a liquid phase, water, and a gas phase, air, carbon dioxide or nitrogen. By changing from a heterogeneous distribution of mesoscopic gas patches to a homogeneous distribution of pore scale gas bubbles, we observe a significant steepening of the high-frequency asymptote of the attenuation. A transition from one dominant attenuation mechanism to another, from mesoscopic wave-induced fluid flow to wave-induced gas exsolution dissolution (WIGED), may explain this change in scaling. We observe that the high-frequency asymptote, for a homogenous pore scale gas bubble distribution, scales in accord with WIGED.

Temperature-dependent elasticity

Figure 8 of Bakker et al. (2016)  
Pressure change for the removal of a 200 m thick ice cap. A) Temperature-independent model; B) Temperature-dependent model; C) Difference between the two.

How temperature-dependent elasticity alters host rock/magmatic reservoir models: A case study on the effects of ice-cap unloading on shallow volcanic systems

Bakker R.R., Frehner M. and Lupi M.

Earth and Planetary Science Letters (2016), 16–25

Go to article


In geodynamic numerical models of volcanic systems, the volcanic basement hosting the magmatic reservoir is often assumed to exhibit constant elastic parameters with a sharp transition from the host rocks to the magmatic reservoir. We assess this assumption by deriving an empirical relation between elastic parameters and temperature for Icelandic basalts by conducting a set of triaxial compression experiments between 200 °C and 1000 °C. Results show a significant decrease of Young's modulus from ∼38 GPa to less than 4.7 GPa at around 1000 °C. Based on these laboratory data, we develop a 2D axisymmetric finite-element model including temperature-dependent elastic properties of the volcanic basement.

As a case study, we use the Snæfellsjökull volcanic system, Western Iceland to evaluate pressure differences in the volcanic edifice and basement due to glacial unloading of the volcano. First, we calculate the temperature field throughout the model and assign elastic properties accordingly. Then we assess unloading-driven pressure differences in the magma chamber at various depths in models with and without temperature-dependent elastic parameters. With constant elastic parameters and a sharp transition between basement and magma chamber we obtain results comparable to other studies. However, pressure changes due to surface unloading become smaller when using more realistic temperature-dependent elastic properties. We ascribe this subdued effect to a transition zone around the magma chamber, which is still solid rock but with relatively low Young's modulus due to high temperatures. We discuss our findings in the light of volcanic processes in proximity to the magma chamber, such as roof collapse, dyke injection, or deep hydrothermal circulation. Our results aim at quantifying the effects of glacial unloading on magma chamber dynamics and volcanic activity.

Krauklis-wave resonance in fractures

Figure 4 of Shih & Frehner (2016)  
Spectrograms of receiver time signals generated by an S-wave propagating through the intact sample (a) and through fractured samples with a fracture inclination angle of 30°, 45°, and 60° (b-d).

Laboratory evidence for Krauklis-wave resonance in fractures and implications for seismic coda wave analysis

Shih P.-J.R. and Frehner M.

Geophysics 81 (2016), T285–T293

Go to article


Krauklis waves are of major interest because they can lead to resonance effects in fluid-filled fractures. This resonance is marked by seismic signals with a dominant signature frequency, which may reveal fracture-related rock properties. In our laboratory study, we used homogeneous Plexiglas samples containing a single well-defined (i.e., manufactured) fracture. We recorded the signals obtained from propagating ultrasonic P- and S-waves (source frequency: 0.6, 1, and 2.25 MHz) along a sample without a fracture and samples with a fracture with different inclination angles of 30°, 45°, and 60° with respect to the short axis. The experimental results obtained from an incident S-wave confirmed that the presence of the fracture led to resonance effects at frequencies lower than the dominant source frequency, which slowly decayed over time in the recorded seismic coda after the first arrival. The resonance frequency was independent of the fracture orientation and the source frequency. We have interpreted this narrow-banded coda signal as a resonance in the fracture, and the frequency at which this occurred was an intrinsic property of the fracture size and elastic properties. To verify our laboratory results, we used an analytical solution, which provided a relationship between the fracture width, fracture length, resonance frequency, and temporal quality factor (i.e., exponential decay over time). The temporal quality factor obtained from our laboratory data agreed very well with the analytical solution. Hence, we concluded that the observed signature frequency (approximately 0.1 MHz) in the seismic coda was indeed a resonance effect. Finally, we have developed possible applications on the reservoir scale to infer fracture-related properties based on seismic coda analysis.

Page URL:
Thu Jul 20 14:48:10 CEST 2017
© 2017 Eidgenössische Technische Hochschule Zürich