Section 2.1: Physics of Earthquakes and Volcanoes


Our research covery a broad range of topics and contributes mainly to the GFZ coordinated research on Natural Hazards (research unit RU4), plate boundary systems (research unit RU2) and Resources (research unit RU5).

The kinematics of rupture and slip for large earthquakes can be resolved if a combination of near and far field broadband multi-sensor data is included in modern rupture inversion schemes. With the improved quantity and quality of such data we tested and applied in-house methods to a set of large earthquakes at different plate boundaries, including modern approaches of high frequency rupture tracking, rapid kinematic and slip inversions, automatic aftershock centroid moment tensors, and ground motion prediction equations. Recent applications are illustrative of the variability of rupture processes for different earthquakes. In contrast to the simple, smooth rupture of the Mw 8.2 Illapel (Chile) event, the Mw 8.1 Iquique earthquake, which took place along the same plate boundary, had a very heterogeneous rupture speed and slip rate. Combining information from main and aftershocks we suggest a novel inversion approach for the complex rupture history of the Mw 7.8 2016 New Zealand earthquake, which ruptured several crustal and near surface faults with a large impact on the distribution of areas affected by near field ground shaking and landslides. Accurate seismic source models are helpful to better assess the ground-shaking as a major component of seismic hazard evaluation.

The better forecast of aftershocks, or more generally, of earthquakes, which are locally triggered or induced by stress transients, is an important ongoing task in time dependent seismic hazard. Independent of the development of purely statistical, operational, time-dependent models we need to improve our physical understanding of earthquake triggering. The triggering process is very complex and is influenced by coseismic (static and dynamic) stress changes as well as transient aseismic processes. We test and compare purely statistical and physics-based static trigger models and their forecast uncertainties for several case applications, including loading scenarios from postseismic stress build-up, fluid/magma intrusions, aseismic creep or rupture nucleation phases.

Earthquakes triggered and induced by human activities have been receiving global attention over the last few years. We target different key questions on induced seismicity, such as the occurrence of larger and felt events and the discrimination and characterization of induced seismicity. The development of control and management systems of anthropogenic seismicity requires the ability to monitor and analyze weak and microseismic events at shallow depth. We improved our monitoring and processing tools by developing automatic, waveform-based detection, location and source inversion methods. The analysis of non-double couple components of such weak earthquakes is quite challenging, but helps to understand the physical processes behind induced seismicity. We have also used physics-based trigger models for the first time to develop a probabilistic discrimination scheme for depletion-induced earthquakes at gas and oil reservoirs.

Volcanoes are complex, multi-hazard systems involving different processes over large temporal and spatial scales. We focus on volcano physics, develop and apply innovative methods for volcano monitoring and study volcanism and processes of magma migration in the underground, and the hazards associated with such activity. We investigate processes culminating in eruptions, and different types of volcanism, such as frequent and small-scale steam driven eruptions at dome building volcanoes, rare but large-scale explosive eruptions and caldera formation, and effusive eruptions and rifting events.

Monitoring methods include traditional ground based techniques, such as seismicity, tilt and GPS monitoring, and also novel techniques such as radar interferometry (InSAR), photogrammetry, pixel tracking, and drone based sampling.

With numerical simulations we explain the data, the long term effect of surface loading, landslides and sector collapses, or crustal magma pathways. We also model the deformation caused by pressurized conduits and magma reservoirs. The computer based models are extended by analogue lab experiments to better understand the controls on the dynamics of magma and fluids in the Earth’s crust and associated faulting and deformations.

We also respond to volcanic crises in our field of expertise, communicate with the media, and send experts to monitor activity and for assistance in hazard analysis.