TL;DR
This paper applies moment tensor inversion techniques, typically used for tectonic earthquakes, to analyze icequakes on the Gornergletscher in Switzerland. They found that most icequakes are caused by tensile crack openings (like surface crevasses), but also identified a rare subset caused by shear faulting and others originating from deeper within the glacier, likely due to water lenses.
What Problem Does This Paper Address?
Understanding the internal processes of glaciers is crucial for predicting their behavior, especially in the context of climate change. While seismic events (icequakes) offer a window into these processes, their source mechanisms have been poorly understood. Previous studies often relied on indirect evidence like hypocentral locations or waveform characteristics. This paper aims to precisely determine the seismic source mechanisms of icequakes using full-waveform moment tensor inversions, a technique that provides detailed information about the forces at play during an earthquake. This allows for a more nuanced understanding of brittle deformation in glacial ice, including the possibility of shear faulting and the role of water.
Methods
The researchers adapted a standard full-waveform moment tensor inversion method, usually applied to moderate-magnitude tectonic earthquakes (Mw > 4), to analyze icequakes (Mw ~ 1.5). They used simple frequency and distance scaling along with Green's functions for a homogeneous half-space to model the seismic wave propagation. They also employed constrained inversion schemes: one assuming a deviatoric source (pure double-couple or CLVD) and another representing a combination of a tensile crack and a double-couple. The quality of the fits was assessed using variance reduction (VR). They carefully selected stations and seismograms to minimize noise and artifacts, considering factors like P-wave arrival polarity and alignment.
Datasets
The study utilized seismic data collected from dense seismic networks deployed on the Gornergletscher, Switzerland, during the summers of 2004 and 2006. This dataset comprised over 80,000 seismic events, part of a larger project investigating the yearly drainage of the nearby Gornersee lake. The paper focuses on analyzing specific clusters of these icequakes, including shallow events, a shallow cluster with atypical waveforms, and events at intermediate depths.
Key Insights and Ideas
- Most icequakes are tensile cracks: The vast majority of icequakes analyzed are consistent with tensile crack openings, likely caused by surface crevasses widening. This explains the dominant Rayleigh waves and compressive first motions observed.
- Shear faulting in ice is possible: A small but significant subset of shallow icequakes exhibited characteristics indicative of shear faulting (double-couple sources). This is a crucial finding, demonstrating that shear failure can occur in glacial ice, not just tensile failure, even without major events like glacier surges.
- Deep icequakes reveal water's role: Icequakes originating at intermediate depths (around 100m) also appear to be tensile crack openings. The authors suggest that the presence of water lenses at these depths reduces effective stress, enabling tensile fracturing, which is otherwise inhibited by high hydrostatic pressure.
- Method adaptation is key: Successfully applying earthquake seismology techniques to icequakes highlights the versatility of moment tensor inversion and the importance of adapting them to the specific characteristics of glacial environments.
Key Results
- Tensile crack openings: Inversions of typical shallow icequakes consistently showed a dominant isotropic component, consistent with tensile crack opening.
- Shear faulting events: A small cluster of shallow icequakes (less than a dozen) were identified as having double-couple sources, indicating shear faulting.
- Intermediate depth events: Three events at ~100m depth were modeled as tensile crack openings, suggesting hydrofracturing facilitated by water lenses.
- Volumetric changes: Tensile crack-type events showed volumetric changes between 100 and 400 cm³, while shear-type events showed negative volumetric changes, likely due to inversion artifacts.
- Moment magnitudes: The calculated moment magnitudes for the analyzed icequakes ranged from approximately -2.3 to -1.5.
Limitations and Future Work
- 1D velocity model: The primary limitation acknowledged is the use of a 1D velocity model for Green's functions. This simplification does not fully account for the glacier's topography or the complex geometry of the glacier bed, which can affect waveform modeling, especially for basal icequakes.
- Free-surface effects: The proximity of shallow sources to the free surface can introduce numerical limitations, potentially affecting the resolution of the isotropic component and leading to artifacts like artificial negative volumetric changes.
- Basal icequakes: The study did not present moment tensor inversions for basal icequakes due to the complexity of modeling reflections from the glacier bed.
- Future work: The authors suggest that future efforts should focus on calculating source parameters for basal icequakes, potentially requiring more advanced 2D or 3D seismic velocity models. They also propose developing procedures to search the entire dataset for shear-type events, particularly those occurring during lake drainage events, to better understand their triggers and implications. Further investigation into the frequency dependence of these source mechanisms is also warranted.
Summary updated