Over the last decade, Distributed Acoustic Sensing (DAS) has gained increased attention as a cost-effective and practical ground motion monitoring technique, due to the unprecedented temporal and spatial sampling possibilities. Considering these inherent advantages, DAS-collected strain rate measurements can be beneficial as addition to traditional particle motion sensors, or potentially replace them under certain circumstances. A DAS system, consisting of an interrogator and attached optical fiber, is conventionally employed in a single (inline) direction providing 1D longitudinal strain rate measurements along the installed fiber. In this thesis, we explore unconventional fiber configurations to retrieve additional wavefield components. To investigate the potential of these unconventional fiber geometries, we perform numerical tests, discuss real data from a field campaign in Yverdon (Switzerland) and report on our analysis of synthetic and field data. Both synthetic and field results show that a loop- and cross fiber lay-out provide thepseudo-divergence, defined as the sum of the local spatial velocity gradients (∂xvx+∂yvy), which is proportional to the full wavefield divergence at the free surface. We confirm that the fiber loop & cross shape predominantly capture Rayleigh surface waves, but are insensitive to Love wave energy. An important factor regarding the loop & cross behaviour is the configuration size with respect to the wavelength (λ). Reduced amplitudes are recognizable for loop & cross dimensions spanning more than 1/5λ, followed by additional phase distortions from 1/3λ and completely altered waveforms for even larger loop & cross sizes.We show that a spherical configuration (composed of 3 independent orthogonal fiberloops) is theoretically attractive by enabling the retrieval of both the full divergence and the individual spatial gradients in the three orthogonal directions. The free surface-relation between the pseudo- and full divergence can be exploited to approximate the local ratio of the shear and pressure wave velocity (VS/VP ), which is demonstrated with synthetic data by means of spectral element modeling. However, the field results are less convincing due to, supposedly, poor cable-ground coupling as a result of required dense cable winding.Finally, we introduce a local approach to estimate the effective shear velocity (VS) without the need for larger arrays. We achieve this by combining local gradients and particle acceleration data, obtained from the fiber loop and one centred 3C-geophone respectively. Our novel approach is validated with synthetic data and we present promising field data results. The proposed 4C-configuration mitigates traditional finite-difference errors related to tilted or inaccurately spaced ground motion sensors. Furthermore, our novel arrangement offers the advantage of azimuthal independence, which can be of interest for VS-estimations from earthquake-, traffic- or potentially ambient noise recordings. The local character of the effective VS-computation is dependent on the incident P-wavelength and is therefore potentially exploitable with regard to 1D shear velocity subsurface models.The presented synthetic and field results in this thesis show the potential of employing unconventional fiber configurations with regard to multi-component gradient sensing. This work further contributes to the development of attractive DAS applications at various scales, especially related to wavefield separation (ground-roll suppression), shallow static corrections and near-surface characterization.

In the summer of 2020, ETH researchers installed a 9 km long fiber-optic cable on the Rhonegletscher (Switzerland), covering the whole glacier extent from accumulation to ablation zone.
The fiber was then interrogated with a DAS system, turning it into a distributed seismic antenna. The DAS system recorded continuously for one month on more than 2000 channels at 1 kHz sampling rate. The large data volume (~18 TB) renders manual event picking and categorization practically unfeasible. For such DAS monitoring experiments, automatic event detection and classification might become indispensable.
Different types of icequakes and meteorological effects are visible on the raw strain-rate recordings. Most cryoseismic events (stick-slip icequakes and surface crevassing) are of short-time duration and small spatial extent. Since most events are only visible on fractions of the array, the data is divided into smaller sub-windows in time and space.
Different array processing techniques are discussed to obtain a low-dimensional feature representation of each sub-window. The output from the array processing is used as input for an unsupervised clustering algorithm assigning a class-membership to each sub-window. This is applied to parts of the Rhonegletscher dataset. Meaningful clusters are returned that corre-
spond to seismic events or noise.
The analysis yields a preliminary overview of signal types contained in the Rhonegletscher dataset and their spatio-temporal distribution. This method can be used to obtain an event catalogue of clustered signals. This could be used as basis for further analyses such as seismic imaging, template matching or supervised machine learning.