Meeting the Stationary Phase Requirement for Local Ambient Noise Interferometry through Polarization Analysis

Abstract:

High frequency surface waves (1 Hz – 10 Hz) are potentially useful for local scale interferometry in areas with anthropological noise sources such as traffic, machinery or mining and other forms of industrial activity. Successful application requires a homogenous noise source distribution, a condition which is however rarely met in local deployments of short duration.

The objective of this study is to investigate the feasibility of polarization analysis as a pre-processing step to minimize the contribution of noise sources which violate the stationary phase requirement. Polarization analysis applied to a single 3C station can be used to discriminate between body and surface waves and can also inform about the incidence angle and azimuth of the incoming seismic energy. In case of surface wave interferometry, selecting ambient noise windows with a dominant azimuth according to the station pair azimuth can act as a polarization filter with the aim to improve interferograms and dispersion curves.

We test this approach on data from an active mining environment in the Eastern Alps of Austria. 125 3C nodes were deployed in November 2016 for the purpose of active seismic measurements, but the nodes recorded continuously for a period of 3 weeks and were subjected to ambient noise surface wave interferometry. Although the beamforming response shows dominant noise sources to the East and South of the array, the data is also abundant in noise contributed by mining activity (blasting, machinery, vehicles) within the array. These sources are non-stationary, and a lot of station pairs result in poor quality interferograms and dispersion curves. To overcome this problem, we calculate polarization attributes (azimuth, incidence and linearity) to extract surface waves from the continuously recorded data and identify time windows where ambient noise has favorable azimuth regarding the source-receiver pair geometry. Those pre-selected time windows are used for surface wave interferometry and dispersion analysis.

Lastly, we compare our results with conventional surface wave interferometry (without pre-selection) and discuss benefits and limitations of the approach.