Eigenstructure analysis

This thesis aims to address two major problems in geophysical data processing: (a) noise attenuation and (b) spatia-temporal spectral analysis of geophysical wavefields.

In order to enhance the signal-to-noise ratio of multichannel time series I propose methods that exploit the ability of the Singular Value Decomposition to extract coherent information from multi-channel data. In particular, I describe techniques to improve the signal-to-noise ratio of 2D and 3D seismic data sets.

In the second part I developed an algorithm based on the multiple signal classification method to retrieve high resolution estimates of the frequency-wavenumber energy distribution of shear Alfvén waves recorded by the SuperDARN radars network. This method allows us to estimate phase velocity of shear Alfvén waves and retrieve information to constraint physical models that describe field line resonance (FLR) phenomena occurred in the Earth's magnetosphere.