Instantaneous Polarization Attributes in the Time-frequency Domain: Application to Wave Field Separation

In this contribution we propose a method of wave field separation from multicomponent data sets based on the continuous wavelet transform (CWT). We present different approaches for obtaining the time-frequency dependent instantaneous polarization attributes for multicomponent data sets (2C, 3C or more). Using these attributes, we show how to construct filters tailored to separate (filter) different wave types followed by an inverse wavelet transform to obtain the desired wave type in the time domain. The proposed methods are applied on synthetic and experimental data for illustration.