Spatial-Domain Synchrosqueezing Wavelet Transform and Its Application to Seismic Ground Roll Suppression

High-precision time-frequency (TF) analysis (TFA) based on the synchrosqueezing wavelet transform can improve the TF resolution and sharpen TF representations (TFRs). However, this method works only for time-varying signals and cannot characterize spatially varying signals with unique spatial properties. Herein, we propose extending the time-domain synchrosqueezing wavelet transform (TSWT) to the spatial domain, yielding the spatial-domain synchrosqueezing wavelet transform (SSWT), and we introduce a velocity parameter. First, we apply the proposed SSWT to several synthetic signals to test its feasibility. From such experiments, we observe that the SSWT is capable of characterizing non-stationary, spatially varying signals with high resolution and robustness, characteristics that are inherited from the TSWT. The SSWT can also maintain its signal reconstruction ability. Furthermore, we apply the SSWT to suppress ground roll in seismic data processing. Through examples of both synthetic datasets and field datasets, we conclude that the SSWT can accurately characterize spatially varying signals and could have many potential applications in the field of signal analysis and processing.