Efficient array partitioning method for F-J transform to extract multimode dispersion curves from ambient noise data
The recently developed F-J transform method relies on the array configuration of seismic stations,which directly affects the imaging results of the velocity structure beneath a specific study area.When the underground structure of the study area varies gently and the station spacing is relatively uniform,it is sufficient to simply move a fixed-size rectangular box gradually along the latitude and longitude.However,if the subsurface structure of the study area is complex and variable,especially when the station spacing is uneven,the array needs to be divided flexibly according to the specific geographical features.This paper introduces an array partitioning method suitable for regions characterized by complex subsurface structures and non-uniform station spacing,exhibiting a certain level of generality.By applying this method to the Alaskan and northwestern Canadian regions,we have validated that our partitioning approach significantly enhances the resolution of the results compared to the simplistic latitude and longitude-based moving array method.This improvement provides a solid foundation for subsequent inversion studies aiming to investigate the three-dimensional velocity structure of the research area.