Broadband Finite Frequency Ambient Noise Tomography: A Case Study in the Western United States Using USArray Stations

Abstract

AbstractWe demonstrate the feasibility of finite frequency ambient noise tomography (ANT) in mapping broadband surface wave phase velocities in the western United States using data from the USArray. We obtain cross‐correlation functions (CCFs) of ambient noise data between pairs of stations and then construct a 2‐D phase sensitivity kernel for each CCF between a pair of stations by treating one station as a “virtual” point source and the other as the receiver. Based on a standard, iterative, linearized inversion technique, we develop a finite frequency ANT method and use this method to construct phase velocity maps. These phase velocity maps are very similar to other reported maps generated by established tomography methods based on ray theory at short periods. At intermediate and long periods (>40 s), phase velocity maps based on finite frequency ANT significantly differ from those based on ray theory ANT; however, they are similar with those based on Helmholtz tomography which uses earthquake surface wave data and considers finite frequency effects of surface waves. Our results suggest that finite frequency effects need to be considered in intermediate‐ and long‐period ANT if one intends to resolve upper‐mantle structures with subwavelength scales using ANT.