Array analysis of <i>P</i>-wave polarization data for constraining near-surface shear wave speeds: Application to the Valley of Mexico seismic network

SUMMARY We present an array method for constraining near-surface shear wave speed values from P-wave polarization data. As the direction of P-wave particle motion is determined both by the incident ray parameter and the local surficial shear wave speed, the latter can be inferred from three-component seismic records if the ray parameter is estimated first. Previous studies have analysed data recorded at a single station from various earthquakes in order to estimate near-surface shear wave speed. In contrast, this paper analyses data recorded at an array of stations. As such, this method can estimate both ray parameter data for the recorded earthquakes and near-surface shear wave speed values at the studied sites. And as an advantage, the method can analyse records of local and regional earthquakes where ray parameter information cannot be reliably estimated using 1D reference velocity models. Additionally, by processing polarization data spanning multiple events and stations, our method provides robust estimates for near-surface shear wave speed values. We apply this array-based method to a seismic array covering the Valley of Mexico. Analysed records correspond to regional earthquakes with epicentral distances of roughly 300 km. The corresponding ray parameter values for all earthquakes were estimated from the delay of the seismic waves arrival throughout the network as all stations have a reliable time reference. Near-surface shear wave speed values estimated from the array-based polarization analysis are correlated with major geological features of the area such as the Trans-Mexican Volcanic Belt and the local basin. We also apply the presented method to a minimal data scenario. The flexibility of this method allows it to employ a single ray parameter value as input, producing robust estimates for near-surface shear wave speed values within our case study. This feature becomes useful in a variety of scenarios, such as utilizing seismic data lacking timing information. The presented method constitutes an accessible and inexpensive alternative to study shallow seismic structure, opening opportunities to expand and complement current tools and techniques to assess seismic hazards.