Improving C1 and C3 empirical Green’s functions from ambient seismic noise in NW Iran using RMS ratio stacking method

Abstract

Listen

Translate article

The retrieval of stable and reliable empirical Green’s functions using ambient seismic noise plays a major role when studying the Earth’s structure at various scales. High-resolution noise correlation functions are obtained in the NW of Iran by processing techniques including dividing the continuously recorded raw data into short (i.e., 10 min) overlapping (i.e., 80%) time windows. We compare four stacking methods (i.e., linear, RMS, RMS ratio, and Nth-root stacking methods) to study robust and stable inter-station empirical Green’s functions. Our results indicate that the new RMS ratio method of stacking would be the optimal method to stack coherent signals. In other words, this method tackles problems including low signal-to-noise ratio (hereafter SNR) value, distortion of wave shape, and phase instability/unstable polarity treatment. In addition to noise correlation functions, we propose another strategy for the computation of the empirical Green’s functions. In this technique, the cross-correlation of scattered coda waves of the calculated noise correlation functions is performed individually. In addition to coda window length, we also investigate another effective parameter, the geometry of various virtual stations to obtain reliable empirical Green’s functions from the scattered coda waves of correlation functions with high SNR. The error of the velocities of Rayleigh wave empirical Green’s functions is on the order of approximately 0.6%, when compared to ambient seismic noise and scattered coda waves for a period band range of 3–10 s.