Exploring the feasibility of seismic monitoring using ambient noise coda Q: Experiments in the Aegean (Greece)

Ambient noise cross-correlations or auto-correlations provide near real-time information about subsurface properties. Changes in the Green’s function obtained from auto-correlations/cross-correlations inform us about velocity changes in the medium (dv/v). Numerous studies have found good correlation between dv/v and medium changes related to a large earthquake, volcanic activity, or even seasonal changes at shallow depths. Another seismic parameter which helps estimate such medium changes is the rate of decay of coda waves or the coda Quality factor (Qc). Low Qc estimates from earthquake data has been shown to represent cracks/fracture as well as fluid migration. Application of ambient noise data for Qc estimation are relatively recent (past 5 years), especially in the Alps and Japan, where good correlation was found with the regional geology. Qc measurements using ambient noise data can ensure continuous monitoring unlike those from earthquakes. In this work, we evaluate the feasibility of monitoring medium changes with Qc by using ambient noise data from Greece. We perform autocorrelation of noise data from the permanent network in Greece over a period of 2010-2021. Preliminary analysis shows a seasonal pattern in Qc at several stations when considering short period bands, which is likely related to seasonal changes at shallow depths due to precipitation. Stations with clear seasonal pattern in the dv/v are correlated with the seasonal pattern in Qc, which confirms that Qc perturbations indeed represent physical changes. Stations which show weak seasonality in dv/v have a seasonality pattern in Qc with a lag of similar number of days. These results suggest than Qc based monitoring has the potential to act as a supplementary data set to dv/v and may even provide more information about the nature of medium change based on whether Qc is related to scattering or intrinsic attenuation.