Mapping the seismic noise field in Europe: spatio-temporal variations in wavefield composition and noise source contributions

Abstract

SUMMARYWe study spatial and temporal characteristics of the microseismic noise field across Europe. Rather than focusing on the areas of noise generation, the scope of this work is to characterize, at the scale of Europe, the spatio-temporal evolution of the noise wavefield that results from the interplay of the seismic noise sources and the propagation effect. To that end, we perform single station analysis in three period bands (PB1: 2.5−5 s; PB2: 5−10 s and PB3: 10−20 s) using three-component seismic data recorded by ∼1000 broad-band stations in the time period 2011–2019. We calculate, for each period band, station and day, a set of parameters that are practically possible to apply to a large data set, yet yields insight into the spatio-temporal evolution of the wavefield. These parameters are: the total energy level, the dominant period of the Primary and Secondary microseismic peaks, the horizontal direction with the most energy, the horizontal direction of the dominant Rayleigh waves and the square root of the energy ratio between the horizontal and vertical components. The analysis of these parameters shows that the noise field in Europe is dominated by surface waves from the North Atlantic Ocean with, in PB1 and PB2, an additional and significant contribution from the eastern part of the Mediterranean Sea. The relative contribution of these two source regions depends on the season, the influence of the eastern Mediterranean Sea being strongest in summer. The map of the peak period of the Primary and Secondary microseismic peaks indicates that the relative contribution of these two source regions is frequency dependent: the period of the Primary microseismic peak exhibits an overall increase with distance to the North Atlantic Sources, because of stronger attenuation of high-frequency wave contents. By contrast, the period of the Secondary microseismic peak is simultaneously influenced by sources in both the North Atlantic Ocean and eastern Mediterranean Sea. We show that in both microseismic peaks (PB2 and PB3), the wavefield is dominated by Love waves, as the horizontal components have the highest energy at approximately 90° angle to the direction of elliptical polarization. Moreover, our results show that lateral heterogeneities in the crust have a major influence on the noise field. In particular, the propagation directions of Love and Rayleigh waves show strong dependency on location (but not on time of year), with very sharp boundaries for example at the edge of the Alps. Thus, the scattering that takes place in the heterogeneous Alpine crust partly randomizes the directions of the microseismic wavefield in particular in PB1 and PB2. Finally, we show that the temporal evolution of the amplitude ratio between the horizontal and vertical components reflects the relative amounts of surface waves from the North Atlantic Ocean with respect to body waves from sources in the Southern Hemisphere. Thus, this ratio can be used as a proxy to identify time periods where body waves are significant in the noise wavefield.