Seasonal Periodicity and noise Discrimination of Microearthquakes at the Garm Test Area

Abstract

We studied the time series of the total number of weakest earthquakes (microearthquakes) with magnitudes M ≤ 2 recorded by 15 seismic stations of the Garm test area in 1975–1985, as well as the microseismic noise level at different frequencies of about 0.005–50 Hz recorded by the Garm station of the IRIS observation network for the periods 2006–2009 and 2013–2019. The main goal of the research was to find a correspondence and, if possible, cause-and-effect relationships between the quasiperiodic components present in both types of time series. Seismic events recorded by less than four stations and not included in the main earthquake catalog due to the impossibility of reliable determination of their hypocentral parameters were used as microearthquakes. At each station, only the number of such events per day was recorded. To identify stable periodicities, all series were studied using spectral and periodogram analysis methods, adapted to work with nonstationary signals. The initial data series were analyzed, as well as their regularized variants (logarithms and ranks of the number of events). Regularization made it possible to reduce the contribution of nonstationary effects (random bursts of activity, increase in the number of microearthquakes during periods of aftershock series, etc.) and ensured robustness of the results. The analysis revealed that of all possible rhythms common for the two types of time series, only the seasonal one is significant. The seasonal rhythm of microseismic noise at different frequency channels differs in phase and amplitude. High-frequency noise (channels of the IRIS system with a center frequency from 1 to 47 Hz), falling in the frequency range of the measuring and recording channels of the seismic stations of the Garm test area, is in antiphase with the seasonal variation in the number of recorded microearthquakes at all stations, with some specificity for individual stations. The antiphase change in the number of recorded microearthquakes and the level of microseismic noise in the recording band of the seismic stations suggests that the reason for these changes is the noise discrimination of microearthquakes when they are identified on seismograms.