Temporal change in seasonal variation of mean amplitudes at high-latitude seismic stations

Abstract:

Some seismic stations at high latitudes (greater than ∼ 60-70◦) show seasonal variations in average bulletin amplitude attributed to winter freezing of adjacent seas. Lower wave noise from frozen-over water allows lower-amplitude signals to be reported. We select 22 seismic stations in the International Monitoring System (IMS) of the Comprehensive Nuclear-Test Ban Treaty Organization (CTBTO) at latitudes >60◦, of which most show clear seasonal variations. Amplitudes from the Reviewed Event Bulletin (REB) of the CTBTO International Data Centre (IDC) were averaged monthly. For some of these stations, which had existed before they were incorporated into the IMS, bulletin amplitudes from the International Seismological Centre (ISC) extending back to the 1960s were used, allowing longer- term trends to be followed. The monthly average bulletin amplitude is higher in summer months when higher noise levels prevent lower-amplitude signals from being recognised, hence the stations are less sensitive in summer. Several stations show a gradual decrease in average bulletin amplitude over the past decade, with both the high and the low points of the seasonal variations decreasing. Some stations at lower latitudes also show a gradual decrease in average bulletin amplitudes, without the seasonal variations. We interpret these results as showing a gradual increase in network sensitivity leading to more low-amplitude arrivals being associated with an event and hence reported in a bulletin. For stations showing seasonal variations in sensitivity, the interplay between network and station sensitivity changes, and step changes in sensitivity caused by instrument upgrades, generally obscure trends. We therefore compare the average bulletin amplitudes with averages of the ”ANP/2” noise amplitude measurement reported by the IDC since 1999, which are essentially randomly-timed samples of the station noise in the mb passband (0.8-4.5 Hz). The effect of changes in network sensitivity on the ”ANP/2” measurements is judged to be negligible, hence they can be used to interpret the ISC average bulletin amplitudes back to the 1960s.