Abstract
Previous work on background noise at seismic stations in Greenland has shown minimum seismic noise in the winter months for frequencies around 1 Hz and maximum seismic noise in the winter months for periods around 6 sec. We have analyzed microseism data from three new digital seismic stations installed during the summer of 1991 in northeast Greenland at Nord, Daneborg, and Scoresbysund. We determined seasonal and station-to-station variations in spectral power density between August and December in the frequency band between 10 sec periods and 5-Hz frequencies. These variations are in agreement with previous studies at periods of 1 and 6 sec. During the summer months, all three stations recorded a minimum for the average spectral power density in the microseism band between 10- and 5-sec periods. From about 3-sec periods to at least 5-Hz frequencies, the average spectral power density is at a maximum during the summer at all three stations. Conversely, the winter months have a maximum in spectral power density between 10- and 5-sec periods and a minimum between about 3-sec periods and at least 5-Hz frequencies at all stations. Station-to-station average-spectral-power-density comparisons show that Nord and Daneborg are roughly comparable over most of the frequency band between 10-sec periods and at least 5-Hz frequencies. Scoresbysund has a systematically higher spectral power density between 8-sec periods and at least 5-Hz frequencies. Overall, Nord had the lowest background seismic noise, at some frequencies approaching the values of a low noise model.
We determined average direction of approaches in the 8- to 4-sec period band for each station during the months of August and November; these determinations agreed with previous studies. The predominant average direction of approaches were: southwest for Nord, south for Daneborg, and southeast for Scoresbysund. Although the microseism amplitude is larger and the direction-of-approach scatter is smaller during the winter months at all three stations, the direction-of-approach mean is apparently independent of season. A large number of storms develop around Iceland and typically track northeast, giving rise to large amplitude microseisms at Scoresbysund but relatively small amplitude microseisms at Daneborg and no microseism activity at Nord. This complete lack of microseism energy at Nord (and to a lesser degree Daneborg) from known frequent microseism sources in the Greenland Sea is shown for one 5-day period in August 1991. Other studies have shown that thick sediments in the Atlantic Ocean's continental margins are responsible for the absence of short-period surface waves from mid-ocean ridge earthquakes that have paths traversing such continental margins. Thick sediments act to attenuate, scatter, and disperse short-period surface waves. Indirect evidence indicates that the northeast Greenland shelf has thick and variable sediment layers. Because the paths of surface waves to Nord (and to a lesser extent Daneborg) originating from typical storms in the Greenland Sea have long path lengths traversing the northeast Greenland shelf, we conclude that this is the likely explanation for the lack of southeast directions from Nord (and to a lesser degree Daneborg) in the observed microseism direction of approaches.
Automated detection and association of surface waves
