S11C-01 INVITED 08:05h
Methods for Detecting Microseism-Generating Ocean Waves
The likeliest field conditions for the generation of oceanic microseisms are the occurrence in an ocean wavenumber spectrum of energetic wave components travelling in opposite directions. This produces ${\em partial \; standing \; waves}$ at the sea surface. Besides the possibility of some reflection of wave energy from a steep coastline, the hindcasting of wave spectra (using the observed wind field) together with ${\em in \; situ}$ wave measurements with directional buoys, has demonstrated that such conditions are much more common over deep water than was previously thought. A simplified theory of the effect, confirmed by laboratory experiments, will be presented. The occurrence of partial standing waves also explains the generation of low-frequency sound (microbaroms) in the atmosphere, as has been shown by certain acoustical observations due to Evers and Haak (2001). Lastly, a new method will be proposed for detecting partial standing waves by aerial observations and by altimeter records from orbiting satellites.
S11C-02 INVITED 08:20h
Mid-Ocean Microseisms: Coastal Source Areas and Historical Wave Climate Implications
The Hawaii-2 Observatory (H2O), located about half way between Hawaii and California, is an excellent site for studying the origin and propagation of microseisms since it is located in the mid-ocean far from shorelines and shallow water. During the period between Dec 26, 2001 and January 24, 2002 on Leg 200 of the Ocean Drilling Program, the bridge crew of the JOIDES Resolution took environmental measurements (wind speed and direction, wave height, etc.) for comparison with the data collected by the H2O seismic system nearby. Comparison of the ship's weather log with the seismic data at frequencies from about 0.2 to 0.5 Hz shows a strong correlation of seismic amplitude with wind speed and direction, implying that the energy reaching the ocean floor (4977 m below) is generated locally by ocean gravity waves at the sea surface. These signals result from the well-known double-frequency (DF) microseism wave-wave interaction mechanism. Near-shore seismic stations on the U.S. West Coast see similar short period DF (SPDF) spectra, also generated locally by wind seas. SPDF microseism amplitudes lag sustained changes in wind speed and direction by several hours, with the lag increasing with wave period, most likely associated with the time necessary for the development of opposing seas for DF microseism generation. Correlation of swell height above H2O with microseism energy at lower frequencies, 0.1- 0.2 Hz, is poor, implying that the long period DF (LPDF) signals have their origin at distant locations. Correlation of the H2O seismic data with NOAA buoy data, with hindcast wave height data from the North Pacific, and with seismic data from mainland and island stations, defines likely source areas of these 0.1 - 0.2 Hz signals. Most of the microseism energy at H2O between 0.08 and 0.2 Hz appears to be generated by high amplitude storm waves impacting long stretches of coastline nearly simultaneously. At near-coastal seismic stations, both SPDF and LPDF microseism levels are generally dominated by local generation at nearby shorelines. Thus wave climate reconstructions from archived seismograms at near-coastal stations along the West Coast, e.g. during the 1940-41 strong El Ni\~{n}o winter, predominately reflect local ocean wave variability.
http://www.soest.hawaii.edu/h2o/
S11C-03 INVITED 08:35h
Locating Ocean Microseism Sources With Seismic Directional Information
We present two different methods to calculate the source azimuth of ocean generated microseismic noise on land, one from beamforming using the vertical components of a seismic array, the second from particle motion analysis at single three-component stations. We confirm our location capabilities by comparison of the microseism propagation direction with ocean wave height maps. Coupling from ocean wave to seismic energy occurs when swells hit coastlines, but it is modulated by coastal geometry. Ocean microseism generated by Atlantic storms are registered at arrays in Wyoming and southern California; such transcontinental noise events occur several times every year with a duration of 1-2 days each. We have processed continous noise registered from 1997-2004 at the ANZA array in California and investigate the influence of El Nino-Southern Oscillation events on noise levels and directions on land. We also present a real-time automated system that generates maps of noise directions from continuous seismic records. A better understanding and location of microseism sources are the first steps to a possible future use of the ocean-generated continuous noise as a signal for seismic investigations of shallow earth structure.
S11C-04 08:50h
Approaches for the Detection of Microseism Generating Regions from Space
Ocean microseisms have been suggested as a potential source for tomographic imaging of the Earth. An optimal tomographic approach requires that the source regions for ocean microseism generation be known. Longuet-Higgins (Philos. trans. R. Soc. London, 1950) has suggested that standing waves in the ocean are responsible for the generation of microseisms, so that a potential way of determining the regions responsible for microseism generation is the determination of which regions of the ocean exhibit standing wave activity. Although ocean standing waves have been observed and modeled in a laboratory setting, there is currently no remote sensing instrument capable of providing direct observations. In this paper, we consider remote sensing or modeling measurements which might provide observations indicative of potential standing wave activity. The first option, radar altimetry, provides a measurement of the significant wave height and, at lower resolutions, sea surface skewness. It is expected that non-linear interactions will result in a spatial modulation of the wave field and an increase of surface skewness. We review the accuracy of current ocean altimeters and conclude that, while these type of observations might be possible in the future, current altimeters do not have sufficient accuracy to retrieve the expected signals. An alternate approach is to use wave directional spectra to estimate the standing wave energy which might be observed. There are two potential sources for the estimation of directional wave spectra: wave action models (WAM's) or synthetic aperture radar (SAR). We review the capabilities of each of these measurements and conclude that they present a useful potential source for estimating regions of microseism generation. To test this conclusion, we present results of comparisons between SAR data collected off the California and WAM data with seismic data.
S11C-05 09:05h
Swell Activity in the Southern Pacific From Seismic and Infrasonic Data
A temporary network of 10 broad band seismic stations has been deployed in French Polynesia for the Polynesian Lithosphere and Upper Mantle Experiment (PLUME). The stations are installed either on volcanic islands or on atolls of the various archipelagos of French Polynesia to complement the geographic coverage provided by the permanent stations. At all sites, the proximity of the ocean generates a high microseismic noise level. The power spectral density of the seismic data show clear peaks in the 0.05 to 0.09 Hz frequency range (11 to 20 s period), corresponding to typical swell frequencies. In this single frequency peak, the swell-related seismic signal is elliptically polarized and is contained within the horizontal plane. We measure hourly its amplitude and azimuth and demonstrate that in this frequency range, the amplitude of the microseismic "noise" shows very similar variations from station to station and is strongly correlated with the swell amplitude predicted by the NOAA, wind-derived, "WaveWatch" models. Deducing the swell direction from the ground particle motion has to be done with care since the island ground motion can be strongly controlled by local swell refraction processes. We find cases, however, such as in Tahiti or on roughly circular atolls, for which the azimuth of the incoming swell may be well deduced from the seismic data. This therefore demonstrates that the swell-related seismic signal observed in French Polynesia in the single frequency peak can reliably be used as a proxy for swell amplitude and azimuth. The presence of an infrasonic array installed in Tahiti also provides the opportunity to use microbarometric signal to characterize the swell activity. For a period of low wind (which is a strong noise generator) and high swell, we evidenced a clear correlation between the microseismic and infrasonic noise amplitude, together with the predicted and the locally observed swell amplitudes, suggesting that such infrasonic data can be used reliably during quiet wind conditions, to retrieve the swell amplitude but not its azimuth.
S11C-06 09:20h
Strongly localized near-coastal generation areas for $\sim$4 s microseismic noise in the North-East Atlantic
Mapping the generation areas for ocean-generated microseismic noise is of importance both for the planning of future ocean bottom experiments and for the use of historical seismological microseismic noise data as a proxy for ocean wave height in historical climate studies. We report on the results of a large aperture (400 km), 3 month pilot deployment of 10 ocean bottom stations (7 seismometers/hydrophones, 3 hydrophones) in the North Atlantic south of Iceland. The noise level was found to be highly variable both with time and for the different stations, and shows marked variability between bands. By correlating noise levels recorded on the ocean bottom stations as well as on nearby land stations with wave height data for the same period, we map out estimated generation areas for microseismic noise. For high frequency noise (0.5--2 Hz) the noise turns out to be predominantly related to local waves, and for lower frequency noise ($<$0.3~Hz) noise seems to originate mainly in a few source regions near the coasts of Ireland, Scotland and Iceland. The generation areas for the ocean bottom instruments were similar to those of land instruments in the same region. It should thus be possible to estimate the generation areas for proposed future ocean bottom installations based on nearby land stations, and thus allow better estimates of expected noise levels than those derived from regional averages of typical weather conditions. The directivity of noise suggests optimised array designs for noise suppression. Below the microseismic band and at the relatively noisier stations, the vertical seismograms at frequencies below $\sim$0.3 Hz could be substantially enhanced by removing the part correlated with the horizontal components, suggesting that noise introduced by instrument tilt was considerable at those stations.
S11C-07 INVITED 09:35h
Excitation of earth's incessant free oscillations by Atmosphere-Ocean-Seafloor coupling
The observation on long period seismograms of continuously excited free oscillations of the Earth was first made by japanese scientists in 1998, during intervals free of significant earthquakes. Since then, attention has focused on elucidating the physical mechanism responsible for them. The sum of all small earthquakes remaining in the records cannot explain the observed level of excitation. The mechanism must be shallow, as fundamental modes appear to be preferentially excited. It shows seasonal variability, with peaks in the northern and southern hemisphere winters. Stochastic mechanisms involving turbulent motion in the atmosphere have been proposed as well as random distribution of sources around the world. An alternative potential source of excitation of the continuous oscillations is in the oceans, but so far the observations, based on the computation of spectra or correlations of signals across long time series, do not have the spacial and temporal resolution to locate these sources. We have developed an array-based method to detect and locate sources of very long period surface wave energy, utilizing the dispersive properties of Rayleigh waves. Our basic approach uses data from two large aperture arrays of very long period seismometers (BDSN in California and F-NET in Japan). We stack the data after projection to the center of each array, and look for directions of arrival of maximum amplitude in the stacks as a function of back-azimuth, taking into account the array response. We show that, for each array, there is a well defined preferential direction, which is stable over one season but changes significantly from winter to summer. The fluctuation as a function of time of the maximum stack amplitudes are correlated across the two arrays and point to the northern Pacific ocean in the northern hemisphere winter and the southern Oceans in the summer, correlating with changes in the global distribution of maximum wave height. We infer that the background oscillations originate primarily in the oceans, and are caused by a non-linear coupling mechanism involving the atmosphere (winds), the oceans (transfer of energy from ocean waves to infragravity waves) and the seafloor (transfer of energy to elastic waves).
S11C-08 09:50h
Observations of Infragravity Waves at the Monterey Ocean Bottom Broadband Station (MOBB)
Ocean bottom broadband seismic observations show increased noise level when compared to land recordings. Infragravity waves are an important source of background noise at periods longer than 20 seconds. Long-period background noise can partially be removed by post-processing (Crawford and Webb, 2000; Stutzmann et al., 2001). At the same time, observations of infragravity waves can help us better understand their generation from short-period waves. We investigated the correlations between the infragravity waves recorded at MOBB and the short-period ocean wave data recorded at nearby ocean buoys. MOBB was installed 40 km offshore in the Monterey Bay at a water depth of 1000 m in April 2002 in collaboration between Berkeley Seismo Lab and Monterey Bay Aquarium Research Institute (MBARI) (McGill et al., 2002). It comprises a three-component broadband seismometer with a temperature sensor, a water current meter measuring current speed and direction, and a differential pressure gauge (DPG). The station is continuously recording data which are retrieved, on average, every three months. We calculated power spectral density (PSD) of the MOBB vertical component for 1-hour long periods for all the data recorded until June 2004. The results were compared to the spectral wave density (SWD) of the ocean waves as recorded on the eight regional NOAA buoys located from offshore Southern California to Alaska, and the six local NOAA buoys located offshore Northern California. The results clearly show that the width of the infragravity wave band recorded on MOBB was best correlated with the energy of the ocean waves recorded at the local NOAA buoy, located just 23 km W of MOBB. To further explore observed correlation between the MOBB PSD and the buoy SWD, we focused on several 7-day periods and calculated correlation coefficient between 1-hour long MOBB PSD and buoy SWD values. The highest correlation was observed for the closest buoy located just outside Monterey Bay. The results show that the 7-17 sec ocean waves were best correlated with seismic waves with 30-200 sec period. A weaker correlation was observed with the other two buoys located over the continental shelf to the north. Correlation with the buoys located to the south and those located further offshore was much smaller. We plan to further investigate these observations as well as compare them with the results obtained from island (FARB) and other land BDSN stations. We will also include pressure and tides data to better understand where and how the energy is transferred from the ocean to the seismic waves. We will use the obtained results to remove the long-period noise from the MOBB vertical channel to improve the quality of the seismic data.