S13F-01 INVITED
Relocating and Characterizing the 10 Feb 2006 "Green Canyon" Gulf of Mexico Earthquake Using Oil-Industry Data
On 10 Feb 2006, a magnitude 5.2 earthquake occurred in the Green Canyon deep-water block of the Gulf of Mexico offshore Louisiana. At the time, this was the largest earthquake in the Northern Gulf of Mexico since at least the early 1970's. This earthquake is of particular interest to the oil industry because of its location in an area of significant underwater infrastructure development. The earthquake is also of interest to the earthquake seismology community because of its unusual intra-plate location and anomalous signature. The earthquake's radiated high-frequency body waves were unusually weak compared to its surface waves, and attempts to fit the earthquake with a typical double-couple source mechanism have been unsuccessful. The event could not be well located using traditional seismic monitoring networks because of the lack of any nearby observations, especially to the South. Two nearby oil-exploration seismic surveys did serendipitously record the event from the South, however: the BP-BHP Atlantis Ocean-Bottom-Node survey, and the CGG Green-Canyon phase VIII multi-client streamer survey. These surveys' instruments were not designed for recording frequencies below 5Hz, so the earthquake signals were weak -- much weaker than the high-frequency signals from the surveys' own airguns (which is what the instruments were designed to record). However, by low-pass filtering the data and beam-forming the arrays (of several hundred receivers each) we were able to extract usable earthquake arrivals. These new seismic observations were then used by the USGS in a first-arrival traveltime inversion to relocate the earthquake. We also measured the phase velocity and azimuth of the earthquake arrivals across the arrays, and used those azimuths as an independent method of estimating the event's location. The event was also recorded by a SeaStar Tension-Leg Platform, which reported an apparent local subsidence of the seafloor of\ .8 inches associated with the event. We are currently attempting to identify the most geologically plausible scenarios that are consistent with all the available data, and determining how these might be tested by further field observations.
S13F-02
Microearthquakes Beneath the Endeavour Hydrothermal Vent Fields: Insights Into Reaction Zone Processes
From 2003-2006, a novel seismic network comprising seven short-period corehole seismometers and a broadband Guralp CMG-1T OBS was deployed using remotely operated vehicles in a subseafloor configuration on the Endeavour segment of the Juan de Fuca mid-ocean ridge. The seismic monitoring array was one part of a multi-disciplinary prototype NEPTUNE experiment designed to investigate the linkages between seismic deformation, hydrothermal fluxes, and microbial productivity along oceanic plate boundaries. The seismic network recorded high-quality data that illustrate the advantages of using an ROV to deploy seismometers in well- coupled configurations that are also away from the effects of ocean currents. A preliminary analysis of the first year of Keck seismic data was undertaken during a research apprenticeship class taught in the fall of 2004 at the University of Washington's Friday Harbor Laboratories. Eight post- baccalaureate students obtained a preliminary catalog of nearly 13,000 earthquakes on the Endeavour segment. Two of the apprentices conducted a second-pass analysis to refine the locations of ~3000 earthquakes that are within or near the network. Further analysis of these proximal earthquakes has focused on the application of cross-correlation and relative relocation techniques, the determination of focal mechanisms using P-wave first motions and P- to S-wave amplitudes ratios, and improved estimates of earthquake magnitudes. The results show that the entire Endeavour segment was seismically active during 2003-2004. Within the network, the earthquakes are located in tight clusters centered at ~2 km depth in the inferred location of the hydrothermal reaction zone immediately above a crustal magma chamber imaged by seismic reflection studies. The number of earthquakes below each hydrothermal vent field correlates with the heat flux measured by other researchers and the vertical thickness of this reaction zone, inferred from the distribution of seismicity, is ~0.5-1 km. Preliminary focal mechanisms show a transition from ridge-perpendicular extension beneath the rise axis to ridge-perpendicular compression on either side over horizontal distances of <1 km. This is most simply explained if the stress field above the magma chamber is locally influenced by either the injection of magma or over-pressured hydrothermal fluids. This observation, in combination with a reaction zone thickness that is at least an order of magnitude larger than that inferred from thermal boundary layer theory, suggests that the heat transfer between the magmatic and hydrothermal systems is not simply the result of cold water attacking hot magma. Instead, it appears as though the magmatic system also attacks the reaction zone from below.
S13F-03
LOCAL AND DISTANT RECORDING OF THE KALAPANA, HAWAII BENCH COLLAPSE OF 16 AUGUST 2007
On 16 August 2007 at 01:24 GMT, a small event took place on the Southeastern shore of the Big Island of Hawaii, with an initially assigned HVO duration magnitude of 1.7. It followed two larger earthquakes at essentially the same location, on 13 and 15 August. This epicentral location was the site of a documented, but untimed, bench collapse reported on the HVO web site. In a preliminary study, we analyzed both seismic waves recorded at the local IRIS station POHA located on the Big Island, 75 km from the epicenter, and hydroacoustic signals recorded at Christmas Island and in Polynesia. The local signal at POHA for the 16 August event is dominated by a 4-minute long signal peaked at about 3 Hz starting 1.5 minutes after the P wave, which is absent from the 13 and 15 August records. Similarly, hydroacoustic records at XMAS and in Polynesia show T waves of long duration and relatively low amplitudes, with a Duration-Amplitude discriminant [Talandier and Okal, 2001] D = -4.95 at XMAS. This value is lower than observed for most earthquake sources, and also than computed at Wake for the 1998 PNG landslide. These properties identify the 16 August 2004 as a bench collapse, with the event at 01:24 probably representing a small earthquake nucleating the process. The combination of the seismic and hydroacoustic records provides a unique opportunity for the detailed study of this phenomenon, which until now had been caught only rarely on distant instruments. A full report based on a more comprehensive dataset will be presented. ~
S13F-04
Performance of the hybrid method for waveform and travel-time analyses based on a comparison with the hypocenters calculated from the ocean-bottom-seismometer network
The locations and focal mechanisms of earthquakes provide basic information for understanding the features of earthquakes and their focal areas such as faulting, fault orientation, stress fields, etc. Ito et al. (GJI, 2006) estimated the accuracy of the centroid locations calculated by the regional CMT inversion method, in which the highly dense broadband seismic network in Japan was used. The accuracy in the horizontal and vertical directions was approximately 5 km when the seismic stations were located around the focal area. However, the accuracy of the horizontal distribution decreased by more than half when the distribution of the seismic stations was restricted azimuthally, i.e., for an analysis of offshore earthquakes. It should be noted that the depth accuracy did not decrease when the stations were restricted. In this paper, we first develop a new method for investigating the hypocenter and focal mechanism of offshore earthquakes occurring at a distance from the landward seismic network. Next, we apply this method to small or intermediate earthquakes occurring in the subduction zone along the Japan Trench. Finally, we discuss the reliability of our method by comparing them with accurate hypocenters obtained by the ocean-bottom-seismometer (OBS) network. In our method, we determine a hypocenter and focal mechanism by using both the observed waveform and the travel times of P and S waves; this method is henceforth referred to as the hybrid method. First, we calculate the focal depth and focal mechanism at an initial epicenter by the moment tensor inversion method using the waveforms observed at the National Research Institute for Earth Science and Disaster Prevention (NIED) F-net broadband and Hi-net tiltmeter networks. Next, we fix the focal depth calculated by the moment tensor inversion method and determine the epicenter based on the travel times by using the least square method. With regard to the hybrid method, we repeatedly used the moment tensor inversion method and the least square method. After 2002, we have repeatedly deployed the OBS network in the Miyagi-oki region, which is located on the landward slope along the middle part of the Japan Trench. In this region, interplate earthquakes with seismic magnitudes exceeding 7 repeatedly occur every 40 years. An interplate earthquake with a magnitude of 7.2 occurred in 2005. Using the data obtained from the OBS network, Hino et al. (EPS, 2007) accurately calculated the hypocenters of aftershocks. Here, we compare the hypocenters obtained by the hybrid method with those calculated by Hino et al. (EPS,2007); they are found to be consistent, and the accuracy in the horizontal and vertical directions is within approximately 6 km. One of the nodal planes of the aftershocks occurring near the plate interface exhibits a high consistency with the slope of the subducting plate. By using the bootstrapping method, the estimated depth error in the hybrid method is found to be approximately 3 km, and 10 degrees for fault parameters such as strike, dip, and rake, which are calculated from the moment solutions assuming a pure double-couple source. These results show the validity of the hybrid method for calculating the hypocenters and focal mechanisms of offshore earthquakes.
S13F-05
Seismicity and Tectonics of the Northeast Caribbean from a Passive Ocean Bottom Seismograph Study
The Northeast Caribbean Plate boundary is prone to tsunamis due to the occurrence of a) subduction zone earthquakes along the Lesser Antilles and the Hispaniola and Puerto Rico trenches, b) other large earthquakes due to more local tectonic activity, such as in the Mona Passage (1918 tsunami) and in the Virgin Islands basin (1867 tsunami), and c) moderate earthquakes that destabilize steep underwater slopes and cause a tsunamigenic landslide. Recent seismicity has been concentrated to the north and west of the British Virgin Islands, in the region sometimes referred to as the Sombrero Seismic Zone (SSZ). In the combined seismicity catalog maintained by the Puerto Rico Seismic Network (PRSN), several hundred small to moderate magnitude events can be found in the SSZ prior to 2006. However, beginning in 2006 and continuing to the present, the rate of seismicity suddenly increased, and a new locus of activity developed to the east of the previous location. Ocean bottom seismographs (OBSs) were deployed in the Northeast Caribbean for three months in 2005 and for six months in 2007 to record earthquakes in the SSZ in an effort to better constrain earthquake locations, to identify crustal features, including a possible rupture in the subducting North American Plate, that could cause tsunamigenic events and to clarify the tectonics that lead to larger earthquakes and tsunamis. Data were analyzed jointly with records from permanent stations of the Puerto Rico Seismic Network (PRSN), whose stations are located on islands to the south of the Sombrero zone and are therefore biased and incomplete without offshore complements. Our goal was to identify, locate, and estimate focal mechanisms for small- magnitude seismic events in this corner of the Caribbean. In the 2005 data set, consisting of recordings by nine OBSs and the ten closest PRSN stations, we identified and analyzed thirty-five events on at least four OBSs and one or more PRSN stations and an additional fifteen events that were observed on four or more OBSs only. Upon relocation, the jointly recorded events revealed tighter clustering and clear biases with respect to original PRSN locations. Changes in epicenters were greatest in the northeastern limit of activity, while there was a consistent progression toward deeper focal depths for events closer to the island of Puerto Rico. The average change in focal depth for the cluster of five events closest to Puerto Rico was 35 km. Events located solely with OBSs were consistent with the relocated events. Most of the relocated events are consistent with the depth and dip of the subducted North American plate in this region and may therefore be associated with active subduction, rather than the shallow tectonics suggested by original locations. Events recorded during the longer 2007 OBS deployment are being analyzed to explore this association and its implications for tearing of the North American Plate lithosphere during subduction.
S13F-06
Modeling acoustic wave propagation in the Southern Ocean to estimate the acoustic impact of seismic surveys on marine mammals
According to the Protocol on Environmental Protection to the Antarctic Treaty, adopted 1991, seismic surveys in the Southern Ocean south of 60°S are exclusively dedicated to academic research. The seismic surveys conducted by the Alfred-Wegener-Institute for Polar and Marine Research, Bremerhaven, Germany during the last 20 years focussed on two areas: The Wedell Sea (60°W - 0°W) and the Amundsen/Bellinghausen Sea (120°W - 60°W). Histograms of the Julian days and water depths covered by these surveys indicate that maximum activities occurred in January and February, and most lines were collected either in shallow waters of 400 - 500 m depth or in deep waters of 2500 - 4500 m depth. To assess the potential risk of future seismic research on marine mammal populations an acoustic wave propagation modeling study is conducted for the Wedell and the Amundsen/ Bellinghausen Sea. A 2.5D finite-difference code is used. It allows to simulate the spherical amplitude decay of point sources correctly, considers P- and S-wave velocities at the sea floor and provides snapshots of the wavefield at any spatial and temporal resolution. As source signals notional signatures of GI-, G- and Bolt guns, computed by the NUCLEUS software (PGS) are used. Based on CTD measurements, sediment core samplings and sediment echosounder recordings two horizontally-layered, range-independent generic models are established for the Wedell and the Amundsen/Bellinghausen Sea, one for shallow (500 m) and one for deep water (3000 m). They indicate that the vertical structure of the water masses is characterized by a 100 m thick, cold, low sound velocity layer (~1440 - 1450 m/s), centered in 100 m depth. In the austral summer it is overlain by a warmer, 50 m thick surface layer with slightly higher sound velocities (~1447 - 1453 m/s). Beneath the low-velocity layer sound velocities increase rapidly to ~1450 - 1460 m/s in 200 m depth, and smoothly to ~1530 m/s in 4700 m depth. The sea floor is mainly covered with soft fine-grained clayey or silty sediments, so that P- and S-wave velocities of 1550 and 200 m/s and a wet bulk density of 1400 kg/m3 are assumed. In a first step the acoustic impact of one seismic line of 10 - 20 km length is computed for the two generic models, assuming a typical shot interval of 15 s and a ship speed of 5 kn. The acoustic impact is determined by running the finite-difference scheme once, shifting the resulting wavefields in space and time according to the movement of the ship and the shot interval, and summing-up the appropriate snapshots of the propagating wavefield. As results, time-dependent contour maps of the cumulative peak-to-peak, zero-to-peak, rms and sound exposure levels are derived. From these contour maps time-dependent exposure histories and histograms of the received sound pressure levels are extracted for animals staying at fixed depth and range positions along the seismic line. Different hearing abilities of low-, mid- and high-frequency cetaceans are taken into account by applying the M-weighting filter characteristics. In a second step the cumulative sound exposure of several parallel and intersecting seismic lines is computed. The layout of the lines is derived from the cruises ANT-XIV/3 and ANT-XXIII/4 to the Wedell and the Amundsen/Bellinghausen Sea, which on average had the closest seismic line spacings of former cruises to both regions. http://www.awi.de
S13F-07
The Role of Shear and Interface Waves in the Excitation of T-waves
T-waves are late arriving phases on a seismogram which travel a significant portion of their journey from the source to the receiver along a slow oceanic path. Park, Soukup and Odom [2001] proposed a modal scattering mechanism which permits energy from steeply arriving rays impinging on the ocean bottom to be converted to shallow grazing angle rays corresponding to low order modes known to comprise the T-wave signal. The Scholte interface waves are crucial to this energy transfer as they have anti-nodes nearly coincident with the ocean bottom. Any seafloor roughness acts as a secondary source located right on the Scholte wave anti-node. This allows energy to be efficiently transferred from high energy modes to lower energy, and slower, T-wave modes. In fact the presence of the Scholte waves is crucial to the existence of T-waves. Some finite shear strength in the bottom sediments and/or upper ocean crust is essential for the existence of the Scholte waves. Elastic interface waves do not exist the boundary between two fluids. The effect of the the shear modulus of the ocean bottom sediment and ocean crust on the excitation of the interface waves and T-waves is discussed.
S13F-08
Bottom Interaction in Long-Range Ocean Acoustic Propagation
Four Ocean Bottom Seismometers, consisting of one hydrophone and one vertical component geophone channel each, were deployed below the deep vertical line array (DVLA) on the North Pacific Acoustic Laboratory/ Long-range Ocean Acoustic Propagation Experiment (NPAL/LOAPEX). Data, sampled at 500sps, were recorded continuously from the deployment in September 2004 for over 100days. The OBS's were on the seafloor in about 5,000 m of water and would be in the shadow zone for long range, ducted propagation in the sound channel. The goals of the deployment were: 1) to quantify the amount of energy that leaks out of the sound channel into the shadow zone, 2) to measure the relative sensitivity (signal-to-noise) of seafloor hydrophones and vertical component geophones to long- range signals and 3) to study the physics of earthquake generated T-phases. Leakage was observed out to 3200km on the vertical component geophone and out to 1000km on the hydrophone. Propagation velocities to the furthest ranges correspond primarily to sound channel paths (not surface or bottom reflected paths). The ratio of the pressure to vertical velocity varies between arrivals on the same trace. Some arrivals are extremely robust (appear on each transmission over 30hours with an arrival time varying be less than +/- 7msec) and some arrivals are ephemeral (magnitude varies tremendously over 2-3 minutes ) (eg at 500km range). (The OBS deployments were funded by contributions from ONR, NSF and WHOI. Data analysis funded by ONR.)