A Detailed Near-bottom Survey of Large Gas Blowout Structures Along the US Atlantic Shelf Break Using the Autonomous Underwater Vehicle (AUV) SeaBED

The autonomous underwater vehicle (AUV) SeaBED was deployed along the edge of the Virginia/North Carolina continental shelf in July 2004 to investigate a series of crack-like features. Prior investigations indicated these are large (up to 4km long, ~50m deep) gas blowout structures. In addition, the features were carefully mapped with a multibeam bathymetric sonar and sampled for geochemical and sedimentological analysis. The AUV completed 16 successful dives and collected microbathymetric, temperature, salinity, and methane concentration data, along with approximately 36,000 photographs while maintaining an altitude of approximately 3m above the seafloor. The AUV was deployed without seafloor transponders. Instead, the vehicle's navigation was based on a doppler inertial system. The accuracy of the navigation was tested by comparing seafloor depth, measured by the vehicle, with the shipboard swath bathymetry, and by computing the differences in seafloor depth where the vehicle tracks cross. The mean crossover error is 2.18m, showing the vehicle's navigation is actually quite accurate. Photos taken near the crossovers will be compared to determine the exact location of the crossovers. Dissolved methane concentration data, collected using a METS sensor, show elevated concentrations throughout most of the survey area, with the largest anomalies being very localized (50-100m across). Photos, color corrected for attenuation in average seawater, are red-shifted wherever high dissolved methane concentrations are detected, providing additional evidence for seepage in those areas. No other visual evidence of methane venting, such as authigenic carbonates, chemosynthetic communities or gas bubbles, is recognized in the photos, although shells, fish and invertebrates are commonly seen. Methane concentration data at most track crossovers are inconsistent, indicating the magnitude of the anomaly may be affected by time dependent factors such as tides, currents or instrument drift. We favor the explanation of methane flux being modulated by tidal variations, in keeping with observations made at cold seeps on the western coast of North America. We will test this hypothesis by minimizing the crossover errors as a function of tidal stage.