University of Texas at Austin, United States

The Ursa Basin, at approximately 1000 m depth on the eastern levee of the Mississippi Canyon, Gulf of Mexico continental slope, is an extraordinary natural laboratory to investigate large-scale aseismic slope failure phenomena. Recent studies have suggested that seismic activity recorded in this area, is a result, at least in part, of shallow gravitational sliding rather than deep-seated tectonic processes. Extensive occurrence of Mass Transport Deposits (MTDs) in the Ursa Basin, both in time and space is also documented by multibeam and seismic reflection data. In June 2005, Integrated Ocean Drilling Program (IODP) Expedition 308 drilled three Sites adjacent to major Recent failures and mud-volcano type fluid escape structures, and through a series of MTDs of Pleistocene to Holocene age. Along these holes a complete suite of logs, sedimentological and geotechnical data were acquired, which illuminate factors that control failure initiation, provide insights into the failure mechanism itself and allow characterization of the hazard potential from future slope instabilities. On seismic data most MTDs appear as transparent bodies, and this appears to indicate some degree of remobilization of the failure mass, yet they do not appear to have moved far from the failure initiation area. Little overburden was removed because the failed masses did not evacuate the failing zone. MTDs that appear as seismic transparent bodies do not exceed 60 to 70 m in thickness. The first 60-70 m of overburden in Ursa Basin are characterized by very high porosity, decreasing down hole from 80 to 55%, and water content, rapidly decreasing from 100 to 40%, i.e. from values near the liquid limit to very close to the plastic limit. From measurements of porosity and stress state, we infer that fluid overpressure, derived from rapid sedimentation, is the most likely factor that initiated/controlled slope instability in the past. Fluid overpressures result in effective stresses that are 50 to 70% lower relative to hydrostatic conditions. Isotropically consolidated, undrained triaxial tests suggest low cohesion and a friction angle around 28 degrees for fine-grained mudstones. This indicates that, given the slope geometry, nearly lithostatic overpressures were/will be needed to trigger slope failure. Profuse evidence of fluid escape structures, including large mud volcanoes, near failure scarps might indicate that such conditions existed in the past.