First-order, second-moment (FOSM) approximations provide an efficient way to assess submarine slope stability across large areas for which digital bathymetric data are available. This is demonstrated using 20m bin 3D seismic seafloor data for a deepwater area with typical geotechnical soil properties. Results are obtained in terms of a factor of safety mean and standard deviation for an infinite slope with pseudo-static seismic loading. From this the probability of sliding is calculated for each bin without the computational burden of Monte Carlo or other iterative methods. Because these types of probabilistic model incorporate parameter uncertainty into their input and output, they can be used to support decisions about the value of additional data collection, or justify more sophisticated analyses that may help to reduce output uncertainties. In addition to providing detailed maps of the probability of sliding, the analysis produces global statistics that allow insight into the broader response of the system to seismic shaking.
Evaluation of deepwater geohazards commonly entails assessment of slope stability either to understand the geologic history of a project area, or to anticipate the risk associated with future events, such as major earthquakes. This can be done qualitatively based on the presence or absence of past landslide deposits; semi-quantitatively using simple measures such as slope angle or gradient; or quantitatively using limit equilibrium slope stability analysis (e.g. Mackenzie et al., 2010). Limit equilibrium methods are widely known and attractive because they integrate the essential physics of sliding and allow evaluation of rare or unprecedented conditions (for example the effects of a large future earthquake). However, they also require specification of geotechnical variables, such as sediment shear strength, thickness and unit weight, in addition to some description of slope geometry (minimally the slope angle).