Resumen
A simplified geostatistical approach was adopted to assess the effect of spatial variability of soil properties on slope stability analysis in order to understand continental margin geologic processes and potential geohazards for an area of the central Scotian Slope, offshore Nova Scotia, Canada. The analyses are conducted on piston core samples, thus are restricted to ~12 m sub-seabed; however, the approach provides insight into the general effects of spatial and temporal variability. Data processing using geostatistics and assessment of spatial correlation are used to characterize the current dataset. A deterministic assessment was performed for both non-spatially averaged and spatially averaged core sections. The results indicate that the estimated factor of safety increased by about 30% when spatially averaged values were used. A probabilistic model is introduced to assess reliability of the slope. The approach makes use of estimates of both the mean and variance of input random variables (e.g., Su and ?b). The model uses an exact probabilistic formulation for the total stress stability analysis and a Taylor series approximation for the effective stress stability analysis. In both cases, the mean and variance of the factor of safety are computed, leading to estimates of failure probability. The results suggest that the deterministic analysis is conservative with respect to slope reliability, although they do not lead to an estimate of the probability of failure. While these results indicate sediment instability is largely unlikely under static conditions, the reality is that many examples of submarine slope failure are observed in the geologic record. These results suggest that cyclic loading (earthquakes) or pre-conditioning factors (elevation of pore pressures) are critical for slope instability on the Scotian Slope.