The safety of embankments under seismic conditions is a primary concern for geotechnical engineering societies. The reliability analysis approach offers an effective tool to quantify the safety margin of geotechnical structures from a probabilistic perspective and has gained increasing popularity in geotechnical engineering. This study presents an approach for probabilistic stability analysis of embankment slopes under transient seepage considering both the spatial variability of soil parameters and seismic randomness. The spatial varying soil parameters are firstly characterized by the random field theory, where a large number of random field samples of the soil parameters can be readily generated. Then, the factor of safety (FS) of the embankment slope under seismic conditions corresponding to each random field sample is evaluated through performing seismic stability analysis based on the pseudo-static method. A hypothetical embankment example is adopted in this study for illustration, and the influences of shear strength parameters, seismic coefficient, and the external water level on the embankment slope failure probability are systematically investigated. Results show that the coefficient of variation of the friction angle and the horizontal scale of fluctuation have more significant effects on the embankment slope failure probability. Besides, the seismic coefficient also affects the embankment slope failure probability considerably. For a given external water level, the failure probability corresponding to the downstream slope of the embankment is larger than that in the upstream slope.
