Abstract
The advantages and disadvantages of using a back analysis of slope failures to evaluate soil shear strength are discussed. A methodology is presented herein that allows the implied level of reliability associated with soil shear strength parameters back calculated from slope failures to be estimated. A reliability approach is also used to estimate the probability of failure for a given limit equilibrium slope stability method, design factor of safety, and combination of back calculated Mohr-Coulomb shear strength parameters, c' and φ'. The methodology is illustrated using 39 landslides in the Orinda Formation in the San Francisco Bay area. The impact of additional case histories in the same geologic setting, i.e., a larger data set, on the required design factor of safety for a given probability of failure is also investigated.
