AN INVERSE ANALYSIS ALGORITHM OF DIFFERENT TRIGGER FACTORS FOR SLOPE FAILURE HAZARD ZONATION

Abstract

  This paper proposes an inverse analysis algorithm of different trigger factors for slope failure hazard zonation, based on Structural Equation Modeling (SEM). In this model, causal factors (i.e., geographical information and satellite remotely sensed data) are used as “observed variables,” while trigger factors caused slope failures are regarded as “unobserved latent variables.” Through measurement equation in SEM, the Inverse Estimation (IE) maps of trigger factor are produced with respect to two kinds of trigger factors that are “rainfall (Case-1)” and “earthquake (Case-2),” separately. As a final outcome, the differences between these IE maps are delineated on a “Trigger Factor Influence (TFI) map.” The TFI map and its interpretation would be useful for assessing differences of trigger factors with respect to slope failures, simultaneously.