Bayesian calibration of rheological parameters for predicting displacement evolution of a dump slope in open pit mining

Long-term displacement prediction of dump slopes in open-pit mines is challenging because of the time-dependent behavior of slope materials and the uncertainty of rheological parameters. This study develops a Bayesian calibration framework based on a fractional Burgers analytical solution to predict the displacement of the Baorixile open-pit mine dump slope using Global Navigation Satellite System (GNSS) monitoring data. Markov Chain Monte Carlo (MCMC) sampling is employed to infer posterior parameter distributions and update displacement predictions. The results show that the posterior predictions reproduce the monitored displacement histories well and are significantly more concentrated than the prior predictions. The inferred parameters exhibit different levels of identifiability, posterior dependence, and time-dependent sensitivity. Compared with the classical Burgers model under the same Bayesian calibration framework, the fractional Burgers model provides a clearer decomposition of delayed elasticity, long-term viscous deformation, and temporal nonlinearity, leading to stronger mechanical interpretability. The proposed framework provides an uncertainty-aware approach for displacement prediction and mechanism-oriented deformation interpretation of the studied dump slope.