Modeling and managing uncertainties in rock-fall hazards

The assessment of rock-fall hazards is subject to significant uncertainty, which is not fully considered in general practice and research. This paper reviews and classifies the various sources of the uncertainty. Taking a generic framework for risk assessment as source, a probabilistic model is presented that consistently combines the different types of uncertainties, in order to obtain a unified estimate of rock-fall risk. An important aspect of the model is that it allows for incorporating all available information, including physical and empirical models, observations and expert knowledge, by means of Bayesian updating. Detailed formulations are developed for various types of information. Finally, two examples considering rock-fall risk on roads, with and without protection structures, illustrate the application of the probabilistic modeling framework to practical problems.