Deformation Assessment Considering an A Priori Functional Model in a Bayesian Framework

Abstract

Monitoring Earth-surface deformations with surveying instruments plays a key role in the knowledge of dynamics and the evolution of geophysical phenomena. During this stage, it is important to adopt an appropriate check of the statistical significance of point displacements. We verified the advantages of using a statistical Bayesian approach with a simple a priori geophysical model instead of the frequentist methods usually adopted. In particular, we analyzed global positioning system (GPS) coordinate time series in two applications. First, by using a dedicated device, we applied movements of known amplitude to a GPS antenna. Such known imposed values were used as a term of comparison both for the frequentist and for the Bayesian approach. Furthermore, we analyzed data from GPS campaigns carried out on a landslide located in the central part of Italy. The results of both investigations (imposed and unknown movements) show more valuable outcomes by using the Bayesian approach.