Application of a Generalised MPK Model with Data Fusion Approaches for Landslide Risk Assessment

The constitutive modelling of unsaturated soil behaviour within the MPK framework using data fusion approaches are discussed to envisage a practical approach for the evaluation and monitoring of landslide-prone areas. The present study focuses on improvements in subsurface suction distribution estimations via electrical resistivity tomography (ERT), which can be used to generate quasi-continuous suction profiles (QCSP) for numerical modelling. An explicit petro geophysical transfer function between suction and resistivity is presented herein, which was derived from the Waxman-Smits electrical resistivity model and the Van Genuchten equation for soil water retention curves. For efficient numerical modelling, a selective correction approach (SCA) is presented for petro geophysical transfer function updating procedure. The data acquisition and analysis architecture are envisaged within a data cloud platform, which is highly attractive within the context of internet-of-things (IoT). Although the focus of this work is on suction estimations, a similar approach can be used for estimation of other constitutive variables required within the MPK framework.