ROCKFALL FAILURE ANALYSIS BASED ON SFM PHOTOGRAMMETRY AND RAINFALL EFFECTS IN THE CASE OF A VERY BRITTLE CLIFF IN DJBEL KORBOUS(NORTH-EASTERN TUNISIA

Djbel Korbous (Cap Bon) is located in north-eastern Tunisia, where the issues of landslides hazards have recently emerged. Despite the multitude, the complexity, and the high degree of dangerousness of these instabilities, the region remains virgin without any previous numerical support or basic data for a risk study.

It is an important relief dominating the southeastern edge of the Gulf of Tunis and presents a special site for tourism and therapy. The main road, which is the only access to it, is a coastal road passing along a damaged cliff and suffers from several instabilities and rockfalls of different sizes reaching the sea and destroying the protection measures.

Structure for motion photogrammetry, being a simple, practical and inexpensive monitoring tool, was an efficient choice to initiate the modeling of a very vulnerable and inaccessible cliff. The present study aims at defining a multi-risk assessment methodology that fits the characteristics of a touristic area with no pre-existing digital field data.

The first part aims at the development of a high-quality digital terrain model for the entire site, about 12km², which required photogrammetric acquisitions from the sea and GPS surveys to realize the georeferencing. Thus the mapping of phenomena and the characterization of potential and fallen volumes allow the constitution of the first instabilities inventory for the zone. In addition, many terrestrial LiDAR scans were realized to complete the first model as well as several local photogrammetric acquisitions were done to capture the Spatio-temporal instabilities’ evolution and to understand the mechanism of rupture, the cliff degradation and the rainfall effects.

The second part of the study is to constitute a quantitative and qualitative rockfall hazard zoning based on 2D and 3D trajectory modeling (RocFall, RockyFor3D and Trajecto3D).

As a result, SFM photogrammetry enabled the reconstruction of a high spatial resolution data support, providing a reliable and detailed 3D model for a site where there was no numerical data. Subsequently, we define the main structurally controlled failure types and we could detect and map all potential instability sources from steep slopes.

The heavy rainfall episodes can show a strong dependence on rockfall activity and rainfall intensities through a continuous degradation of the rock strength and the increase of pre-existing fractures displacement.

Combining different rockfall numerical models all together allowed a complete and comparative analysis methodology to predict the rockfall runout distance and the propagation areas with an estimation of the fallen blocks' velocities and energies.

Keywords: Rockfall, SFM photogrammetry, trajectory, rainfall-effect, Korbous, Tunisia