Multisource data integration to investigate a 3D landslide morphology affecting an urban area: case of Bordj Bou Naama (west Algeria)

Situated in the northern part of Algeria, Ouarsenis is one of several area affected by landslide phenomena. Indeed, 60 houses, a stadium and a lot of sports infrastructures are affected by the landslides. In this perspective, we have chosen a model of landslide with an area of 166 700 m2 affecting the southern part of the Bordj Bou Naama city. In order to characterized the landslide structure, we used multisource data (geological, topographical, geophysical and geotechnical). For modeling the 3D landslide surfaces we used three geometric models which are generated from different techniques of interpolation as Inverse Distance Weight (IDW), Minimum Curvature (MC) and Kriging (KO) and applied to the same input data set. The root of the mean square error (RMSE) and visual appearance of the morphology are used to select the best model. Indeed our results show that the KO represents the best model that gives a good result. Quality control is also performed to ensure that the model is suitable for hydro-mechanical modeling. This model show that the total volume of soils moved is ~9.8*105 m3, in which the volume of geological levels is 828 500 m3, 143440 m3 and 11 434.32 m3 respectively for the embankments, colluviums and Brown shale. The bleu shale is far from to be affected by this landslides, indeed the deepest zone that the rupture area affects is located at 12 m of depth. These results seem to be very important in order to plan remediation work in this area.