GIS-based 3D modelling for landslide hazard assessment:

1. GIS-based 3D modelling forlandslide hazard assessment: Séverine Bilgot Aurèle Parriaux Increasing the objectivity of hazard maps.

2. Introduction • 6% of the Swiss territory affected by landslides • Occurrences and potential damages always increasing Necessity to have objective and easy to update hazard maps

3. Identification of susceptibility factors Intrinsic factors: • Resisting forces • Cohesion • Friction (depends on g, a, j) g, c, j • For rock material: =f(geotype, grade, D, GSI) • For loose material: =f(granulometry, plasticity)

4. Identification of susceptibility factors Intrinsic factors: • Resisting forces • Cohesion • Friction (depends on g, a, j) • Driving forces • Gravity (depends on g, a) • Percolation (depends on i) i= grad h • Measured piezometric level (boreholes, springs, etc) • Spatial distribution of K

5. Identification of susceptibility factors Extrinsic factors: • Vegetation

6. Identification of susceptibility factors Extrinsic factors: • Vegetation • Permafrost melting Depending on: • Location (altitude, aspect) • Susceptibility of the formation =f(granulometry, plasticity)

7. Identification of susceptibility factors Extrinsic factors: • Vegetation • Permafrost melting • Anthropic impact

8. Methodology Predispositionmap Factor of safety Pre-existing data Field data + 3D parameters model Hydraulic gradient Hazardmap

9. Methodology Predispositionmap Factor of safety Pre-existing data Field data + 3D parameters model Hydraulic gradient Hazardmap

10. Integration of pre-existing data Collecting data frompreviousprojects Integratingthem in a geodatabase 

11. Field mapping • Phenomena • Geology • Discontinuities • Formations • Fracturation, grade and GSI • Granulometry and plasticity • Hydrology • Springs and wet areas • Supposed associated aquifer • Vegetation • Structures

12. Methodology Predispositionmap Factor of safety Pre-existing data Field data + 3D parameters model Hydraulic gradient Hazardmap

13. 3D parameters model • Determination of the 3D limits of the structural units • Kriging of the dips and strikes in each structural units • Modelling the rock formations with respect to field observation, boreholes information, profiles and dip and strike interpolation • Modelling the loose formations with respect to field observation, boreholes information, profiles and modelled rock formations • Kriging of the parameters inside each polygone or using the database

14. 3D parameters model

15. Methodology Predispositionmap Factor of safety Pre-existing data Field data + 3D parameters model Hydraulic gradient Hazardmap

16. Hydraulic gradient evaluation • Connectingeachpiezometriclevel, spring or wet area to the correspondingaquifer • Evaluating the geometry of eachaquiferaccording to the repartition of K in the model • Interpolating the hydraulichead in eachaquifer • Calculating the hydraulic gradient

17. Methodology Predispositionmap Factor of safety Pre-existing data Field data + 3D parameters model Hydraulic gradient Hazardmap

18. 3D factor of safety calculation • Calculation of the FoSateachslidinglevelindicated on boreholes • For each (x,y) couple, calculation of the FoSateach intersection withfaults • For each (x,y) withoutcalculatedFoS < 1, calculation of the FoSateach z step, keeping the lowest value

19. Methodology Predispositionmap Factor of safety Pre-existing data Field data + 3D parameters model Hydraulic gradient Hazardmap

20. Predisposition map • Applyingbayesianprobability model to evaluate the weight of frozensoil, vegetation and anthropic infrastructures in the susceptibility to landslides • Calculating the predispositionmap • Verifying the coherence of the predispositionmapwith respect to the map of phenomena

21. Methodology Predispositionmap Factor of safety Pre-existing data Field data + 3D parameters model Hydraulic gradient Hazardmap

22. Hazard map

23. Conclusion: at local scale • Applicable on most of instable areas (alreadytested on 10 sites in Switzerland and 2 in Kyrgyzstan) • Allows full transparency of the results and easy updates • Data canbeeasilyre-used (highinteroperability) • Only one common software to use • Toolboxescanbeusedseparately in otherdomains (naturalresources management, tunnel engineering, etc) • In order to reduce the costs and to providebetterresults, development of equivalent plugins for GRASS (in progress)

24. Indicative maps • Data infrastructure and standardizeprocessallows to change the scaleeasily • Methodology applicable atlargerscalewith simple Quaternary structures (withoutlenses): • Usingboreholes data and databases to evaluate the parameters • Usinggeoportals and aerial photographies for maps of phenomena, vegetation and anthropic impact

25. Thank you for your attention! severine.bilgot@epfl.ch