Use of remote sensing methodology in assessing landslide susceptibility at the upstream catchment of Tamakoshi 3 hydroelectric project, Nepal

Abstract

Identification and evaluation of potential large scale landslides within the catchment of a hydropower project is vital to secure the dam, reservoir and downstream areas. Digital Elevation Models (DEMs) used in a Geographical Information System (GIS) allows regional scale landslide susceptibility assessments.

DEM resolution should be selected to fit the purpose and scale of analysis.Regional scale assessments require lower resolutions than detailed slope studies. However, to identify discernible morphological features related to large scale landslides the resolution must be high enough. Highest resolution achieved by a non-commercial DEM is the ASTER GDEM, with a 15m resolutionwhile state of the art High Resolution DEMs (HRDEMs) can have resolutions better than 1m.

Orientation of discontinuities can be found from a DEM using the software Coltop3D. The data can further be used as input for the software Matterocking to perform an automated kinematic analysis using. Coltop3D and Matterocking can therefore give an indication of unstable areas which can be helpful in planning further field investigations.

The method of Sloping Local Base Level (SLBL) was used to determine potential erodible volumes located in the study area, assuming that slope instabilities most likely will occur where spurs are located in the terrain.SLBL therefore gives an indication of where the potential of gravitational erosion is high. The resolution of the DEM used in this thesis was too low to determine morphological features related to large scale landslides.Therefore the SLBL-method was used to create a map over potential large slope instabilities. From this method the slope east of Gongar were selected for further investigations.

Rockfall susceptibility was evaluated using a qualitative multivariate statistical approach. Three factors were included in the assessment: 1) slopeangle, 2) denudation potential (SLBL), and 3) presence of discontinuities.Each factor was rated between 0 and 2. In addition each factor was given a weight according to its influence on stability. On the basis of these factors a rockfall Susceptibility Index (SI) was found for the entire area to create aregional rockfall susceptibility map.

Numerical modelling of the selected slope at Gongar indicates stress anisotropy at the upper part of the slope. This can cause the formation of tension joint and reduce the shear strength of the rock mass. Applied seismicity indicates increased stress and stress anisotropy for the model. ElasticStrength Factor (SF) indicated failure along a circular failure plane, which was confirmed with a plastic Shear Strength Reduction (SSR) analysis. A Strength Reduction Factor (SRF) of 1.1 was found for the numerical model. Applied seismicity reduced the SRF to 0.94, which indicates slope failure. Foliation favors stability at the Gongar slope, and a distinct discontinuityor a deep seated discontinuity must be present for a slope failure to occur.This thesis cannot exclude the presence of such a structure, which requires further investigations.