Compared with other ground failure phenomena like liquefaction, earthquake-induced landslides have been studied relatively little. This can be explained by several reasons.

The usual procedure for assessing the landslide risk is to locate the known landslides, to define the major factors controlling the slope failures, and to extrapolate the results to the rest of the area. Landslide distribution maps are commonly drawn from aerial photo interpretation, remote sensing analysis and geological observations. The basic factors influencing landslide distribution are slopes, geology, hydrogeology, anthropological factors (land use) and seismic ground motion. A detailed analysis of individual landslides is necessary to check and to specify the main factors before the results can be applied to the whole area.

Let’s work with a 3D model

Remote sensing imagery provides synthetic views of the studied areas at different scales. These views are the best tools for studying large tectonic events combined with individual slope movements. Their processing and interpretation are combined with the study of seismic and topographic data in digital format in a Geographic Information System (GIS).

The topography is also represented here in digital format files known as Digital Elevation Models (DEMs), that can be used to compute slopes, shaded relief images and three-dimensional views as block diagrams.

The whole country is covered by RESURS 1 multispectral data at 160 m ground resolution. This allows the publication of colour composites as 1:1,000 000 and 1/500, 000 scale image maps. These image maps are useful for defining the general geologic and structural context and plotting all the observations.

Landsat MSS images provide an intermediate ground resolution of 57 m; they have been published as 1/200 000-scale colour composites.

The different images are combined to enhance compositional contrasts, such as in basement lithology, weathering products and alluvial deposits, and detection of active or passive geological structures.

On regional scale
A picture of the area: Chon Kemin

The geological structures and landslides of the Chon Kemin region are best expressed on SPOT panchromatic imagery at 10m ground resolution.

1/50 000 DEMs of Chon Kemin surface rupture area, rockslides and soft sediment landslides associated with the earthquake of 1911 have been processed together with SPOT P imagery and a historical map of surface ruptures and landslides.

The regional scale GIS includes digital geological maps that provide structural and lithological information and will contribute to the geotechnical definition of rock bodies. Slope maps derived from medium-scale DEMs will be included in the risk assessment.

On local scale
A picture of one detail: Bielogorka

A DEM of Bielogorka rockslides has been built by interpolating the topographic contours from a 1/25000-scale map. The results allowed the construction of block diagrams where the orthorectified aerial photo was draped on the cartographic DEM.

These data will be used with a DEM of the area before the landslides to reconstruct the geometry of the events.

Time for field work

Field observation of the existing landslides has been performed to provide detailed information on the geometry of the landslides, the kind of movement (fall, slide, topple, lateral spread, bulging, etc.) and the direction of motion and height of the displaced soil. This information will be used for a comprehensive classification of landslides, which will allow comparison and modelling of the landslide map with the geological, structural and neotectonic context.

Local geophysical investigations were performed at two sites in the northern Tian Shan. Inverting electrical tomography profiles gave detailed information on the geometry of the near surface geology (up to a depth of 30m).

The aim of the project is to compare the landslide distribution map with all the others, including the topographical and geological maps of the region. From this analysis, a correlation will be made between landslide location and the different factors (geology, slope, hydrogeology, peak acceleration, etc.).

From these results and the detailed study of specific slope failures, the key parameters controlling the landslide occurrence in the Tian Shan region will be pointed out and the landslide risk will be evaluated by extrapolating the results to the whole area. The final output will be a landslide susceptibility map, which will be of direct use for planners.

Though this study is limited to the Republic of Kyrgyzstan, many areas of the earth present problems of landslides triggered by earthquake and can be studied by similar techniques.