Relationship between spatial distribution of landslides in Bhutan and geographical parameters

Landslide is a disaster which is affecting countries with high-relief topography and large-amount precipitation. A typical example is Bhutan. Major roads are sometimes blocked by landslides caused by monsoon-derived intensive precipitation events. To understand where landslide is prone to occur in Bhutan, we need geographical assessments focused on both of spatial distribution of past landslide mass movements and that of precipitation. In this study, landslide features were delineated from high-resolution satellite imagery and Digital Surface Model (DSM) collected by the Advanced Land Observing Satellite (ALOS) operated by the Japan Aerospace Exploration Agency (JAXA). We define three domains located in different river basins as our study site. They are located in the Mangde river tributary, the Wang river tributary, and the Drangme river tributary. Multiple geographical parameters were calculated from ALOS-derived DSM data; i.e. elevation, slope angle, distance from the river, curvature, topographic wetness index (TWI), stream power index (SPI), and sedimentary transport index (STI). Frequency ratio (FR) was calculated by the number of pixels in each class of parameter to evaluate the geographic conditions that are known to be associated with landslides.

The results show that the FR was greater in places with (1) lower elevation, (2) closer distance from the river relative to the entire watershed, and that landslides are more likely to occur under these conditions in all three study areas. The larger FR at lower elevations is presumably due to other factors, such as weathering, which are affected by elevation. The finding that the FR was larger in the area closer to the river is explained by a hypothesis that erosion of the lower part of the slope reduces the stability of the slope and makes landslides more likely to occur. In addition, representative values of geographical parameters in the study areas were compared with each other. The Drangme river tributary area with the smallest elevation and distance from the river has the largest percentage of landslide features. They indicate the same trend as (1) and (2). Thus, elevation and distance from river are important parameters to know landslide prone area in these districts.