This chapter aims to develop landslide susceptibility maps for the Sikkim state in India by combining the analytical hierarchy process, geographic information systems, and remote sensing. The delineation of the landslide susceptibility maps has taken into consideration a variety of data such as density of lineament, slope, lithology, aspect, land cover and land use, road buffer, rainfall, and drainage density. Using both Landsat 8 and ground data in a GIS framework, spatial distribution of maps and map layers of required themes were produced. The appropriate weights based on the Saaty's scale were given to these thematic layers in accordance with their respective significance in the occurrence of landslides in the study area. According to the study area's demarcated landslide susceptibility map, the risk levels were very low (12.52%), low (21.12%), moderate (8.05%), high (31.13%), and very high (27.18%). The accuracy of the study region is computed using the AUC curve using the AHP model landslide map and inventory map, which shows good result with 70% accuracy.
TopIntroduction
Natural hazards comprises of Landslides that can occur on both land and water; they are common throughout the world and are primarily concentrated in hilly areas. Landslides are caused by a variety of phenomena, including earthquakes, rainfall intensity, snow melting rate, and human involvement. (Hutchinson, 1988) defines a landslide as the relatively quick mass wasting process that causes the movement of large masses of rock, rubble, or earth downward when one or more external stimuli are present. The landslide can be described as “the mass movement of soil and rock down a slope” (Couture, 2011). Regarding the kind of debris that slides down a slope, this definition of a landslide is more inclusive. Landslide susceptibility is defined as a quantitative or qualitative evaluation of the classification, volume (or area), and spatial distribution of landslides that exist or may occur in a given area. Landslides usually result from a combo of geo - environmental phenomena, as explained by (Varnes, 1984) and (Hutchinson et al., 1985). The Himalayas provide shelter for millions of people from various nations. About 13% of India's territory is exposed to various levels of landslide risk (Geological survey of India, 2009), which also includes the north east state Sikkim. The North East Himalayas region of seven sister states accounts for 0.18 million square kilometres of the total area. Landslides are a common natural phenomenon in India, particularly during the monsoon season. Landslides are sensitive to rock quality, slope geometry, topographical and natural conditions, and disturbances. The landslides can be classified as fall (separation of soil), flow (soil flow), creep (down slope movement of soil), debris flow (quick development of free soil), lahar (volcanic eruption cause ice melting), mudflow, and slides. Rotational slides and translational landslides are the two main types of landslides.
Sikkim is vulnerable to natural disasters due to heavy rainfall, weak geology, and earthquakes. Due to its land and climatic conditions, Sikkim is prone to landslides. It is effected by land cover and land use by humans. There is loss of infrastructure and invaluable human life. Due to the complex geological setup and extensive rainfall in the terrain, the severity of damages caused by there are a lot of abrupt mass movement in several areas of the Sikkim. Every year, thousands of people are killed in landslides in Sikkim Himalaya (Bhasin et al., 2002). In the year 1968, landslides in Sikkim resulted in the deaths of almost 36,000 people (Kaur et al., 2019). Rainfall, seismic activity, and anthropogenic activities are the main controlling factors for slope instability (V. Kumar et al., 2018; Sonker et al., 2021; Taalab et al., 2018; van Westen et al., 2008; Yu & Huang, 2006). Around the world, both directly and indirectly, landslides have a huge effect. In response, this has led to a rise in injuries and fatalities, as well as negative effects on the economy and ecology, in the nations where landslides are most frequent.
Many landslide susceptibility maps have been developed in recent years the use of statistical approaches based on GIS such as Weights of Evidence (WoE) and Frequency Ratio (FR) model. The fact that the findings of these models demonstrated excellent results with great consistency, that they are relatively easy to use (Akgün & Bulut, 2007; Choi et al., 2012; Dahal et al., 2008; Fayez et al., 2018; R. Kumar & Anbalagan, 2016; Lee & Pradhan, 2007; Mersha & Meten, 2020; Park et al., 2013; Pradhan et al., 2010; Sonker et al., 2021; Vakhshoori & Zare, 2016; Yilmaz, 2009).