An Integrated Approach for Landslide Susceptibility Mapping Using Remote Sensing and GIS

A methodology for landslide susceptibility mapping using an integrated remote sensing and GIS approach is presented. A part of the Darjeeling Himalaya was selected for the model execution. IRS satellite data, topographic maps, field data, and other informative maps were used as inputs to the study. Important terrain factors, contributing to landslide occurrences in the region, were identified and corresponding thematic data layers were generated. These data layers represent the geological, topographical, and hydrological conditions of the terrain. A numerical rating scheme for the factors was developed for spatial data analysis in a GIS. The resulting landslide susceptibility map delineates the area into different zones of four relative susceptibility classes: high, moderate, low, and very low. The susceptibility map was validated by correlating the landslide frequencies of different classes. This has shown a close agreement with the existing field instability condition. The effectiveness of the map was also confirmed by the high statistically significant value of a chi-square test. Introduction Landslides are one of the destructive geological processes which cause not only enormous damage to roads, bridges, and houses but also lead to loss of life. Hence, there is a need for landslide susceptibility mapping for identification of potential landslide areas. Landslides are the result of complex interaction among several factors, primarily involving geological, geomorphological, and meteorological factors. The spatial information related to these factors can be derived from remote sensing data, ground based information, and several other data sources. Digital image processing of remote sensing data has a greater degree of objectivity and reproducibility than that obtained by traditional visual interpretation approach. Geographic information systems (GIS) are a very powerful tool for the integration of different types of data. Over the past few years, there have been significant developments of GIS for spatial data analysis. Efficient landslide susceptibility mapping can be carried out by combining GIS with image processing capabilities. The present study is an attempt towards development of a methodology for landslide susceptibility mapping. This involves the generation of thematic data layers and their spatial analysis within the Darjeeling Himalaya, India. The study also aims to ascertain the utility of remote sensing and GIS. Trends in Landslide Susceptibility Mapping In recent years, assessment of landslide susceptibility in the form of hazard zonation has been attempted in a wide variety of environments and using diverse approaches. The different An Integrated Approach for Landslide Susceptibility Mapping Using Remote Sensing and GIS S. Sarkar and D.P. Kanungo methodologies developed were influenced by the scale of analysis, the availability of input data, and the required details of the hazard map (Brabb, 1984; Nilsen et al., 1979; Varnes, 1984; Wagner et al., 1988; Pachauri and Pant, 1992; Anbalagan, 1992; Sarkar et al., 1995; Mehrotra et al., 1996). These methods are based on the integration of information about the spatial distribution of the factors identified to be important in assessing slope instability. Notable contributions from Carrara (1983) and Chung et al. (1995) presented multivariate statistical models for assessing landslide hazards. Other statistical methods, such as the use of information models and fuzzy set theory, were also applied in this context (Yin and Yan, 1988; Juang et al., 1992; Jade and Sarkar, 1993). To evaluate the combined effect of the factors, the use of GIS in the modeling of landslide hazards using many different parameter maps was attempted by several researchers (Carrara et al., 1991; Kingsbury, 1992; Westen, 1994; Nagrajan et al., 1998; Gupta et al., 1999; Dhakal et al., 2000). Study Area and Geology The Darjeeling Himalaya, encompassing a total area of 3000 km, rises abruptly from the alluvial plains of North Bengal and attains a maximum elevation of about 2600 meters. The area lies between Sikkim on the north, Bhutan towards east, and Nepal towards west. This study focused on Darjeeling hill which lies within latitude 26°56 to 27°8 E and longitude 88°10 to 88°25 N and covers an area of about 254 km (Figure 1). The area is traversed by many ridges and valleys. The major ridges are at elevations of 2000 m to 2600 m while valleys range in elevation from 500 m to 1000 m. The maximum elevation in the area occurs at Tiger hill at 2584 m. The area is dominated by slopes ranging from 15° to 35° while steep slopes of greater than 35° occupy a much smaller area. In general, gentle slopes of 0° to 15° were found on ridges and at places in regions of lower relief. The main land-use practice in the area is tea plantations. The agriculture lands present are mostly associated with habitation. The area is dominated by thick forests, particularly in the eastern part. The annual rainfall in the area is on the order of 3000 to 6000 mm. The rainfall pattern is highly seasonal with the majority falling during the monsoon months of June to October. A maximum monthly rainfall of 580 mm was observed in July 2001 at Darjeeling. P H OTO G R A M M E T R I C E N G I N E E R I N G & R E M OT E S E N S I N G May 2004 6 1 7 Geotechnical Engineering Division, Central Building Research Institute, Roorkee, India ([email protected]; [email protected]). Photogrammetric Engineering & Remote Sensing Vol. 70, No. 5, May 2004, pp. 617–625. 0099-1112/04/7005–0617/$3.00/0 © 2004 American Society for Photogrammetry and Remote Sensing 02-108.qxd 4/5/04 10:46 PM Page 617