Visible and NIR reflectance spectra of the components of a LANDSLIDE zone in the Nilgiris, South India
S. Jayaseelan
Centre for geoscience and engineering
Anna University, Madras-600 025 India
Email: ssanjeevi@ annauniv.edu
S. Sanjeevi
Centre for geoscience and engineering,
Anna University, Madras-600 025 India
Email: ssanjeevi@ annauniv.edu
Abstract
Landslides can occur without any prior indications, thereby causing considerable damage
to human beings, vegetation and landscape. Such unpredictability demands development of tools
and techniques to study the landslides in detail. Landslides and landslide-prone areas can be
identified better, and in a faster manner using remote sensing techniques. The majority of the
methods, however, have been the interpretation of aerial photographs and satellite image.
This discussion paper is concerned with the assessment of remote sensing techniques to
map the components of a landslide zone in the Nilgiris south India. Marappalam landslide zone is
selected for this study using satellite images and for studying the reflectance characteristics of the
components namely, the slide zone, rock outcrops, soil cover, vegetation, sapprolite/ lithomarge
etc. These are studied in the visible and near - infrared regions (0.4µm to 1.1µm) of the Electro
Magnetic Spectrum (EMS).
Analysis of multi temporal satellite image (SPOT 1 - MLA and IRS 1B - LISS2) is done to
identify and delineate the Marappalam landslide. The post slide satellite image reveals that the slide
is located at the intersection of two conspicuous lineaments. As regards the reflectance spectra, it
is observed that the younger landslides have higher reflectance in the visible and near-infrared
regions. Low NDVI (Normalised Difference Vegetation Index) values are indicative of landslide
patches, slide-prone areas or outcrops, thus suggesting the usefulness of NDVI in the study of
landslides. Thus spectral sensing and multitemporal image interpretation techniques may form an
effective basis for the study of landslides and efficient disaster management.
Intorduction
This paper is concerned with the study of the spectral reflectance characterstics of the
components of a landslide zone in the Nilgiris, south India. This study would form the basis for
understanding the exact nature and manifestation of landslides and landslide-prone areas in.multispectral image data. This study assumes significance in the present context of an increasing
risk of landslide occurrence, which demands the use of versatile tools such as remote sensing to
mitigate the hazards.
Aims and Objectives
- To evaluate the applicability of spectral remote sensing in landslide studies,
- To understand spectral behavior of stable and unstable slopes in order to identify them in
satellite image data,
- To relate the spectral properties of the components of the landslide zone to their image
characters,
- To understand the relationship between the percentage cover of land-cover features in a
landslide zone and the spectral signature in satellite images, and
- To identify optimal spectral technique that could be used to delineate landslides, stable and
unstable slopes.
Study area and Methodology
Occurrences of landslides are common in Nilgiri district, particularly during the north-east
monsoon. During 1993, about 408 landslides were reported to have occurred of which Marappalam
(Fig.1) is the severest one in terms of magnitude. This slide occurred 26 km away from
Mettupalayam town on the state highway.
Fig. 1. Location map of Nilgiris
Since the Marappalam area and its surrounding area has experienced such landslide
time and again, it was decided to select this area for the present study.
Physiography, Geology and Geomorphology of study area
The general elevation of the study area varies from 1,100 m to 1,600 m above msl. The
Coonoor peak is the highest one at 2091 m above MSL. The Coonoor stream, which drains the area
has a straight to sinuous course. Two erosional surfaces, at altitudes of 2,400m and 2,100m were,
reported by Parthasaradhi (1974). The rocks exposed at the slide area are garnetiferous,
quatrzofelspathic gneiss and pyroxene granulite with EW foliation trend, dipping steeply towards
south to vertical. The trend of foliation is sub-parallel to the slope.
