REMOTE SENSING AND GIS APPLICATIONS
IN SEISMIC HAZARD STUDIES IN NW HIMALAYAS.

Prasanjit Dash
IIT-Kanpur, TU-Berlin
Submitted May, 1999.

        ABSTRACT:
                                        Integrated approach of remote sensing and  Geographical  Information  Systems
                                        (GIS) is highly useful in evaluation, management and monitoring of natural hazards.
                                        Routine use of remote sensing data and its analysis of a hazard prone area, helps in
                                        monitoring the changes in surfacial feature. The digital image processing techniques
                                        enable the user to get the desired information in a more reliable, quicker and easier
                                        way. In the present study, digital remote sensing has been used in the mapping of
                                        lineaments. Various approaches have been attempted to map the lineaments of the
                                        Chamoli area in the NW Himalayas. Using the lineament map, the orientation of the
                                        stresses in the study area have been determined. Using GIS module, integration of the
                                        past earthquake data, lineaments, stress pattern and the geological information have
                                        been attempted to evaluate the seismic risk in the study area.
        The project work included construction of lineament fabric of the area using Sobel operator,
construction of tectonic map of the area using self-designed 101x101 and 31x31 equdimensional
zero-sum kernels in ERDAS MODEL MAKER, construction of length weighted rose diagrams for
orientation of lineaments using 'C' program and MATLAB software, estimating stress pattern in
different  windows as per concepts of  Mohr's  theory of failure by  'C'  program and MATLAB,
construction of  DEM  using USGS  global  digital  topographic  data  and processing through
ARC/INFO and  image draping  by ERDAS IMAGINE. All the thematic information along with
earthquake data has been put into GIS platform. Finally, in the light of all the available information
neo-tectonic study and seismic risk of the area has been done. Also the most probable reason for
a recent earthquake (28th Mar., 99) in Chamoli has been discussed.
         This work is based on edge-detection techniques of digital image processing,
using various software and  computer programming in 'C'. The satellite data being used
is Indian Remote Sensing(IRS) 1B, LISS1 data having a ground resolution of 72.5 Mts.,
and in four multispectral bands.Along with it, digital elevation model data and earthquake
epicentre data are also used in various layers. NH-44-5 series toposheet of USGS is used
for geocoding and other references along with field study and use of GPS.
         The project  goal included:
1.    Successful lineament detection on satellite imagery in a time and cost limited survey and producing
       tectonic map.
 2.    Emphasis has been given on the spatial edge detection techniques like the Sobel edge detection method.

 3.    Also self-designed zero-sum equidimensional kernels of 31x31 and 101x101size are used for
        mega-lineament detection, as the Sobel operator gives a broken pattern of the edges.

 4.   The lineaments are subjected to further statistical processing and rose-diagram plotting and effort has
        been made to find out the stress pattern of the area, using Mohr's concept of failure.

 5.    Earthquake data are plotted as per there geographical location and effort has been made to find there
        relationship with the major thrusts and faults and the stress pattern of the area.

 6.    For better visual interpretation, DEM has been constructed and various layers are merged to one
        projection.

        The various software used are ERDAS 8.3.1, ARC/INFO, PHOTOSHOP, R2V,
        MATLAB etc. Also computer program in 'C' is used for the statistical processing.

Comments, suggestions, questions, help : Prasanjit Dash