MULTICRITERIA BASED ROUTE ALIGNMENT IN HILLY TERRAIN USING GIS

Abstract

Transport systems play a significant role in the overall development of any country. The rate of growth of transport demand is generally higher than the rate of growth of economy. The adequacy and efficiency of the transport system in a country is a good indicator of its economic prosperity. The road transport in India has increased tremendously over the last few decades, which has necessitated to search for new methods and alternatives that ensure efficient, feasible and faster methods of locating the best route alignment. India, bearing the tremendous pressure of population growth, feels its importance, and therefore a large share of country's budget is now allocated for the development of transportation infrastructure. India has a fairly good road network, except in hilly areas. These hilly areas, especially Himalaya, have extremes of climate conditions, difficult and hazardous terrain, topography and geology. In such areas, due to mountainous terrain, development of rail and even air transport system is difficult due to the prohibitive cost of construction. Owing to their spread, elevation and general ruggedness, most of the regions in the Himalayas are not easily accessible; scattered roads are providing the only means of transport. The demand for roads in hilly areas has increased considering the infrastructural development requirements of these areas, and also the strategic needs. Conventional methods of surveying, like plane table, compass, levelling etc., take lot of time and are more expensive for the extraction of surface profile information for new road alignment. This type of survey is particularly more cumbersome in hilly terrain. The conventional methods of route alignment have gradually been replaced since 1960's by the techniques, which use aerial photographs, and later in 1970's by remote sensing techniques which use satellite imagery. In recent years, it has been established that the use of Geographic Information System (GIS) based spatial database is an efficient and economic method to be employed in planning and management of engineering projects. Remote sensing and GIS techniques, by virtue of their numerous advantages, appear to be an automatic choice for route planning, which requires efficient processing, interpretation and analysis of a large number of spatial data from a variety of sources. The present research has been focused to demonstrate the capabilities of GIS and remote sensing techniques to delineate an effective, economic, and feasible route alignment between Uttarkashi and Kedarnath towns using multi-criteria approach. Bounded by 30° 30'N to 31° 00' N latitudes and 78" 15' to 79° 05' E longitudes, the area of 3811.15 km2 is covered by parts of three districts i.e., Uttarkashi, Tehri Garhwal and Rudraprayag with altitudes varying from 800 to 6500m. In this study area, three routes already exist between Uttarkashi and Kedarnath; (i) Uttarkashi-Tehri- Gaurikund- Kedarnath with a total length of 260km, (ii) Uttarkashi- Rudraprayag-Kedarnath with a total length of 263km, and (iii) Uttarkashi-Deoprayag-. Kedarnath with a total length of 301km. By proposing a shorter and more direct route between these towns, tourists and others normal traffic will benefit. For generation of Triangular Irregular Network (TIN), Survey of India topographic maps. (1:50,000) has been digitized at 40m contour interval, interpolated and resampled to generate slope and aspect maps. Lithological and faults features have been extracted from geological map. The lineaments have been interpreted from a 3 X 3 edge-enhanced filter applied to IRS IC LISS-III image. The lineament buffer map and faults buffer map have been prepared. The drainage features have been digitized from topographic map and upgraded using IRS IC LISS III and PAN imageries for preparation of drainage density map and drainage order map. Landuse/landcover map has been generated from LISSIII image by supervised classification technique. Ten landuse/landcover classes considered are water, dense forest, in builtup area, sediment, agriculture, fallen land, medium dense forest, scrub, barren land and snow. A Digital Elevation Model (DEM) and Normalized Difference Vegetation Index (NDVI) have been included in the classification process to reduce the effect of shadows in the region and to improve the separability among various classes. For separability analysis, Transformed Divergence (TD) has been used. From multisource classification, it has been found that the combination of six bands (LISS-I1I four bands, DEM and NDVI data) using Maximum Likelihood Classifier (MLC) produced the maximum overall accuracy of 92%. In this study, Landslide Hazard Zonation (LHZ) map has been prepared using weighted rating scheme. Slope, aspect, lithology, landuse, drainage, lineaments, faults and soil are used for this purpose based on the importance of each factor influencing landslide hazard. Each sub-class is given rating at 0-9 scale in an increasing order of hazard; zero indicates low hazard and 9 indicates high hazard. All the rating classes are multiplied by weight at 1-9 scale, lowest value indicates low risk of landslides and vice versa. For integration of all layers, arithmetic overlay model is used. The input data layer rating is multiplied by corresponding weights and added for all the layers to get the landslide potential index, which is classified into four classes; very low, low, medium and high. The accuracy waschecked by PAN imageries and existing landslides in the field. A multicriteria tool was developed using Visual Basic programming. In this tool, rank sum, rank reciprocal, rank exponent, ratio estimation and AHP methods are implemented. Multicriteria weighted methods are not available in an integrated manner in any GIS software. An attempt is made to use this tool to get weights by all the methods. The basic inputs for this tool are route parameters and their ranking order. Tool provides a wide variety of choices to compute weights. In this study, all the five methods are used to identify five different routes. IV For multi-criteria based route identification, all the route alignment factors are reclassified at 0-9 rating scale and multiplied by corresponding weight of each parameter to generate combined weight (discrete weight) map using spatial analyst module of Arc GIS software. The resultant weighted maps of five methods have been used for finding best shortest path between Uttarkashi and Kedarnath towns. The shortest path algorithm, called spread algorithm, identified the least weighted path between these two locations. After finding five routes, the best route alignment has been selected based on the minimum length of route, minimum number of bridges and culverts, maximum number of habitations connectivity within 5km buffer zone, maximum number of tourist places connectivity within 5km buffer zone. Cost analysis is done to all five routes based on total length of bridges, culverts and length of road. The best results are achieved by ratio estimation method, followed by AHP method. By using ratio estimation method, a route length between Uttarkashi and Kedarnath towns is only 75.79 km long (including existing road which is common with the proposed alignment over a length of 23.16 km), and therefore, the new road required is 52.63 km, while the existing route between Uttarkashi and Kedarnath is 260 km long. The cost of construction of the optimum route identified by this method is 216.36 million Rupees (approx US$ 54, 09,000). A sensitivity analysis is done for ratio estimation method. It is observed that slope, landuse and drainage parameters are more sensitive (in that order) as compared to other parameters. Among the other parameters, LHZ, lineaments, faults, lithology is found to be less sensitive