Study of road-realignment using geospatial techniques and 3D flood inundation model creation
Student name: Ms Sukanya D Saikia
Guide: Dr Anu Rani Sharma
Year of completion: 2013
Host Organisation: North Eastern Space Applications Centre (NESAC), Umiam
Supervisor (Host Organisation): Mr Victor Saikhom
Abstract: This study demonstrates and investigates the use of Photogrammetric and GIS techniques to carry out digital
elevation model (DEM) & orthoproduct generation and road realignment study in a small region in Umiam,
Meghalaya. National Highway 40, connecting Jorabat and Jowai in the state is the road through which public
commute daily. A reservoir, Barapani Lake, is situated in this region, which falls in the Ribhoi and East Khasi
Hills districts of Meghalaya. A very old dam present in the region holds high risk of breakage. This may cause
severe loss to the public as the National Highway is one of the busiest connecting route. This study has been
adopted to address such issues and take steps to avoid such situations and if occurs, bring up with ways that
can mitigate the disastrous consequences. Digital Elevation Model generation is done using LPS software in
ERDAS 2011. Stereopairs of CARTOSAT – I (2.5m spatial resolution), which are provided with RPC files, are
imported into a block in the LPS environment. With the RPC provided, interior orientation has been done.
Exterior orientation involves automatic tie point generation. It is also done by taking tie points and ground
control points manually with reference images & DEM. Bundled block adjustment, also called triangulation, is
performed with an RMSE of 0.3046589. The DEM generated was edited in Leica Terrain Format using Terrain
Editor module of LPS, with mass points, breaklines and contours simultaneously displayed. A new refined DEM
of 10m is generated and an orthorectified product is obtained from it. Every point in this image is
planimetrically corrected. Seamless Carto-DEM and ortho-rectified Cartosat-1 data are prepared. LISS IV
image (5.8 m spatial resolution) is obtained and the bands are layer stacked. This image is orthorectified with
respect to the CARTOSAT –I orthoimage and Carto-DEM. Digital image processing techniques are applied to
get an enhanced image. The CARTOSAT – I orthoimage and the ortho-rectified LISS-IV-MX are merged using
all the techniques available. Brovey Tranform algorithm with Bilinear interpolation as the resampling method
is found to be the best method to give the best visually appealing image. LULC classification is done in the
better contrast fused image and road features are being extracted. A toposheet of 1:50,000 scale is used to infer
the names of the features into the attribute table. Spatial Analysis is used to identify spatial relationships among
different features. Slope obtained from the 10m DEM and the LULC are given weights and cost raster has been
prepared. Two sources of polygon type and two destinations of point type are prepared. Using the cost raster
and source as inputs, cost distance raster and direction raster are prepared which when fed as inputs with the
destination shapefile gives the shortest route. Two least cost paths have been prepared as per the feasibility.
The shortest path refers to the route incurring the least cost in terms of time, money and any other kind of
impedances. The two alternative roads along with the existing road, are depicted in a 3D environment in
Virtual GIS of ERDAS 2013. Flood levels at 800m to 900m at intervals of 20m are shown in the 3D
visualization. The 3D visualization provides suggestions that can be forwarded or recommended to the decision
makers. Finally a 3D fly through has been prepared.
Keywords: Stereopair, Digital Elevation Model, Orthoimage, Bundled Block Adjustment, Breaklines.
