Application of photogrammetric technique for deriving topographic information in Delta region

Abstract

Digital elevation models (DEM) are crucial for conducting topographical studies in land and water resources management, particularly in flood management. The delta section of a basin is consistently susceptible to flooding due to its low gradient, which poses challenges in analyzing the topographical data. This study attempts to develop a relatively accurate DEM from Cartosat-1 stereo image pairs for delta region of Mahanadi basin. Photogrammetry technique is meticulously performed with minimum RMSE of 2 to 3 pixels. A 10m×10m generated DEM was resampled to 30m×30m DEM for comparison with SRTM DEM. DEM quality was assessed with reference to surveyed cross-section, river embankment profile and ground control points using statistical parameter and accuracy measures such as root mean square error (RMSE) and mean average error (MAE). For river cross-sections, MAE and RMSE of Cartosat- 1 DEM, i.e., 2.70 m and 3.38 m, respectively are found to be lower than that of SRTM DEM, i.e., 4.26 m and 5.32 m, respectively. Statistical and graphical comparison infers that the generated Cartosat-1 DEM is better in representing river bed than that of the SRTM DEM; however, both the DEMs have similar representation for the floodplain elevation.