ESTIMATION OF SHALLOW WATER BATHYMETRY ALONG THE NORTHERN COAST OF BAY OF BENGAL: A REMOTE SENSING-BASED APPROACH

Abstract

Bathymetric information is vital for navigational safety and is utilized for many more activities. Remote sensing data and satellite images are widely used these days to determine shallow coastal areas’ bathymetry at a low cost. This study reviewed different methods for satellite-derived bathymetry and selected the ratio transform method to apply to Landsat 8 imagery. Two images covered the northern coastal region of the Bay of Bengal. They were processed using a ratio-based algorithm calibrated with reference data. A hydrographic chart from the Bangladesh Navy depicted a general overview of the study area. Another chart from the BIWTA, along with GEBCO gridded bathymetry data, was used as reference data for the study. The ArcGIS and ENVI image processing software was used to process and analyze satellite imagery. The correlation between the satellite-derived bathymetry and reference data was also studied at the end of the study. The mean absolute deviation, mean squared error, and root mean square error were also evaluated. Both algorithms were able to extract bathymetry up to a depth of 12 meters with minimal errors. Satellite Derived Bathymetry (SDB) from the Central Coast resulted in an R-squared value of 0.82 with a Mean Absolute Deviation (MAD) of 0.89. SDB from the Western Coast had R2 of 0.81 and MAD of 1.16. After interpolating the algorithm results, contour lines were also generated to visualize the bathymetry. The deviations and irregularities in the contours resulted due to the high turbidity of the coastal waters of the Bay of Bengal.