Open AccessMapping Kuwait bathymetry using passive multispectral remote sensing
Author(s) -
Jasem A Albanai
Publication year - 2021
Publication title -
maǧallaẗ al-kuwayt li-l-ʿulūm
Language(s) - English
Resource type - Journals
eISSN - 2307-4116
pISSN - 2307-4108
DOI - 10.48129/kjs.v48i4.8978
Subject(s) - bathymetry , remote sensing , multispectral image , echo sounding , mean squared error , geology , satellite , environmental science , oceanography , statistics , mathematics , aerospace engineering , engineering
Mapping bathymetry is essential for many fields, including science, engineering, and the military, among others. Bathymetry is extremely important in the scientific field because it is linked to many physical and environmental issues such as coastal erosion, sea-level rise, and water quality. Traditionally, conventional methods, such as pre-measured cable passage, were used to estimate depths. Lately, echo-sounder assessments were used on hydrograph ships. This method is effective, but it is very costly in both economic and time terms. Remote sensing technology provides modern methods for mapping bathymetry, such as the use of active and passive remote sensing. Many satellite sensors cover multispectral bands. Some are commercial, such as IKONOS and WorldView, while others are freely available, such as Landsat 8 and Sentinel-2. In this study, Landsat 8 (15 meters spatial resolution) was used to estimate the depths of the waters of Kuwait, an Arabian Gulf country located on the Northwestern side of the gulf. Ground truthing points (GTPs) were used to build a bathymetric model of Kuwaiti territorial water (KTW) using the ratio transform algorithm (RTA) applied on Landsat 8 data. The results showed a good ability of Landsat 8 and RTA to estimate the depths of Kuwait’s waters, where the relationship between the derived model from Landsat 8 and the GTPs was positive (r2 = 0.9634). Meanwhile, the accuracy of the derived bathymetric model was evaluated by computing the Root Mean Square Error (RMSE = ± 1.66 meters) and Mean Absolute Error (MAE ± = 1.29).