The Effect of Sunglint Correction for Estimating Water Depth Using Rationing, Thresholding, and Mean Value Algorithms

Abstract

Satellite-Derived Bathymetry (SDB) is an alternative for obtaining shallow water depth data. The existence of images with various resolutions, the availability of a complete image band can develop the extraction results. This method is based on the principle of the satellite's visible band to estimate water depth. The mapping of shallow water depth is dependent on water conditions, both its brightness and surface. When the sensor senses a water object, the reflected reflection comes from the surface, and some sensors cannot penetrate the water object. The sun's position and the sensor's point of view when sensing it results in interference from the water surface (Sunglint). The sunglint effect on the image can be reduced by performing RGB band correction with NIR Infrared. This study aims to demonstrate the effect of Sunglint's correction on three SDB approaches, namely Thresholding, Rationing, and Mean Value on Worldview 3 imagery in Karimunjawa Islands, Central Java. This study's results indicate that the Sunglint correction on Worldview 3 imagery affects the depth extraction results. The best results are shown by Sunglint's correction using the Thresholding approach (B2-B7), which produces the best correlation with R2 of 0.7364 and (B7-B2) with R2 = 0.7351. Contrastingly, the lowest correlation was generated using the Mean Value ((B2 + B7) / 2) approach without Sunglint's correction with R2 = 0.4015. So this research proves that the Worldview 3 image with Sunglint correction can provide bathymetry data, especially in shallow waters.

Keywords

• bathymetry mapping
• sunglint correction
• band ratio
• worldview 3
• karimunjawa