Atmospheric correction and analysis on the suspended sediments in Gyeonggi-Bay using GOCI and Landsat ETM+

Abstract

Most of the western coastal area in Korea peninsular is turbid water due to terrestrial inputs and sediment resuspension whereas East Sea is clear water by deep depth and no resuspension. The large spatial and temporal variability in coastal environment such as chlorophyll and suspended sediment are difficult to analyze by cruise sampling. In recent years, satellite images are used to monitor the coastal environment change. For analysis on the environment change of coastal water using satellite images, atmospheric correction is essential to remove the aerosol and rayleigh scattering. In this study, we carried out the atmospheric correction of Landsat and GOCI images using NIR/SWIR, COST model and modified MUMM algorithm respectively. And then we analyzed the variation in SS in Gyeonggi-Bay using Landsat and GOCI images. Modified MUMM algorithm for GOCI atmospheric correction had been tested by Lee et al. (2012). In turbidity area, accuracy of modified MUMM has been significantly improved. Landsat image corrected atmospheric effect using NIR/SWIR band and COST model. Water-leaving radiance of Landsat in clear water area was underestimated. However, water-leaving radiance of Landsat in turbidity area was a good match to the modified MIUMM algorithm result. Daily variations in GOCI-derived SSC showed low values during flood time and high values during ebb time. Seasonal variations in Landsat-derived SSC showed high valul environment such as chlorophyll and suspended sediment are difficult to analyze by cruise sampling. In recent years, satellite images are used to monitor the coastal environment change. For analysis on the environment change of coastal water using satellite images, atmospheric correction is essential to remove the aerosol and rayleigh scattering. In this study, we carried out the atmospheric correction of Landsat and GOCI images using NIR/SWIR, COST model and modified MUMM algorithm respectively. And then we analyzed the variation in SS in Gyeonggi-Bay using Landsat and GOCI images. Modified MUMM algorithm for GOCI atmospheric correction had been tested by Lee et al. (2012). In turbidity area, accuracy of modified MUMM has been significantly improved. Landsat image corrected atmospheric effect using NIR/SWIR band and COST model. Water-leaving radiance of Landsat in clear water area was underestimated. However, water-leaving radiance of Landsat in turbidity area was a good match to the modified MIUMM algorithm result. Daily variations in GOCI-derived SSC showed low values during flood time and high values during ebb time. Seasonal variations in Landsat-derived SSC showed high valu