Satellite detection of HAB in the southern coast of Korea: A Class-based Approach

Abstract

Since 1994, blooms of Cochlodinium polykrikoides (a dinoflagellate) have occurred nearly everysummer in the southern coast of Korea and have caused great losses in the Korean aquacultureindustries. Early detection of C. polykrikoides blooms could be useful in mitigating the damagesfrom the HABs (harmful algal blooms). However, it is very difficult to detect the HAB using satellitedata in optically complex coastal waters such as the southern coast of Korea that is affected byhigh concentration of suspended sediments and colored dissolved organic matter as well asphytoplankton. In this study, using a family of reflectance spectra of C. polykrikoides, we attemptto apply fuzzy classification technique to MODIS reflectance data for HAB pixel detection. Thistechnique may allow identifying distinct optical water types in the southern coast of Korea. Wediscuss the potentials and limitations of the method.ooms could be useful in mitigating the damagesfrom the HABs (harmful algal blooms). However, it is very difficult to detect the HAB using satellitedata in optically complex coastal waters such as the southern coast of Korea that is affected byhigh concentration of suspended sediments and colored dissolved organic matter as well asphytoplankton. In this study, using a family of reflectance spectra of C. polykrikoides, we attemptto apply fuzzy classification technique to MODIS reflectance data for HAB pixel detection. Thistechnique may allow identifying distinct optical water types in the southern coast of Korea. Wediscuss the potentials and limitations of the method.