인공위성을 이용한 북서태평양 연안 해역에서의 적조 및 유해성 조류 증식에 대한 탐지

Abstract

Cochlodinium polykrikoides (p) is a planktonic dinoflagellate known to produce red tides responsible for massive fish kills in Korean coastal waters, particularly in summer and fall seasons. These algal blooms with higher concentrations appear to span several weeks in summer and fall seasons and are widespread in most of the South Sea coastal bays and neighboring ocean waters. Accurate detection of these algal blooms using satellite data are very challenging because of eventual interference of the coastally-influenced suspended sediments and dissolved organic matter concentrations with the standard retrievals algorithms. This study attempts to evaluate the previously developed methods by Ahn, Shanmugam et al. (2006), for detecting the red tide algal blooms from SeaWiFS and Landsat-7 ETM+ image data in Northwest Pacific waters. First, we examine whether the OC4-derived chlorophyll from SeaWiFS imagery provides a means of detecting red tides in turbid coastal waters and then seek whether the classical spectral techniques such as Minimum Spectral Distance (MSD) and Forward Principal Component Analysis (FPCA) adopted from Shanmugam (2002 and 2006) are applicable for this application. SeaWiFS chlorophyll seemed to be useful in locating the areas of red tide blooms, but uniquely identification of Cochlodinium.p from turbid and non-bloom waters remained ineffective with these data. Furthermore, highly discolored waters around the river mouths, estuaries and coastal areas often led to very high chlorophyll falsely indicating Cochlodinium.p blooms in these areas. In contrast, the MSD and FPCA attempted on SeaWiFS and Landsat-7 ETM+ image data seemed to be powerful in tracking the complex patterns of the Cochlodinium.p blooms from surrounding non-bloom and sediment-dominated waters. From this investigation, we believe that the MSD and FPCA coupled with field data for the early detection of HABs can provide state managers with accurate identification of the extent and location