Remote Sensing of Chlorophyll Concentration in a Bay from Landsat TM Data

摘要

Major inner bays in Japan are in chronic eutrophication and suffering from so-called red tide and blue tide or anoxia in summer, which often cause serious damages to coastal fisheries and environments. In order to improve the water quality and reduce the damage, various kinds of countermeasure have been carried out, which include reducing inflow load of nitrogen and phosphorus, removing organic bottom mud and covering the bottom mud with sand etc. Continuous monitoring of water quality is very important to grasp the effects of these countermeasures and to take proper procedures. Mikawa Bay (area = 604km~2, mean water depth = 9.2m) is a shallow and closed inner bay connected to Ise Bay, where the water quality problems have increasingly been serious due to eutrophication. The prefectural government has continued water quality measurement every month at about 20 locations in the bay. More frequent measurements at more locations are however necessary to understand the dynamic behavior of water quality in the bay In the ocean, satellite images have been used for analyzing distributions of water surface temperature, chlorophyll concentration (Chl-a) and turbidity etc. The remote sensing of water quality in inner bays are also expected because satellite images regularly supply simultaneous information of wide area. Water quality of inner bays is usually evaluated through various indices such as COD, SS and Chl-a etc. Among these, Chl-a is the most important index for eutrophic waters. Estimate of Chl-a in the ocean from satellite images has been tackled by many researchers. In particular, various types of algorithm has been developed and verified for the CZCS sensor on NIMBUS7. Most of them relate Chl-a to a ratio between radiance values of two different bands (Gordon et al., 1983). These algorithms are however applicable only to relatively clear water in the ocean where Chl-a is around several mg/m~3. On the other hand, eutrophic water in inner bays sometimes shows very high concentration more than 100 mg/m~3 with high SS and dissolved organic compound, where these algorithms cannot be applied. In this paper, an algorithm that enables us to estimate Chl-a from Landsat TM data is proposed based on field data obtained in Mikawa Bay for several years. Atmospheric effects that deteriorate the quality of information of Chl-a are considered in the algorithm.