Classification of Water Masses and Sound-Scattering Layer Biomass in the Waters off Northeastern Taiwan Using a Fuzzy Clustering Method

Abstract

One of the most productive neritic fishing grounds near Taiwan lies to the northeast of the island on the southeastern margin of the East China Sea continental shelf. In this region, however, the relationships between biomass distribution and oceanographic conditions are complex and thus difficult to evaluate using traditional sampling methods. The present study, therefore, applied a fuzzy clustering method to hydroacoustic survey data (Sv, volume backscattering strength indices of biomass) and CTD (conductivity, temperature and depth) observation data. Four types (A, B, C and D) of water masses were suggested by results of fuzzy clustering analysis. In type A water mass (continental mixed water), which had high temperature (> 26.54 ºC) and low salinity (< 33.942 psu), density (< 22.062 kg/m3) and chlorophyll a fluorescence (< 0.183 mg/m3 ), the biomass indices were higher than –59.37dB. In type D water mass (upwelling of the Kuroshio subsurface water), which had low temperature (< 23.41 ºC) and high salinity (> 34.272 psu), density (> 23.245 kg/m3) and chlorophyll a fluorescence (> 0.251 mg/m3), the biomass indices were less than –65.24 dB, which was about 4 times lower than the biomass value for the type in A water mass. The biomass indices in water mass types B and C were –62.20 ~ –59.37 dB and –65.24 ~ –62.20 dB, respectively. We conclude that fuzzy clustering is a useful method for classifying the water masses in the waters off northeastern Taiwan, and that each type of water mass is associated with a particular range of values of the biomass (Sv) index. Accordingly, application of the hydroacoustic technique and fuzzy clustering analysis to obtain the information about the biomass distribution and water mass types is potentially feasible