Pendugaan Distribusi Vertikal Klorofil-a Berdasarkan Neural Network

Abstract

The primary production quantity depends on the vertical distribution of chlorophyll concentration in the water column. The chlorophyll maximum value not always observed near or at the sea surface, but sometimes lies deeper than bottom of the euphotic zone. In this case, the ocean color sensors cannot measure the chlorophyll maximum value. A shifted Gauss model has been proposed to describe the variation of the chlorophyll-a (Chl-a) profile which consists of four parameters, i.e. background biomass (B0), maximum depth of Chl-a (Zm), total biomass in the peak (h), and measurenment of the thickness or vertical scale of the peak (). However, these parameters are not easy to be determined directly from satellite data. Therefore, in these research, an ANN methodology is used. Using in-situ data 1962 to 1985 in Banda Sea, the above parameters are calculated to derive the Chl-a concentration, sea surface temperature, mixed layer depth, latitude, longitude, and season. The total of 79 profiles of Chl-a and temperature are used for ANN. The correlation coefficient of these parameters are 0.912 (B0), 0.871 (h), 0.986 () and 0.990 (Zm) respectively. After comparing with in-situ data and ANN model, the result show not good enough agreement relatively. Keywords: Chlorophyll-a (Chl-a), Vertical Structure, Artificial Neural Networks (ANN)