The pattern determination of sea surface temperature distribution and chlorophyll a in the Southern Caspian Sea using SOM Model

Remote sensing has changed modern oceanography by proving synoptic periodic data which can be processed. Since the satellite data are usually too much and nonlinear, in most cases, it is difficult to distinguish the patterns from these images. In fact, SOM (Self-Organizing Maps) model is a type of ANN (Artificial Neural Network) that has the ability to distinguish the efficient patterns from the vast complex of satellite data. In this study, the sea surface temperature data and chlorophyll a related to a part of south Caspian Sea were investigated weekly by NOAA satellite for three years (۲۰۰۳–۲۰۰۵) and the annual and seasonal patterns were extracted (elicited) with their relative frequency using the SOM model. In all patterns the Caspian Sea coast has the highest chl-a and when you go away from the shore the rate decreases and when you approach to the middle parts the chl-a is of the least proportion on the sea surface.

Keywords:

Self-organizing maps , SST , Chlorophyll a , Caspian Sea