engleski

Uncovering spatial and temporal patterns of the Adriatic Sea color with self-organizing maps

Fifteen years (1998-2002) of monthly chlorophyll- like pigments concentrations from the satellite images (JRC database) were analyzed, using self- organizing maps (SOM). The idea was to classify the pigments structures of the Adriatic Sea into a number of patterns present in the data by this novel neural network method. The images are represented by a vector in the multidimensional data-space, limited to the Adriatic area. We tried with several numbers of predefined patterns, selecting at the end optimal number of sixteen spatial structures. Distinct neurons are primarily associated with patterns of different pigments concentration in the whole Adriatic area and with a different extension of the color front along the west coast. Rare phenomena and variability limited to smaller areas were less represented in the SOM patterns. Although temporal pigments dynamics is in many areas of the Adriatic Sea synchronous the analysis of annual course resulted in ten temporal patterns. This representation has also demonstrated the ability of SOM method to connect areas with a particular seasonal variation of chlorophyll.

Adriatic Sea color; self-organizing maps; satellite images; chlorophyll time series

nije evidentirano