On possibility of the monitoring of water masses' structure in natural reservoirs by fluorimetry with using neural network algorithms

In the paper it is proposed to use fluorescence of dissolved organic matter (DOM) as a water masses indicator. The DOM fluorescence spectra were registered and for two reservoirs of different types: the costal zone of the North Sea and the Rybinskoe reservoir in the middle of the European part of Russia. In the paper it is shown that using normalized intensity of DOM fluorescence and artificial neural networks (ANN) for analysing DOM fluorescence band shape can provide determining the water masses type in natural reservoirs.