Neuronal Network Models of Phytoplankton Primary Production

Abstract

Empirical models of phytoplankton primary productivity have always played an important role in oceanographic research, mainly because direct measurements of this process are difficult, expensive and time-consuming. Moreover, these models are needed to estimate primary production from the large phytoplankton biomass data sets that are obtained by remote sensing. They are also necessary to carry out instrumental estimates of primary production (e.g. by pump and probe fluorometers) and, in general, to post-process phytoplankton biomass data. Many different empirical models have been developed during the last 40 years and several of them have provided very useful results. The most common formulations among these models are based on simple linear relationships, where depth-integrated phytoplankton primary production depends on phytoplankton biomass within the upper layer of the water column (i.e. the upper attenuation length). For instance: