Application of empirical neural networks to chlorophyll-a estimation in coastal waters using remote optosensors

This paper presents chlorophyll-a estimation in coastal waters off the Gulf of Finland using remote optosensors. Concurrent remote optosensor data and in situ measurements of water quality were obtained in the study area. Significant correlations were observed between digital values and chlorophyll-a measurements. The results as a case study show that the estimated accuracy of chlorophyll-a retrieval using neural networks is higher than the accuracy of chlorophyll-a estimation using regression analyzes in the area. The study also shows one example why remote optosensors are critical to monitor water quality in coastal areas such as the Gulf of Finland.