All of the man-made objects and anthropogenic activities related to the oil industries in water bodies (e.g. drilling, equipment installation, construction of oil platforms and especially oil spillages) have environmental impacts on water ecosystems. The assessment of oil field operations and oil platforms impacts on water quality have become a necessity throughout the world. Phytoplanktons have the lowest trophic level in water environment; therefore the life of all of the other marine animals is highly relevant to phytoplanktonic population. Hence it seems necessary to monitor the spatiotemporal effects of different activities and events, related to oil industries on phytoplanktons. Chlorophyll is an indicator of phytoplanktonic biomass. Area-wide monitoring of Chlorophyll using conventional in-situ measurement methods is expensive and time consuming. However remote sensing measurements can provide a complementary tool for chlorophyll monitoring, but the remotely sensed data inherently contain high level of noise and some of the methods are failed in operational application to satellite data for the chlorophyll retrieval and also it is not easy to retrieve chlorophyll of noisy remotely sensed data in optically complex waters. Therefore robust and accurate inverse modeling methods for the chlorophyll retrieval of satellite data are needed. The aim of this study is to develop an innovative robust and accurate inverse model for the chlorophyll retrieval of Caspian Sea from MERIS (MEdium Resolution Imaging Spectrometer) data. We developed a new hybrid inverse model using the integration of ALM (Active Learning Method) and ANN (Artificial Neural Network) inverse models. The results of this study showed that it is very robust to noise and also demonstrated that it is very useful model for chlorophyll retrieval of remotely sensed data in the coastal (high chlorophyll concentration) and open sea (low chlorophyll concentration) waters. The main outcome of this research is an automatic code which is able to process satellite images automatically and enables robust water quality monitoring of different water bodies.

You can access this article if you purchase or spend a download.