论文部分内容阅读
To realize potential cost savings in coastal monitoring programs and provide timely advice for marine management, there is an urgent need for efficient evaluation tools based on easily measured variables for the rapid and timely assessment of estuarine and offshore eutrophication. In this study, using parallel factor analysis(PARAFAC), principal component analysis(PCA), and discriminant function analysis(DFA) with the trophic index(TRIX) for reference, we developed an approach for rapidly assessing the eutrophication status of coastal waters using easy-to-measure parameters, including chromophoric dissolved organic matter(CDOM), fluorescence excitation–emission matrices, CDOM UV-Vis absorbance, and other water-quality parameters(turbidity, chlorophyll a, and dissolved oxygen). First, we decomposed CDOM excitation-emission matrices(EEMs) by PARAFAC to identify three components. Then, we applied PCA to simplify the complexity of the relationships between the water-quality parameters. Finally, we used the PCA score values as independent variables in DFA to develop a eutrophication assessment model. The developed model yielded classification accuracy rates of 97.1%, 80.5%, 90.3%, and 89.1% for good, moderate, and poor water qualities, and for the overall data sets, respectively. Our results suggest that these easy-to-measure parameters could be used to develop a simple approach for rapid in-situ assessment and monitoring of the eutrophication of estuarine and offshore areas.
To realize potential cost savings in coastal monitoring programs and provide timely advice for marine management, there is an urgent need for efficient evaluation tools based on easily measured variables for the rapid and timely assessment of estuarine and offshore eutrophication. In this study, using parallel factor analysis (PARAFAC), principal component analysis (PCA), and discriminant function analysis (DFA) with the trophic index (TRIX) for reference, we developed an approach for rapidly assessing the eutrophication status of coastal waters using easy-to- including chromophoric dissolved organic matter (CDOM), fluorescence excitation-emission matrices, CDOM UV-Vis absorbance, and other water-quality parameters (turbidity, chlorophyll a, and dissolved oxygen) by PARAFAC to identify three components. Then, we applied PCA to simplify the complexity of the relationships between the water-quality parameters. F inally, we used the PCA score values as independent variables in DFA to develop an eutrophication assessment model. The developed model yielded classification accuracy rates of 97.1%, 80.5%, 90.3%, and 89.1% for good, moderate, and poor water qualities, and for the overall data sets, respectively. Our results suggest that these easy-to-measure parameters could be used to develop a simple approach for rapid in-situ assessment and monitoring of the eutrophication of estuarine and offshore areas.