近岸海域叶绿素a反演研究进展OA

Coastal waters are significantly affected by human activities and have complex optical environments,which leads to great uncertainty in chlorophyll-a(Chla)remote sensing retrieval.Based on the optical properties of Case-2 waters in coastal areas,this paper systematically sorts out the theoretical basis and research progress of chlorophyll-a remote sensing retrieval,focusing on reviewing the application characteristics and applicability differences of empirical models based on blue-green bands,semi-empirical/semi-analytical models,retrieval methods using red and red-edge bands,and machine learning methods in complex coastal waters.The analysis shows that the combined interference of suspended particulate matter and colored dissolved organic matter(CDOM)on spectral signals is the key factor affecting the accuracy of coastal chlorophyll-a retrieval.Different methods have their own advantages and limitations in stability,accuracy and interpretability.The main challenges faced by coastal chlorophyll-a remote sensing retrieval in optical complexity,atmospheric correction,spatial resolution and data support are further summarized,and the development directions of multi-source remote sensing fusion,regional modeling and the combination of physical constraints and machine learning are prospected.