Elisabetta Canuti ^1* Antonella Penna ²

• ¹ Joint Research Centre (Italy), Ispra, Italy
• ² University of Urbino Carlo Bo, Urbino, Marche, Italy

The study investigated the distribution of surface phytoplankton communities in the Baltic Sea using datasets from different seasons and areas. Data collected during six oceanographic campaigns conducted between 2005 and 2008, included high-performance liquid chromatography (HPLC) pigment characterization, measurements of apparent and inherent optical properties, and hydrogeological parameters. The first part of this comprehensive study is focused on the HPLC phytoplankton pigments dataset in relation to hydrogeological conditions. The research highlighted the importance of high quality input data for accurate taxonomic analysis. The results evidenced a seasonal pattern of phytoplankton succession in the Baltic Sea, with diatom blooms in spring, cyanobacterial blooms in mid-summer, and haptophyte and dinoflagellate peaks in late summer and autumn. Several unsupervised machine learning approaches, including Hierarchical Cluster Analysis (HCA), Principal Component Analysis (PCA) and Network-Based Community Detection Analysis (NCA),Principal Component Analysis (PCA) and network analysis, were used to analyze the data and identify phytoplankton communities. Network analysis revealed significant connectivity among communities, which helped to understand the phytoplankton spatio-temporal distribution. Five distinct phytoplankton communities were identified based on biomarker pigments and PCA analysis, including diatoms, dinoflagellates, cryptophytes, green algae, and cyanobacteria. Where the results from PCA and Network analysisNCA were in good agreement, the outcome from the HCA analysis was less helpful in elucidating the phytoplankton structure. The results of the statistical analysis were then compared with traditional approaches such as CHEMTAX and regionspecific bio-optical algorithms, providing new perspectives on the taxonomic composition of phytoplankton groups, functional types (PFTs) and size classes (PSCs). Overall, the study provided valuable insights into phytoplankton dynamics and the effectiveness of different analytical approaches in understanding community structure, providing metrics that can enhance current and future advancements in remote sensing, including support for hyperspectral ocean color remote sensors such as NASA's Plankton, Aerosol, Cloud, and Ocean Ecosystem (PACE) mission.