Temporal-spatial change of phytoplankton size structure from satellite observations in the Bohai and Yellow Seas of China during 2003–2024

Phytoplankton size structure is a key ecological indicator for assessing marine ecosystems status and their response to environmental change; however, long-term monitoring of phytoplankton size has been constrained by limitations in observational data. To address this gap, this study developed a 22-year (2003–2024) daily satellite-derived phytoplankton size class (PSC) dataset for the Bohai and Yellow seas of China (BYS) using an in-situ optimized abundance-based model. Key findings are: (1) micro-phytoplankton dominated the coastal waters all-year round, whereas offshore water exhibited seasonal alternation between pico-phytoplankton (summer) and micro-phytoplankton (other seasons); (2) a significant long-term phytoplankton size miniaturization was detected, with the contributions of micro- and nano-phytoplankton declined at a rates of −0.031% per year (non-significant, P>0.05) and −0.049% per year (significant, P<0.01), respectively, while pico-phytoplankton increased by 0.055% per year (significant, P<0.01); (3) through explainable machine learning model, sea surface temperature was identified as the dominant driver of PSC variations, exhibiting highly nonlinear relationships and distinct threshold effects with shifts in phytoplankton size structure. This validated, long-term satellite PSC product provides novel insights into the environmental forcing of phytoplankton community restructuring in the BYS and offering a critical baseline for assessing ecosystem responses to climate change and anthropogenic pressure.