Analytical Workflow for Tracking Aquatic Biomass Responses to Sea Surface Temperature Changes

Ocean ecosystem services provisioning is driven by phytoplankton, which form the base of the ocean food chain in aquatic ecosystems and play a critical role as the Earth‘s carbon sink. Phytoplankton is highly sensitive to temperature, making it vulnerable to the effects of temperature variations. The aim of this research was to develop and test a workflow analysis to monitor the impact of sea surface temperature (SST) on phytoplankton biomass and primary production by combining field and remote sensing data of Chl-a and net primary production (NPP) (as proxies of phytoplankton biomass). The tropical zone was used as a case study to test the procedure. Firstly, machine learning algorithms were applied to the field data of SST, Chl-a and NPP, showing that the Random Forest was the most effective in capturing the dataset’s patterns. Secondly, the Random Forest algorithm was applied to MODIS SST images to build Chl-a and NPP time series. The time series analysis showed a significant increase in SST which corresponded to a significant negative trend in Chl-a concentrations and NPP variation. The recurrence plot of the time series revealed significant disruptions in Chl-a and NPP evolutions, potentially linked to El Niño–Southern Oscillation (ENSO) events. Therefore, the analysis can help to highlight the effects of temperature variation on Chl-a and NPP, such as the long-term evolution of the trend and short perturbation events. The methodology, starting from local studies, can support broader spatial–temporal-scale studies and provide insights into future scenarios.