Improving the prediction of ecosystem indicators by assimilating biogeochemical data from multiplatform observing systems

Operational forecasting systems rely on data assimilation to improve the predictions of ocean biogeochemistry. However, the integration of multiple observation platforms and the provision of robust uncertainty estimates are still a challenge for current operational systems. We apply ensemble-based data assimilation approaches to integrate multiplatform observations of chlorophyll from remote sensing and insitu measurements. The assimilation skill is quantified by combining controllability metrics for the assimilated variable with metrics measuring the reduction of uncertainty in key ecosystem indicators. The addressed target indicators (phytoplankton productivity and phenology, and carbon cycle) provide insight into marine ecosystem functioning and ocean health. Sensitivity to the assimilation settings is assessed through a series of experiments conducted in a 1D numerical framework, in which different configurations combining the two observational sources are explored. In addition, ensemble multi-platform assimilation is implemented in 3D domains of three Copernicus Marine operational systems, spanning environments from the oligotrophic Mediterranean Sea to the eutrophic North-West Shelf and the Arctic seas. The results consistently indicate a reduction in uncertainties across key ecosystem indicators, including subsurface processes and summer conditions. Notably, the combined assimilation of ocean colour and in situ profile observations substantially decreases uncertainty in primary production dynamics across all systems. However, the transferability of results from 1D to 3D configurations should be approached with caution, as differences in biogeochemical model behaviour significantly affect the outcomes. Overall, these findings support the effective integration of satellite and in situ observations within current operational systems, enhancing the prediction of indicators relevant to ocean health monitoring and ecosystem management