Abstract/Summary

The transfer of carbon dioxide between the ocean and the atmosphere, and within the ocean interior, can be described by constituent “carbon pumps”. These carbon pumps are driven by biological and physical processes. The biological components can be separated into the “biological carbon pump”, which describes the cycling of carbon in the upper layers driven by photosynthesis in phytoplankton, and the “microbial carbon pump”, which describes the bacterial transformation of dissolved organic carbon into a slowly degradable form in the deep ocean. Understanding these processes requires both sophisticated marine ecosystem models and observations of the ocean carbon cycle. This thesis proposes that the simulation and understanding of the carbon pumps can be improved through data assimilation. Data assimilation is the process of incorporating observations (data) into a dynamic model to improve the accuracy of the simulations. This thesis makes use of ocean colour observations obtained by satellite imaging, assimilated into the marine ecosystem model ERSEM. The first objective of this study is to provide evidence that assimilating ocean colour data into a marine ecosystem model improves the simulation of carbon fluxes in the ocean, which is supported by results from identical twin experiments. The second objective is to improve the understanding of the biological and microbial carbon pumps and their variability across different marine locations. This was achieved by comparing the results of ocean colour data assimilation reanalyses at a nutrient rich coastal site and a nutrient-poor open-ocean site. A major finding of this study is that nutrient concentrations control the strength of the biologically driven carbon pumps, with the microbial carbon pump showing dominance in nutrient poor environments.