Predictions of future atmospheric CO2 levels, an essential objective of Earth system science, are influenced by global carbon and nutrient cycles, climate interactions, and feedbacks to the Earth system. Relevant processes operate at different spatial and temporal scales and vary across terrestrial, coastal, and marine ecosystems. Biogeochemical feedbacks may be altered by anthropogenic and natural disturbances, including tropospheric O3, changes in nutrient and hydrological cycles, eutrophication, acidification, land cover/land use change, prescribed and wildland fires, and potential climate intervention strategies. This session focuses on integrated understanding of terrestrial and marine feedback mechanisms that impact the predictability of all aspects of the Earth system, methods for evaluating and benchmarking process representations in Earth system models, approaches for constraining future climate projections (e.g., emergent constraints), and novel applications of artificial intelligence and machine learning to improve predictive understanding of global biogeochemical cycles.