Linkage between multi-model uncertainties and the role of ocean heat content in ocean carbon uptake

Ocean heat content (OHC) plays an important role in ocean carbon uptake (OCU). However, the changes of OHC and OCU are model-dependent and have large bias compared with observations. This makes it difficult to quantify their relationship. Here, we propose a new metric to measure the uncertainty of the relationship between OHC and OCU. The new metric can link the uncertainty with different OCU processes and allow direct comparison of the impact of OHC on the OCU in different simulations. The metric is illustrated in different simulations of the Coupled Model Intercomparison Project phase 5 (CMIP5) in which atmospheric CO₂ is increased by 1%/year. Results show that OHC in 0–500 m plays a dominant role in the OCU for the radiatively coupled (RAD) experiment because warming intensifies the carbon loss in the upper ocean. Relatively, OHC in the intermediate waters (500–2000 m) are crucial for the fully coupled and biogeochemically coupled experiment because this layer largely regulates the OCU. For different ocean basins, the intermediate Southern Ocean and deep North Atlantic are more important for the OCU in the RAD simulation. The metric also suggests the importance of global overturning circulation and the Southern Ocean in the OCU.