New research highlights the importance of seasonal carbon cycle variations in ocean carbon sink estimates

The global ocean has long been recognized as a key regulator of Earth’s changing climate, absorbing about 25% of the human-produced carbon dioxide gas (CO₂) released to the atmosphere each year. To monitor this oceanic “sink” for human-produced CO₂, international efforts such as the Global Carbon Project rely on observation-informed data products and numerical ocean models to estimate fluctuations in ocean CO₂ uptake from year to year. In the past decade, there has been a growing disagreement in the estimates from these two approaches: many global models predict a weaker modern ocean CO₂ sink than the observation-informed estimates suggest, though the causes of this difference remain uncertain. Researchers have been working for many years to understand why the observation-based estimates and models disagree from one another, and why model estimates differ from each other.

A recent study published in Global Biogeochemical Cycles led by Dr. Mar Arroyo, who conducted this work while a graduate student at the University of California Santa Cruz and is now a Postdoctoral Scholar at UW CICOES / NOAA PMEL, examined how different representations of seasonal carbon cycle variations across a number of ocean models used by the Global Carbon Project may impact annual ocean carbon sink estimates. The study finds that ocean CO₂ sink disagreements between models can be directly linked to differences in how they simulate seasonal CO₂ variations in the surface ocean. Within a given year, surface ocean CO₂ levels naturally cycle from high to low as the seasons change. It is well established that, as the ocean accumulates more human-released CO₂, the range of these seasonal CO₂ variations increases: seasonal CO₂ highs get higher and seasonal CO₂ lows get lower. A key finding of this article is that a model’s initial seasonal CO₂ variability dictates its future response to the accumulation of human-released CO₂: models with a larger seasonal cycle range experience a more pronounced expansion in that range over time. Because of this, instead of models converging toward agreement in surface ocean CO₂ cycling over time, the differences between them actually grow.

The differing rates of seasonal CO₂ cycle expansion in models have important implications for ocean CO₂ sink estimates. Surface ocean CO₂ levels, especially in the winter months, are a primary factor in determining how much human-released CO₂ the ocean can absorb from the atmosphere. When models disagree on seasonal CO₂ levels, it creates inconsistencies in the air-to-sea CO₂ difference, directly leading to widening gaps in ocean CO₂ uptake over time (see figure). This indicates that differences in CO₂ seasonal cycle representation directly impact estimates of the annual CO₂ sink within models.

Ultimately, creating more realistic representations of how surface CO₂ levels vary seasonally within these models is essential for providing a more reliable estimate of the changing ocean CO₂ sink. By highlighting the influence of differing seasonal CO₂ cycling in models, this study offers a recommended path forward for reducing uncertainties in model-based ocean CO₂ sink estimates, directly supporting NOAA’s mission to understand and predict the Earth system. By refining historical estimates of the ocean CO₂ sink, researchers gain a clearer picture of how the ocean has responded to human CO₂ emissions so far. This historical accuracy is the foundation for more precise predictions of how the ocean will continue to serve as a climate regulator under different emissions scenarios. Collaborators on this work include Dr. Andrea J. Fassbender of NOAA PMEL and Dr. Keith B. Rodgers from the World Premier International Research Center Initiative - Advanced Institute for Marine and Ecosystem Change in Sendai, Japan.