Mid- and High-Latitude Influences on Carbon Uptake and Ocean Acidification in GFDL CM4 and ESM4

John P Krasting, NOAA / Geophysical Fluid Dynamics Laboratory, Ocean and Cryosphere, Princeton, NJ, United States, Maurizia De Palma, Princeton University, Princeton, NJ, United States and John P Dunne, NOAA Geophys Fluid Dynamic, Princeton, United States
Abstract:
The global ocean is a critical sink of nearly a quarter of present-day anthropogenic carbon emissions, but this additional carbon contributes to ocean acidification. Mid-latitude thermocline and intermediate waters along with high latitude deep waters are the primary storage pathways for carbon and model representations of processes (i.e. thermocline strength, lateral and vertical mixing, AMOC, and Southern Ocean convection) govern the total sequestration of carbon. Furthermore, the ability of the ocean to buffer against acidification varies both regionally and temporally, resulting in different acidification patterns in individual basins. Understanding how ocean carbon uptake and acidification vary among models with different resolutions and biogeochemical schemes is important in interpreting future projections. Using two coupled climate models, CM4 and ESM4, developed at NOAA-GFDL in support of CMIP6, we highlight different spatial distributions of carbon uptake despite both models having similar total ocean carbon uptake. Comparing simulated distributions of CFC tracers with observations, we find both models under-ventilate the deep Southern Ocean - except in the case of centennial-scale polyna-like events that result in full column convection in the Ross and Weddell Seas. In a future warming scenario (SSP5-8.5), the combination of sea ice loss and carbon uptake make the Arctic basin particularly vulnerable to acidification. The projected volume of Arctic ocean waters with calcium carbonate undersaturated with respect to aragonite increases abruptly on a decadal timescale from a baseline of ~10% to nearly 100% during the mid-21st century. Using a machine learning algorithm (K-means clustering), we identified regions with similar ocean acidification responses that can guide further analysis and observations. These results illustrate the importance of both mid- and high-latitude ocean processes for simulations of carbon uptake and acidification.