Plain Language Summary: We detail a new climate model hierarchy, CM4X. CM4X has two model configurations, CM4X-p25 and CM4X-p125, that differ only in the ocean/sea ice horizontal grid spacing. CM4X-p125 outperforms CM4X-p25 for certain climate processes, while maintaining skill levels seen in previous generations for other results. CM4X-p125 requires about 10 times less time than CM4X-p25 to reach pre-industrial control thermal equilibration. Also, CM4X-p125 equilibrates to an ocean state with roughly 400 ZJ less heat content than present-day, consistent with estimates of anthropogenic heat uptake since 1870, whereas CM4X-p25 equilibrates to a state with roughly 1100 ZJ more heat than present-day. Consequently, the CM4X-p125 ocean state has not drifted far from observational estimates. We propose the mesoscale dominance hypothesis to interpret the relatively rapid thermal equilibration of CM4X-p125 to a cooler and more realistic pre-industrial state. Such ocean models result from negligible spurious mixing (from numerical truncation errors) along with an active mesoscale transport and realistic parameterization of small scale (diapycnal) mixing. Noting the preliminary nature of our results, and with caveats detailed in this paper, we suggest that the more rapid thermal equilibration possible from mesoscale dominant ocean models greatly reduces the computational energy footprint of models that are not mesoscale dominant.
Plain Language Summary: We examine simulations from a new climate model hierarchy, referred to as Climate Model version 4X (CM4X). The finer grid component of the hierarchy, CM4X-p125, out shines its coarser sibling, CM4X-p25, for certain processes of interest for climate studies, though in others the results are not dramatically distinct. Each case study reveals the advances made by moving from the predecessor CM4.0 climate model to finer grid spacing in either the atmosphere or ocean. Even so, there remain many unresolved problems that help to guide further research and development goals and strategies.
The ventilation of the central Labrador Sea is important for the uptake of ocean tracers and carbon. Using historical ocean observations, we construct a simple multiple linear regression model that successfully reconstructs the decadal variability of the upper ∼2,000 m of the central Labrador Sea water properties based on observed indices that represent two different open-ocean ventilation mechanisms. The first mechanism is the modification of deep ocean properties through local decadal variability of the Labrador Sea deep convective mixing. The second, more novel, mechanism is the climatological convective vertical redistribution of upper central Labrador Sea temperature and salinity anomalies associated with the nonlocal large-scale subpolar Atlantic Multidecadal Variability and the Atlantic Meridional Overturning Circulation. The ventilated decadal central Labrador Sea signal subsequently spreads into the western subpolar North Atlantic. The results have important implications for predicting decadal ventilated signals in the Labrador Sea that are associated with the large-scale climate variability.
Zhang, Rong, and Matthew Thomas, June 2021: Horizontal circulation across density surfaces contributes substantially to the long-term mean northern Atlantic Meridional Overturning Circulation. Communications Earth and Environment, 2, 112, DOI:10.1038/s43247-021-00182-y. Abstract
The Greenland Sea is often viewed as the northern terminus of the Atlantic Meridional Overturning Circulation. It has also been proposed that the shutdown of open-ocean deep convection in the Labrador or Greenland Seas would substantially weaken the Atlantic Meridional Overturning Circulation. Here we analyze Robust Diagnostic Calculations conducted in a high-resolution global coupled climate model constrained by observed hydrographic climatology to provide a holistic picture of the long-term mean Atlantic Overturning Circulation at northern high latitudes. Our results suggest that the Arctic Ocean, not the Greenland Sea, is the northern terminus of the mean Atlantic Overturning Circulation; open-ocean deep convection, in either the Labrador or Greenland Seas, contributes minimally to the mean Atlantic Overturning Circulation, hence it would not necessarily be substantially weakened by a shutdown of open-ocean deep convection; horizontal circulation across sloping isopycnals contributes substantially (more than 40%) to the maximum mean northeastern subpolar Atlantic Overturning Circulation.
