Low, T., Maroon, E., and Yeager, S.G. (2024)
Presented at:
Ocean Sciences Meeting 2024Global climate models, such as the Community Earth System Model version 2 (CESM2), have long struggled with their representation of North Atlantic currents. The misrepresentation of North Atlantic currents cause biases in sea surface temperature (SST), salinity, and density that degrade climate further downstream. These surface biases also affect buoyancy-forced circulations, including the Atlantic meridional overturning circulation. Surface-forced water mass transformation (WMT) can be used to understand how surface biases affect the overturning circulation. Here, we introduce a new observation-based WMT benchmark dataset and then compare it to 1-degree and 0.1-degree CESM simulations. the coordinated ice-ocean reference experiments version 2 (COREv2) and the Japanese Reanalysis surface dataset for driving ocean-sea ice models (JRA55-do), paired with multiple different sea surface temperature datasets including HadISST, OISST, EN4, and ECCO. We find that observation-based WMT from JRA55-do is weaker than that from COREv2. Using different SST and surface salinity datasets impact WMT strength and shifts the max in WMT to heavier or lighter density classes. When comparing observation-based and CESM WMT, we find that only the WMT in the 0.1-degree CESM is consistent with the observation-based benchmarks in the subpolar North Atlantic. Decomposing WMT shows that these differences between 1-degree CESM and its 0.1-degree counterpart are mainly due to isopycnal outcrop placement and area, as well as differences in surface heat fluxes.
