Sensitivity of the sea level change to model resolution in Max Planck Institute Earth System Model (MPI-ESM) simulations

Abstract

Rising sea level driven by global warming is a serious threat to the coastal communities whose livelihoods depend on the coastal ecosystem. Hence, effective adaptation measures are needed to mitigate the predicted dangers of sea level rise. To be able to make informed decisions and develop successful adaptation strategies, accurate and reliable sea-level projections are essential for the coming decades. However, current climate models do not have the resolution to explicitly represent mesoscale processes in the ocean component; as a result, these processes are parameterized. The limited description of mesoscale processes in these models results in systematic errors in simulating ocean circulation properties, affecting the accuracy of sea-level projections. To study the dependence of future sea-level change on ocean model resolution, the dynamic sea level (DSL; sea surface height above the geoid) change patterns in a climate model featuring an eddy rich ocean component are compared to those of state-of-the-art coarser resolution versions of the same model. The study examines the impact of spatial resolution on sea level projections using Max Planck Institute Earth System Model (MPI-ESM) simulations. Three different spatial resolutions are taken into consideration: low resolution (LR) of about 1.5°, high resolution (HR) of 0.4°, and eddy-rich resolution (ER) of 0.1°. In the first part, the DSL changes for each configuration are analyzed by comparing the time mean of the SSP5-8.5 climate change scenario for the years 2080–2099 to the time mean of the historical simulation for the years 1995–2014. The ER model, which resolves mesoscale processes, projects a higher DSL increase in the North Atlantic sub-polar region, the Kuroshio region, and the Arctic Ocean compared with models with parameterized eddies (HR and LR). In addition, a smaller DSL increase is observed in the band at 40°S in the Southern Ocean, compared to HR and LR models. The differences between the two model categories, in these regions, can be attributed to alterations in regional circulation patterns. This study also revealed that large-scale DSL change patterns, such as dipoles in the North Atlantic and North Pacific and belt-like DSL change patterns in the Southern Ocean, are not resolution dependent. However, the finer details differ in the coarse resolution configurations from those in ER by either the magnitude or the location. Therefore, non-eddy-resolving studies can still be used to understand large-scale changes. Yet, eddy resolving studies are necessary for smaller-scale sea level change research, because future adaptations and mitigation measures will primarily be smaller-scale in nature. The second part of the study uses prescribed identical surface forcing anomalies of heat, momentum, and freshwater fluxes to understand the mechanisms of the different responses of the ocean in various resolutions. The heat-flux perturbation causes a major fraction of North Atlantic sea-level change and when the ocean component is better resolved, the negative DSL change in the sub-polar region (southeast of Greenland) indicates a decrease in both magnitude and size. The sea-level decline is associated with an increased heat transport divergence that is caused by larger (smaller) northward heat transport at the northern (southern) boundary of the sub-polar region. Between the simulations, no significant differences in heat-flux-induced weakening of the Atlantic Meridional Overturning Circulation (AMOC) and the associated heat transport changes at the southern boundary exist but ER shows significantly lower changes at the northern boundary. These lower changes also yield an enhanced decrease in thermosteric sea level in the Eurasian basin of the Arctic and a smaller increase in thermosteric sea level in the GIN Seas in ER. In parameterized simulations, surface freshwater perturbation results in an increase in thermosteric sea level south of 60ºS. This is attributed to the reduction in surface salinity, which increases the stability of the water column and inhibits deep convection. Consequently, heat transported by the circumpolar deep water accumulates at greater depths, resulting in a substantial thermal expansion of the water column. By contrast, the ER simulation does not show a significant thermosteric increase because the water column was already more strongly stratified in the control experiment state, yielding lower sensitivity. In conclusion, this study sheds light on the complex nature of regional sea level change and emphasizes the necessity of high-resolution modeling to capture the dynamics of how eddy fluxes of heat and salt respond to climate change. Coarse resolution models with parameterized meso-scale eddy processes have limitations in accurately representing these dynamics. Eddy-resolving models have the capability to provide more accurate and reliable sea-level projections, which are essential for developing effective adaptation strategies to future sea-level rise.