PNNL

The Science

Ocean waves influence how heat moves through the upper ocean, which can affect the energy available to tropical cyclones. Yet many Earth system models do not explicitly represent wave-driven ocean mixing, leaving uncertainty about how this process influences storms over many years. Researchers compared North Atlantic simulations with and without ocean-wave coupling in the Energy Exascale Earth System Model. Adding waves produced weaker tropical cyclones and reduced storm activity, particularly in the region where many Atlantic hurricanes develop. Wave coupling also intensified existing biases in sea surface temperature and vertical wind shear, a difference in winds at different heights that can inhibit storm development. Surprisingly, the wave-coupled simulation showed less storm-induced surface cooling despite stronger mixing during storms.

The Impact

The findings showed that whether including a physical process may improve a model’s simulation of tropical cyclones depends on existing model biases. In this configuration, wave-driven ocean mixing intensified existing errors in the simulated background environment, producing conditions less favorable for hurricane development. The analysis also distinguished two competing effects of waves: stronger mixing during individual storms and changes to the ocean that develop over longer periods. The longer-term changes dominated the simulated cooling response, because waves produced deeper, more uniformly warm upper-ocean layers that reduced the upper-ocean thermal stratification before storms arrived. This distinction helps explain why model results can differ from expectations based only on storm-scale interactions. The research identified background-state biases as an important consideration when evaluating ocean-wave coupling in Earth system simulations.

Summary

Researchers investigated how explicitly representing ocean surface waves affected simulated North Atlantic tropical cyclones and their surrounding conditions in the Energy Exascale Earth System Model version 2.1. They compared two otherwise similar historical simulations spanning 1978–2014, analyzing 1979–2014 after an initial adjustment year. Both simulations used a regionally refined atmosphere with approximately 25-kilometer resolution over the North Atlantic. In the wave-coupled run, WAVEWATCH III exchanged information with the ocean model to represent wave-dependent Langmuir mixing. The team compared storm statistics with observed hurricane tracks and evaluated ocean temperatures and atmospheric conditions using ERA5 reanalysis.

Both configurations underestimated intense storms and hurricane activity in parts of the Atlantic, but the wave-coupled version performed worse. Its modeled storm-track density matched observed spatial patterns less closely than the baseline simulation (correlations of 0.60 versus 0.73). Stronger cold sea surface temperature biases were associated with a stronger subtropical high-pressure system and increased vertical wind shear, likely contributing to fewer simulated storms. Wave coupling also deepened the pre-storm mixed layer by about 11 meters over the tropical Atlantic and weakened temperature differences immediately below it. Although storm-driven mixing deepened the mixed layer more with waves (about 5.8 versus 4.1 meters), the resulting sea surface cooling was smaller on average due to the weakened temperature differences between the mixed layer and the layer below it. The authors concluded that the effect of ocean-wave coupling depends on the model’s background conditions and noted that their configuration represented wave impacts through ocean mixing only but not direct wave-atmosphere interactions, which can also play an important role through a modulation of surface drag and air–sea fluxes.

Contacts

DOE Program Manager Contact

Sally McFarlane
Earth and Environmental System Modeling
Sally.McFarlane@science.doe.gov

PNNL Contact

L. Ruby Leung
Pacific Northwest National Laboratory
Ruby.Leung@pnnl.gov 

Research Contact

Karthik Balaguru
Pacific Northwest National Laboratory
Karthik.Balaguru@pnnl.gov

Funding

This work was supported by the U.S. Department of Energy, Office of Critical Minerals and Energy Innovation, Integrated Energy System Office, and by the U.S. Department of Energy, Office of Science, Biological and Environmental Research program as part of the Energy Exascale Earth System Model project. Computational resources were provided by the National Energy Research Scientific Computing Center.