How ocean physics affects the stability of climate models

What is it about?

Within the Atlantic, warm salty waters are carried into northern latitudes where they give up their heat, and freshened by precipitation. When the former process dominates the latter, the waters become denser and sink, allowing more warm waters to be drawn into the North Atlantic. The shutdown of this circulation in the past has been associated with rapid climate change over the North Atlantic, leading to concern about whether global warming might result in a similar shutdown in the future. Many of the coupled climate models used to project the impacts of global warming forward into the next century fail to reproduce the current distributions of temperature and salinity. The first generation of climate models corrected for such biases by adding "corrections" to the fluxes of heat and moisture between the atmosphere and the ocean and recent work suggested that this may result in models that are systematically less stable to global warming. This paper evaluates whether this argument still holds when the source of the biases is not inaccuracies in the simulation of the atmosphere, but in the representation of ocean physics. Because climate models are very complex, we perform our studies in a simple box model, where we can carefully control what aspects of the model are being varied. We focus on a number of parameters which vary across coupled climate models. We find that correcting the fluxes does help produce a more realistic estimate of how far the overturning is from collapse if the models are biased in terms of Southern Ocean winds or eddy activity, but not for other physical biases.

Featured Image

Why is it important?

This paper helps to explain why we might (or might not) trust the models used for projecting climate change, and helps motivate efforts to better constrain key parameters such as lateral mixing rates from observations.