Book Review Modeling

Small scale, big effect

Su, Zhan, et al. “Ocean submesoscales as a key component of the global heat budget.” Nature Communications 9.1 (2018): 775. doi:10.1038/s41467-018-02983-w

We hear a lot about resolution in our increasingly digital lives. Most consumers worry about it in terms of imaging ­– think 4K Ultra High Definition televisions or 12-megapixel camera sensors in iPhones. But the concept of resolution applies in all digital domains, including the climate models scientists use to make predictions about our planet.

For decades, climate scientists have used computational models to study the Earth system as a whole. They create digital approximations of the world by dividing it into a 3-dimensional grid and writing mathematical rules that govern the physical relationship between the parameters they are studying. When climate modelers discuss resolution, they are referring to the size of the boxes in the grid.

Most global climate models have resolutions of 10s of kilometers, setting a lower limit to what types of physical phenomena can be described. This is akin to taking a picture with a low-resolution camera – you might be able to see a person’s face, but cannot quite make out their eye color. In their new paper, Dr. Zhan Su and his colleagues at CalTech and the Jet Propulsion Laboratory argue that climate models have been missing the effects of submesoscale processes due to limited resolution.

Figure 1 – A satellite image of submesoscale structures in the Baltic Sea. Notice how the feature swirl together and the 20 km scale bar (Adapted from Su et al., 2018).

The term submesoscale describes physical features in the ocean that occur at spatial scales of 10s of meters to several kilometers. Scientist often describe eddies as submesoscale phenomena (fig. 1). These structures are often observable from satellites and vessels. Recent work by other groups has suggested that they account for about half of the variability in vertical velocity in the ocean. The implication is that relatively small, local features play an outsize role in how water moves between the ocean’s surface and interior.

The water circulated between the surface and inner ocean governs the magnitude and direction of air-sea heat exchange. The heat flux between the ocean and the atmosphere is a critical component of the Earth’s climate system. In fact, scientists estimate that the ocean has taken up a substantial amount of the heat associated with anthropogenic climate change. Su recognized that incorporating the effects of submesoscale features might dramatically change the output of climate models. The point is not purely academic – climate models are used to inform policy decisions on the international stage.

Figure 2 – Output from the high resolution model corresponding to March, 2012. The insets are details from the Northern and Southern hemispheres (adapted from Su et al., 2018).

Su and his team leveraged powerful computers at the NASA Advanced Supercomputing facility to run a global model with about 2 km resolution. Each cell in the model represented 1/48 of a degree of latitude. The researchers forced the model with atmospheric and tidal information. The team then had it output predictions for a 14-month period corresponding to September of 2011 to November of 2012 (fig. 2). A separate, coarser model was run with the same inputs for comparison.

Su’s high-resolution model demonstrated that submesoscale features enhanced the amount of vertical heat transport across the entire global ocean. The mixing caused by the small-scale features generally resulted in a warmer surface and cooler inner ocean. The effect was more substantial in the winter when storm activity caused the ocean to be more energetic. At most grid points, the higher resolution model exhibited greater upward heat flux than the coarser simulation.

The group argues that their results demonstrate that submesoscale dynamics exert an important influence on the global the ocean. The effect of these relatively small processes could also have an impact on atmospheric dynamics. Su points out that the only way to assess the interplay is to develop models that more tightly couple the ocean and atmosphere at these resolutions.

Fleshing out the details of the interaction between small processes in the ocean and the climate is critically important. If the submesoscale is as influential as Su and others suggest, then the models used to predict the future of our planet might be in need of a big adjustment.

Leave a Reply

Your email address will not be published.