The secrets of one of the world’s most advanced sea-ice models

Researchers from NERSC and colleagues in France, Germany, and Mali have improved how to simulate sea-ice movement in the Arctic. Their sea-ice model neXtSIM with a new and unique rheology is very good at producing accurate sea-ice forecasts and can possibly enhance climate predictions. 

Sea ice in the Arctic. Photo: Espen Storheim/Nansen CenterSea ice in the Arctic. Photo: Espen Storheim/Nansen Center

Are you wondering now what “rheology” means? Don’t worry, we’ll get to that. But first, let’s start with sea-ice modelling and what it is good for. Researchers use computer models to investigate how sea ice in nature behaves, both in the past, the present, and the future. To do that, they use mathematical formulas to recreate processes as close to reality as feasible. This way they can for example forecast sea-ice conditions such as thickness and extent for the upcoming days (short time scales) or predict how fast the ice will melt with a warming climate (long time scales). Since the Arctic is warming much faster than the rest of the planet, scientists are extremely interested in what goes on there, both right now, and in the future. 

Sea-ice conditions are important for human activities in the Arctic, such as shipping, fishing, tourism, and research. The position of the sea-ice edge also has important implications for life in the ocean – along the ice edge, animals find a lot of food that is otherwise scarce. And more reliable climate predictions for the coming years are necessary to develop efforts for effectively adapting to climate change. To be able to forecast how the sea ice will behave on short and longer timescales is therefore relevant on different levels.

The sea-ice model neXtSIM has been developed over the past years by researchers at NERSC in collaboration with researchers from France, to produce the best sea-ice forecasts and to improve climate predictions involving sea ice. It is functioning really well now, and its latest modifications are introduced in the recent article “A new brittle rheology and numerical framework for large-scale sea-ice models”. neXtSIM is already producing forecasts available to everyone with the new rheology 10 days ahead of time, benefitting people operating in the Arctic Ocean. But how do researchers generally get a model to simulate the ice movement correctly?

In the Arctic winds and ocean currents act on sea ice, and the ice reacts to these forces. The study of how materials deform and move in response to a force is called rheology, the word originates from Greek and translates to “study of flow”. In sea-ice models, the rheology is a description with mathematical formulas of how sea ice responds to winds and ocean currents.

Different sea-ice models use different formulas (rheologies) to define the sea-ice response to winds and ocean currents. Many simulate sea ice like something soft, but we know that it cracks in a brittle way when responding to force. All mathematical descriptions of sea-ice rheology have some kind of flaw, and the researchers at the Nansen Center and other institutions involved in developing the sea-ice model neXtSIM have been using the so-called Maxwell-Elasto-Brittle rheology in the past years. It performed well on short time scales, but they were not satisfied with its performance when predicting changes over multiple years. So, they investigated how to modify this specific rheology to make neXtSIM produce even better short- and long-term forecasts.  

Sea ice breaks in distinctive patterns that span the entire Arctic – known as Linear Kinematic Features (LKFs). This image shows the shear (difference in motion on either side of a crack) as observed by satellite (left), simulated by the new rheology (centre), and simulated by a more traditional rheology (right). Image from “A new brittle rheology and numerical framework for large-scale sea-ice models”, CC BY-NC-ND 4.0.Sea ice breaks in distinctive patterns that span the entire Arctic – known as Linear Kinematic Features (LKFs). This image shows the shear (difference in motion on either side of a crack) as observed by satellite (left), simulated by the new rheology (centre), and simulated by a more traditional rheology (right). Image from “A new brittle rheology and numerical framework for large-scale sea-ice models”, CC BY-NC-ND 4.0.

Interestingly, rocks and sea ice break in a very similar way, and researchers in France developed a sea-ice rheology based on rock mechanics. The sea-ice modelling team at NERSC implemented it in neXtSIM, and the team around Einar Ólason recently further improved that rheology. Extensive testing has gone well: neXtSIM-produced sea-ice forecasts and actual satellite images of the Arctic show the same features. The model is now able to simulate nature as closely as currently possible.

Ólason and his colleagues have recently published a paper on the new rheology, which they call the brittle Bingham-Maxwell rheology. This new rheology is the last step in the chain from rock mechanics models to a full-scale sea-ice model. It includes a way to include damage propagation: When a crack forms in sea ice, it freezes over and is pushed shut again, but this former line is more prone to breaking again, because the ice is weaker along it. Including formulas for this allows our researchers to represent deformation happening to the Arctic sea ice on small and large scales, from hundreds of kilometers down to kilometers, or even meters. This means that cracks in the sea ice can be simulated reliably, and the forecasts produced with neXtSIM are generally better than forecasts produced with other sea-ice models. Better simulation of the cracks gives a better representation of atmosphere-ocean-ice interaction in the model, which will hopefully lead to more reliable climate predictions. 

Researchers in the Nansen Legacy project work both on the ice in the Arctic and in offices across the country. They all have the same goal: to find out as much as possible about the changing Arctic. Credit: Nansen Legacy/Andreas WoldenResearchers in the Nansen Legacy project work both on the ice in the Arctic and in offices across the country. They all have the same goal: to find out as much as possible about the changing Arctic. Credit: Nansen Legacy/Andreas WoldenThe authors Einar Ólason, Guillaume Boutin, Anton Korosov, Pierre Rampal, and Timothy Williams from NERSC have been part of the Nansen Legacy project. This paper constitutes one of their most important contributions to the project. Einar Ólason on the recent article, Nansen Legacy, and neXtSIM: “We went into the Nansen Legacy project wanting to demonstrate how the rheology we were using in neXtSIM would give us insights into atmosphere-ocean-ice interactions – only to find out that the rheology itself wasn’t good enough. With this latest work, we’ve addressed that problem and we now understand the processes better and we have a powerful new tool at our disposal to address the questions we originally posed. This is emblematic of cutting-edge research, where you think you understand what is going on, only to be proven wrong by your next result. It’s this constant discovery that makes the work so fascinating!” 

 

Reference:

Ólason, E., Boutin, G., Korosov, A., Rampal, P., Williams, T., Kimmritz, M., Dansereau. V., Samaké, A. (2022). A new brittle rheology and numerical framework for large-scale sea-ice models. Journal of Advances in Modeling Earth Systems, 14, e2021MS002685. https://doi.org/10.1029/2021MS002685

Click on image to enlarge. Visiting MyOcean (https://myocean.marine.copernicus.eu/) gives you an even better resolution.Click on image to enlarge. Visiting MyOcean (https://myocean.marine.copernicus.eu/) gives you an even better resolution.

Add comment

Login to post comments