New ice is full of salt

When sea ice freezes, its salt separates out, and heavy salty water sinks to the bottom. However, it turns out that this process take place more slowly than we thought, which may have implications for our climate models.

Lars Henrik Smedsrud of the Bjerknes Centre at the University of Bergen and Ragnheid Skogseth of the University Course on Svalbard have been studying what happens when new ice and brash ice form at sea – a subject which they have been remarkably alone in studying recently. For while many scientists have been interested in what happens to ice that has been around for some time, few of them have chosen to study what happens right at the beginning, when the ice is still in the process of freezing:

“We have cited only one article, by Martin and Kaufman from 1981, which looked at much the same topic, but to the best of my knowledge very little has been done since their study,” says Smedsrud.

Studying fresh slush

Together with Skogseth and others, for many years in succession he has spent several cold weeks at the edge of major leads, or channels, in the pack-ice around Svalbard. Large, open channels are also known as “polynya”, a word derived from the Russian word for just this feature of the pack-ice.

“Such channels often develop when pack-ice is blown away from the land,” explains Smedsrud.

In polynya, new ice forms throughout the winter, which makes them ideal for studies of the properties of completely fresh sea-ice, particularly because the brash ice, the rather slushy mixture of ice and water that characterises the first phase of ice formation in turbulent water, forms more rapidly than ice that freezes solid.

Smedsrud and his colleagues created an artificial eight by ten-metre lead in Van Mijenfjorden near Svea on Svalbard, but they have also collected samples from a natural polynya in Storfjorden.

Driving force for ocean currents

There they measured the thickness of the ice layer and the types of ice it was made up of, and studied the salt content of both the water and the ice. The latter turned out to be unexpectedly high.

This came as a surprise to more people than just the research group itself. If you have driven on winter roads in Norway, you already know how salty water freezes at a lower temperature than pure freshwater.

“But all saltwater will freeze if the temperature falls far enough. Seawater, which has a salt content of about 34 parts per thousand (ppt), usually freezes at around -1.9o C,“ explains Smedsrud.

However, the salt is not incorporated into the ice crystals, and it is therefore gradually separated out through a number of different processes. Instead, it dissolves in the water beneath the ice, making it even more saline and thus more dense. This makes the water sink and flow close to the seabed, while less saline water flows into the same area at the surface. These mechanisms are believed to be an important driving force for many of the ocean currents in our waters.

“The densest water in the Arctic was formed in just this way,” says Smedsrud.

Brash ice holds onto its salt

However, the models we use to understand such phenomena are inadequate. For example, they assume that the salt dissolves out of the ice relatively rapidly – while the data of Smedsrud and Skogseth suggest that this process take place relatively slowly.

“The brash ice is simply much more saline that people thought. In some of our measurements its salt content is nearly as high as normal seawater. Sooner or later the salt content will sink, but how quickly this happens is of great significance for our models,” explains Smedsrud.

“The ice in Framstredet between Svalbard and Greenland, for example, is mostly more than a metre thick and may have a salt content of 5 – 6 ppt. Such ice is more than a year old – so the salinity of new ice is much higher”. And we need to remember that Smedsrud’s measurements were made in brash ice, which forms more rapidly than solid ice. This is why we assume that the release of salt takes place more rapidly in polynya, where there is brash ice, than in other conditions when seawater freezes solid; so while the process takes place slowly enough where Smedsrud and Skogseth made their measurements, there is every reason to believe that it happens even more slowly elsewhere.

Hazardous waters

Smedsrud and Skogseth have already published their findings in the April issue of the journal “Cold Regions Science and Technology”, but ProClim, the project of which it is a part, will continue until the end of this year. Smedsrud therefore just come back from yet another field trip to Svalbard, and he will soon be returning there on the Coastguard vessel KV Svalbard, which will break through the ice cover to the polynya in Storfjorden.

“This year the ice is so thin that there will be no problem getting there by sea, but until now we have been flown in by helicopter and had our base in a little cabin on Edgeøya,” explains Smedsrud. All the same, operating in this way is not completely free of hazards – the scientists are isolated from civilisation on the remote island of Edgeøya (to put it mildly), and if the motor on their small boat should fail and the wind take hold of them, they risk being blown all the way to Bjørnøya.

But field-work is important. Today’s current models are not fine-meshed enough, operating as they do with a resolution of about 50 kilometres; where
Smedsrud and Skogseth have been operating, Storfjorden is scarcely 50 km wide.

Revolution under way?

“This means that Storfjord is only at about “noise level” in the model calculations. When you realise that the ice crystals we are studying are only 2 cm in diameter, you can see that there is quite a large gap between the model and reality. We have no chance of estimating how many such crystals you might find in Storfjorden in 2050, but at some point or other we will have to determine how heavily the wind is blowing from the land, and how thick the brash ice is – such parameters will have to be baked into a model. For the moment, we are working on a model with a resolution of 2 km, but even in that model we assume that the salinity of the ice is constant,” says Smedsrud.

Both he and the other scientists in the ProClim project realise that they will not be able to revolutionise climate modelling on the basis of this project alone.

“It’s a step in the right direction, but we hope and believe that when we have a better understanding of these problems, it will be possible to insert small-scale processes into our large climate models and thus improve their resolution.”