Today's topic is Sea Ice Growth and Melt. It sounds fairly straightforward. But like just about everything else that goes on in the Arctic, there's a lot more than meets the eye. Key points I want to get across. First, sea ice, what is it? It's any type of ice that forms from the freezing of seawater. Other key point, in the Arctic Ocean, there's this fairly fresh low density surface layer, which greatly aids in the formation of sea ice. Also, sea ice growth is a distillation process. In other words, the ice that forms is at a considerably lower salinity than the seawater that it started with. For example, if we have a surface salinity of 30-32 parts per 1,000, or what we call practical salinity units, the ice that forms is considerably less salty than that. An old ice, which we call multi-year ice, I'll get to that in a bit, is fresh enough that you can actually use it as drinking water, and we used to actually do that in the field. Also, ice deforms in the Arctic. There's deformation processes that result in ridges and keels and leads, and it's a way that sea ice can easily, very quickly thicken. Also, ice in the Arctic is getting younger and it is getting thinner as well. We've talked about the reduction in ice extent. How many square miles, how many square kilometers of ice there are. But the ice is getting thinner and it's also getting younger. Sea ice growth. Let's think of the basics. Well, if we want ice to form, then think of it just if it's a freshwater lake. What happens is you have to cool the water column from the surface. So autumn comes on in the Arctic. What happened, of course, is the sunsets or starts to set and the surface layers of the ocean will cool. Well, what will happen is that the density will increase and that cooled water will want to sink. But it will also then want to draw warmer waters up from below. Now, in a freshwater lake, you'd have to continue and cool down the entire column of water, until we got to something called the temperature of maximum density. You have to cool the entire column down. Well, that would take a while, of course. Now, the thing is in the Arctic Ocean, this fairly fresh surface layer means that we don't have to cool down that entire column of water. We only have to cool it down maybe 20 or 30 meters. That's because the fresh water on the top is really light water, I should say. Really light water, but it's relatively light water, relatively low density water, and it wants to inhibit mixing. Why it wants to inhibit mixing, is because the freshwater on the top, again it means it's less dense, and the density increases with depth, and that's because the salinity increases with depth. What we have in the Arctic Ocean, there's a strong pycnocline. That is a strong increase in density with depth, determined by the halocline, a strong increase in salinity with depth. Because of that pycnocline, you only have to cool down the top 20 or 30 meters of the water before the ice can form. I'm showing this through this little pattern here. There I have a figure that's showing density increasing from left to right and depth increasing downward. That surface layer I'm trying to showing here is cold but fairly fresh. A lot of that freshness turns out to be due to river runoff. River runoff into the Arctic Ocean, keeping it fairly fresh, result in a low density layer. Then density increases quickly as we go down, and that is associated with what we call a pycnocline. A rapid increase in salinity with depth, is also causing that rapid increase of density with depth. Then below that, we have this Atlantic layer which is warm but fairly salty. If that Atlantic layer would mix upwards, it would be hard for the sea ice to form, but it's kept at bay because we have that fairly fresh surface layer. Now, first step in ice formation typically is frazil ice, little needles or spicules of ice, maybe a few millimeters long. When very needle-like appears, very characteristic, but that's the first step in the ice formation. Second, as I'm showing in this photograph, is grease ice. It has a gray greasy appearance and you can see it on the left side of this photograph, there's grease ice forming there. On the right side of the photograph that's still open water. Grease ice is greasy appearance of ice as these frazil ice crystals start to agglomerate. Now, if the ocean is calm, we have something called, nilas ice that forms. Theses nice smooth sheets of ice, and you can see, this is a nice photograph of it and you can see these people, I'm not sure if they're tourists or what ship that is, looking at that beautiful nilas ice and taking photographs of it. If we have a rougher ocean, we instead tend to form pancake ice. Here's a photograph of pancake ice and you can see why it's called pancake ice because it looks like pancakes. A lot of these have upturn edges because these ice flows tend to bump together and shoves the ice up along the side to bit. It looks like a white [inaudible] pad, but it's known as pancake ice. Then the ice continues to thicken until we get to what we would call pack ice. That pack ice in areas can be rather thicker, three meters thick, even thicker if there's ridging or rafting. Now, thermodynamic ice growth, how does ice basically grow? Well, one way it grows is through thermodynamic processes. I'm trying to show with this little cartoon here, two things at work. There is an ocean heat flux to the bottom of the ice. But there's also an upward conductive heat flux through the ice to the surface and then that heat would be lost there through longwave radiation losses. Well, if that upward conductive flux through the ice exceeds that ocean heat flux to the bottom, we have what we would call an energy flux divergence. If that's happening, ice is going to form at the bottom of the ice, that is at the ice-ocean interface. The bigger the difference between that upward conductive flux through the ice and that ocean heat flux to the bottom, the stronger the ice growth, the more rapid the ice growth. Now, snow atop the ice cover would diminish the growth rate. That's because, snow is a rather poor conductor. So instead of conducting the heat through the ice to the surface, now you have to go through the snow cover layer, but that snow cover layer does not conduct heat very well at all. So snow atop the sea ice cover would tend to reduce the growth rate. But we can also have deformation. One process is rafting. If we have compressive forces so that two ice flows are coming together like that, what can happen is one of them will fail, and one will write atop the other, or raft atop the other. This is a way that the ice can get thicker in a different way than this thermodynamic process. It's just by rafting so we can get a lot thicker ice through that process. Another way is through ridging. We bring these ice flows together again like that, but both of them fail and so we have a ridge that develops on the top and we also have a corresponding keel underneath the water as well. So we have a ridge and a keel, but again, resulting from these compressive forces on the ice. Now, this is just a nice photograph of shore ridging and what happened here is that the ice was being pushed up against the shore and the ice grounded upon the shoreline, but there were forces still acting on it and the ice failed and so we got these ridges. Those ridges were all maybe three meters tall or something like that. This photograph is just a number of us out there taking a look at them because of course we're scientists and we are interested in all this stuff. A question, how thick can undeformed sea ice get? The answer is that it could get up to about three meters in thickness. Harder to get it much more than three meters? Because, when you get really thick, it's very hard to conduct the heat away through the ice. You just can't conduct enough heat away through the ice. It's just much harder for ice to grow. This makes sense. Actually, when ice is initially very thin, it grows quickly because we can conduct a lot of heat away, but as the ice gets thicker, it gets harder and harder to conduct the heat away just because it is thicker and we get up to about three meters. That's about the limit that we can get for the thickness of undeformed ice. Now, if we ridge it or raft it, of course we can get it considerably thicker. Now what about melt? Well, one way when we think about melt is just to reverse this whole process we talked about the thermodynamic ice growth. Well, if you could get a case now where the ocean heat flux to the bottom is greater than the conductive heat flux out through the ice, then you will get melt at the bottom. Of course, the bigger the difference in the conductive heat flux to the ice and that ocean heat flux to the bottom, the stronger the melt. But as we get further into the melt season, we have a situation where there's no conductive flux at all. Basically that ice is what we call isothermal, it's the same temperature throughout, namely at the freezing point. Then you can have the ocean heat flux herding it from the bottom causing melt, but you're also going to have surface melt because the sun rises as spring and summer comes along. There's energy fluxes to the surface, a notably solar radiation, and so you get surface melt. The ice thins or melt from both side, from the top and from the bottom. When we get later into the season, say August, the surface melt tends to go away and most of the remaining melt is actually from the bottom, from the ocean heat flux. From ocean heat that was picked up and stored through the summer. Now, quite typically in mid-summer anyhow both of those processes are at work, the top melt and the bottom melt, but later in the melt season its really the bottom melt that takes over. Ice age is an interesting concept. If we look at the Arctic Ocean and the ice in it, there's really two types of ice. Firstyear ice, that is ice that grows in a single growth season. Such ice starts to grow in September, and then we get to the melt season and May, and the Ice that grew during that period is firstyear ice. Now through the summer, a lot of that firstyear ice melts away, but some may survive. That which survives the summer melt season becomes multiyear ice. Now come the next year, some of that multiyear ice, called second-year ice as well, but that melts away, but some of it's survives, and so now it ages another year and so on. It used to be that you'd find ice in the Arctic Ocean that may have been over a decade old, 10 years old. Now, generally, not always, the multiyear ice is the thicker ice you find in the Arctic, and that's because every winter it has time to grow some more. That's not always the case because if we have ridging and drafting, fairly thin ice can also become rather thick ice as well. A question. True or false. There's more multiyear ice in the Antarctic than in the Arctic? The answer is false. In the Antarctic, there's lots and lots of ice growth through the autumn and the winter of that hemisphere, but almost all of it melts away in summer, really most of it melts away. In the Arctic, by contrast, we always have some ice left at the end of the melt season, that's first year ice that survived, or multiyear ice that is maybe two, three, four, even five years old. At some point when that multiyear ice is all gone, by definition, that we will be a seasonal ice regime entirely. In other words, you go out there and August and September, and there would be no ice to speak of. Much more similar to the case that we currently see in the Antarctic. Now, what about this ice age, or the age of the ice. We can actually track this using satellite data. The image on the left is showing the amount of ice of different age classes we had back in 1985. Just focus your attention on the red. The reds is ice that was at least four years old and was a lot. The image on the right is showing what we had in 2019, almost all of that really old ice is gone. The ice is getting younger, it's getting thinner as well because it's losing that older, thicker ice. This figure here is showing the change in the age classes through time. Starting on the left to the right, that's going from the mid 1980's into the very recent period. The key is just to look at the stuff on the red, that's the ice that's at least four years old and we see we're losing it very quickly. As I just said, once we lose all the multiyear ice, sea ice in the Arctic will be but a seasonal beecher. Thank you.
