DR. JAMES OVERLAND>> ...here at PMEL and those online. I thought I'd give my annual update on the Arctic, 

And the big takeaway this time is that we have quite a few observations that are outside the envelope of previous experience.

And that's a real science quandary.

Is it an extreme, existing set of events and we just by random have a new extreme or have things changed enough that it's a new Arctic.

And it's not easy to support or to work out which it is. For science and statistical certainty is on one end but if you want to make public actions relating to global warming and so on, you probably want to use the new information. 

So in looking for changes overall for the last 40 years, it's good to look at just more than one variable. So we worked on a...what's called a SWIPA assessment, with various variables. 

So here we have summer air temperature on the top.

The second one is permafrost ground temperatures in Alaska.

The third one is a satellite measure of greenness. So that's how much we've transitioned from low tundra plants to shrubs, ice loss, snow cover duration.

This is mostly April and May.

And now that have no snow compared to where they were before. And then changes in glaciers and especially Greenland.

So if you look at all of these we start seeing changes in the mid-1990s. And there was enough information by the mid-2000s to really say there was a collective change overall. Then they continued.

There's some detail here, like the tundra seemed to hit a plateau where the amount of snow cover was impacting that the maximum amount of shrub growth and so on. And Greenland tends to have a lot of individual events rather than the more smooth curve.

The biggest effect shown before is the shift from mostly old thick multi-year ice across the Arctic, to mostly first-year sea ice.

And satellites are able to tell the difference, as old thick ice doesn't have much salt in it, no salinity where the first-year sea ice still has some salt in it. And that signal can be picked up by satellite so the main thing here is to look at the shift from the old thick ice in red to the mostly first-year ice in blue. And then we have the timing as you go from left to right.

And so after a minimum amount of sea ice in summer in 2007, right here, we had a big chunk and not much of a recovery but not a continuing down...downward until this last year where we had another big drop. And so we've replaced 60% of the old thick ice with first-year ice.

This is from...PMEL had a NOAA aircraft three years ago.

This is 200 miles north of Alaska.

And 20 years ago, we would be walking around here, you would have 20-foot high pressure ridges and would all be broken up. 

Where now it's basically really the thin, smooth, first-year sea ice.

So this shows a huge difference in my career from 20 years ago, walking around in roughly the same place and now seeing the ice is totally different.

So this would basically verify what the satellite was showing in the last picture.

So what kind of changes do we have going on?

Certainly we're losing the extent of sea ice in summer and one of the biggest is when will the Arctic be ice-free during summer?

And on the right here is the downward trend of the ice area.

And you notice there was a huge drop in 2007. And at that time we thought, well, maybe there's a tipping point that we went to such a low extent and so you have a lot more open water. Open water you can absorb more heat from the sun and so there would be a real big positive feedback of the ice going away quickly.

Well, that didn't happen. After 2007, there was more ice in 2008.

Then we had another minimum amount of summer sea ice in 2012.

And it went back to more ice.

So the blue line here is the long-term downward trend and so when we have these extremes they actually recovered back to the long-term downward trend.

And it looks like with global warming, we have an increase in temperatures and the ice is pretty much locked into that change in temperature.

So as we warm up we have less ice but if we were able, it looks like fortunately, if we were able to stabilize global temperatures it would probably stabilize the amount of sea ice we have.

But there are other changes that are irreversible.

One is permafrost, which for ten thousand years we've had a water and soil mixed together and frozen which is the permafrost.

And if you melt the permafrost and the water runs away and you're left with the dirt, and so you can't put that back together.

So permafrost is an example of one-way changes.

The other thing is shown in this picture that some on the North Slope of Alaska, you see all this open water in the background--no ice--so more open water, you can have bigger waves.

So when the ice was there, it kept the wave height down. Now that the ice is gone you have bigger waves when a storm goes through and that means more erosion along the coastline. So that's another one where things are speeding up.

Looking at the "big picture" for the long-term future, based on climate models--along the bottom here we have 1900 to 2000, 2000 to 2100.

And the blue lines here are the expected temperatures for an aggressive but not totally unreasonable mitigation scenario. 

That's sort of the best we can hope for. 

And so if like globally if by 2050, we had that scenario where we would hold the global temperature to an increase in two degrees, that's pretty much what the international community is trying to do. 

But if you look at the Arctic at the same time, you're over four degree increase which is another difference in the Arctic compared to where we are now.

So the Arctic is going to have a huge change, no matter what, even if we start stabilizing.

And winter temperatures are even larger than the annual.

The red line is basically doing nothing to slow down CO2 increases.

And basically what we've already put in and what we'll put in for the next 20 years is pretty much known and already there. So out to 2040 or so, we're pretty much locked in to these changes, there's nothing we can do to slow things down. But if we started mitigation now, that's the difference between the red line and the blue line. So that the impact of mitigation is really more towards our grandchildren's than us in the next 20 years.

So looking at some of the other effects. I mentioned permafrost. Permafrost, if it melts, that affects existing buildings that fall apart and it also can release methane and water. And that's barely started now. The big increase...uh, or, decrease in permafrost area, in the next...a fairly large impact for the next 20 years.

So we haven't really seen much permafrost impact but we should see quite a bit in Alaska, Canada and in Russia in next 20 years.

The other big effect is loss of glacier ice.

As you know, the sea ice is floating on the ocean, so when the sea ice melts it doesn't affect sea level but its land glaciers, when they melt, contribute about 40% to the rise in sea level.

The other 40% is just the warming of the oceans, which PMEL works on as well.

The thing to note, we looked at two different scenarios. The high emissions in the red and the low emissions in the blue and purple. But the big thing was the difference between how much the glaciers contributed by 2030 and by 2080. 

And there's a huge increase from the glacier impact in the second half of the century rather than what's going on now. So there's not much of a sea level effect now or for the next 20 years but it'll be huge in the second half of the century.

And the reason is that there's a lag between warming temperatures and the rate that glaciers melt. 

So, and that's about 30 years or so. So they call it commitment that that we've already had more of a temperature increase than the equilibrium to our current temperature will take about 30 years to impact.

So if we stabilize global temperatures, say, in the next 30 years or so, the glaciers will not stabilize before 2100. So the sea-level effect is going to continue throughout the century even if we stabilize the rest of the climate. So that's the big wild card.

Greenland's a big effect and Antarctica is another big effect.

So most of that information is already known. 

But what I want to bring up is some big changes in the last three or four years.

Here you're looking down at the North Pole. 

On the left are winter temperatures from winter 2016. 

And the red area are temperatures that were six degrees above the normal temperature.

And that was just about double the previous record. So not only did we exceed the previous record, we exceeded it by a lot.

So this is the kind of surprises that we're seeing in the Arctic data. 

And part of that had to do with bringing warm air in from mid-latitudes and an approximate low-level jetstream is following these purple lines here.

And you bring warm air across Alaska into the Arctic in this situation. And you bring warm air from the Atlantic Ocean across Scandinavia and spiraling into the North Pole.

So it wasn't just background overall warming, part of the contribution was more connection with the mid-latitudes as well. So that was winter 2016.

And this is this previous winter. And again we had extreme...

[AUDIENCE LAUGHTER]

We had extreme temperatures over the Arctic that hadn't been seen before. 

In between 17 was well above normal but it wasn't the extreme. So here within three years we've had, like, double the previous record and again it related to bringing some air in from the mid-latitudes.

In addition after this...so the low...

you go back...

the purple here are low pressure but there's high pressure...

in between...

[AUDIENCE LAUGHTER]

MAN IN AUDIENCE>> Go faster, Jim.

[AUDIENCE LAUGHTER]

DR. OVERLAND>> ...in between, that this high pressure, uh...

when this is done in March actually rotates out of the Arctic.

Next slide.

So some of that high pressure air moves south.

And high pressure, the winds go clockwise around the center of the pressure. 

And they got extreme cold temperatures and snow during late winter in Europe. They call it the "Beast from the East".

Next slide.

So another multiple sequential change in the Arctic is the amount of ice there was in winter.

And that's been a minimum amount compared to the long-term record as shown on the left.

The last four years had the minimum amount of ice during wintertime. 

So on the left you've got the amount of ice it's increasing into winter and we go from December through April. And all those colored lines are how much ice you had for different years. And the last four are well below the climatology.

So to some extent these summer or winter temperatures and particularly now the sea ice, it's not just a one-year extreme. There were four years in a row. And this was really unusual because if you have thin ice, it grows faster than thick ice. 

So even if you had a minimum amount of sea ice during summer or a late freeze-up in fall, the theory was that the ice would grow really fast and but by springtime you would lose any year-to-year memory that the ice would just go back to its normal winter thickness. That's not the case for the last four years. We've got a new situation where it looks like we're carrying some memory on the warmer thinner ice from year-to-year.

And that's really different from before.

On the right side we have the delay in fall freeze-up overall.

The Alaska, I'm going to talk a bit more but it's not only Alaska, it's north of western Russia that we have a....the ice is not freezing up. You know, that link there's ability of at least a month or so compared to where it used to be.

Another thing is when you're looking at summer, some of the climate models say, well, the summer ice would go away in 2016 but if you look at the long-term trend, that's 2030 or 2040, but if you look in detail at a picture of the ice during summer, there's all these holes between the ice floes where the...or the solar heating can penetrate. So it looks like the ice is getting thinner and thinner every year. So rather than the ice going away by cutting back less and less overall extent, it looks like just one year we might hit the balance where the ice was thin enough that there's enough solar heat in the summer to melt it out. So this is why a lot of us think, you know, this could happen within the next 10 years or so.

So a big deal for PMEL and people in Alaska and those concerned are the big effects that happened this year in Alaska.

And in the fall, in December, the picture on the right, the red, is the ice cover. You can see that there's still a whole lot of open water north of Bering Strait. Normally by this time, the ice would be down into the Bering Sea.

And as you see here's the total amount of ice in the 1st of December and it's the lowest we've ever had. And it's 50% below the long-term average. So there's a whole lot more open water in the fall. You could sail a boat through Bering Strait well into the fall.

When we come to the time of maximum ice, normally in the Bering Sea, and the picture on the right, the ice age can be down near the Pribilof Islands. They're covering about half of the Bering Sea. And there was very little ice in the Bering Sea in February. And the plot on the left-side here is different dates from late January into April and all the dots show the maximum ice extent in all the previous years. And this year, we're only about a third of the ice than in most previous years.

And yes, this could be a big deal for the ecosystem because when we normally have an ice-covered Bering Sea, we've cooled the water down to the freezing point minus one and a half or so. Then in spring you melt that out, it stabilizes the water column. The top can warm up. But the cold temperature, called a "cold pool", remain near the bottom, and that's the water that the Euphausiid, it's a small shrimp that are really prized by the fishery, live. So there's probably going to be no real cold pool this year that would support the fisheries. So we'll see what happens with that.

FOCI and Phyllis has a mooring right about here. Normally they have 150 days of ice cover and this year they have no ice cover at all at that place. And FOCI from both building 3 and building 4, heading up to really document what's going on because this is such an outlier compared to what we had. 

Some of the loss of ice, it was a maximum in February then we lost some ice. This had a lot to do with some southwest winds. Normally the winds are from the northeast blowing the ice south. This year we had, was unusual with the southeast wind. 

So part of the loss was due to the winds, not just the overall effect. 

And looking at the first day when the Bering ice starts warming up.

Well, you can see that this last year it's already started warming up where normally, you don't start melting until in June. So here's another example of the sea ice in the Bering Sea being completely out of where we've been before.

So in summary, we've seen the global modeling warming everything up over the next century with about a four degree increase for the Arctic. But things could be faster based on what we're seeing now. And one of the pieces of physics is that we now have all these open water areas so when there was ice there, the surface temperature could be like minus six or minus ten degrees Celsius. If it's open water it's like zero or minus one so even though it's cold it's a lot warmer than if there had been ice there. So when we have a storm bringing warm air into the Arctic, it can reach a lot further than it used to. And that's why we've seen these examples of warming temperatures reaching the North Pole and so on. So this is a new piece of physics that the changing in the jetstream and the winds, there tends to be random from year-to-year, but now the thermodynamics of open water is helping. 

And that's going to be a long-term effect in the two regions. Also warmer air is less dense and increase can change the height of pressure surfaces and that can change the wavy Jetstream. And that was going on last December when we had all this ice free and warm temperatures over Alaska. It helped the wavy jetstream so Alaska was really warm and helped lock in the wave but then the wave descends down with northerly winds bringing cold temperatures to the East Coast.

And so this is all kind of controversial but last December looked like a really clear case where the Arctic doesn't cause the cold weather on the East Coast but can help reinforce an existing pattern. So these are two new things that we're seeing. And so the question is that it looks like we're moving faster towards a new Arctic in the last three or four years. It's not just a long-term trend. We may have a period of accelerating change because we have more interaction between the surface and the atmosphere and the mid-latitudes and the Artic related to being outside previous experience. 

And you know, how do we reconcile this with the model projections? This indicates that things are probably faster than the models were saying but we can't really quantify that. And so that makes the idea that we should take more precautionary work based on what we're seeing now, extrapolating what we're seeing now. So thank you very much!

[APPLAUSE]

Have any questions from the room?

MAN 1 IN AUDIENCE>> I have one, Jim. You had that glacier melt diagram showing the possible scenarios. There was the scale on those bars you had there, how much when Greenland melts, how much it will rise and sea temperature we could talk about. Or did you purposely not put that on there?

[AUDIENCE LAUGHTER]

DR. OVERLAND>> I think in the paper we had to scale...it's in gigatons so it's nothing we can relate to. If you convert it over to sea level, we're looking at half a meter to a meter by the end of the century. And so what, you know, what we're seeing now is in maybe 10 centimeters. So that's the difference between the near-future and the far-future is between 10 centimeters and a meter. And in Antarctica it's the big wild card that...they were very conservative in the last IPCC report, saying, well, maybe 40 centimeters by the end of the century. But they were leaving out some of the Arctic and large possibilities and Antarctica's a real wild card. If you do a probability, there's like a 10% chance of a meter and a half, for example, which is just based on the uncertainty from Antarctica.

MAN 1 IN AUDIENCE>> Yeah, I read the other day if the entire face glacier were to melt, it's starting to move faster now, that's 87 meters of sea-level rise right there. It's as big as a state, so, yeah.

DR. OVERLAND>> Yeah, if they always say, well, you know, it could be 10 feet if all of the ice in Greenland would melt. The Greenland ice is in this bowl, and that's really misleading. The amount that could melt or runoff would be more like this 50 centimeter effect even if everything that could runoff, did. So, at least for Greenland, that's more of a...more realistic.

Sue?

SUE>> So Jim, your next to last question there, I wonder if you could comment a little bit on it because as we've talked so many times, observations are galloping ahead of model predictions. No matter how you do the ensembles, not that I know about the modeling but--I wonder if you could just comment a little bit on that. I mean how...the models are missing it. So how do you take a next step in that regard? Because observed information, as you just said, the last three years...really stunning.

DR. OVERLAND>> Well, it's a little disappointing 'cause the last two rounds of assessments and models, there really wasn't much of an improvement in the models. And we're in the next round that's going to come out in 2020 or 21, and they started early to improve the physics so the hope was that the next round will be better but it's not clear that that's going to happen. 

Part of the difference is they can't do our clouds very well. And our clouds...the ice wants to compensate. Too many clouds, too little ice and vice versa. And so, you know, the point from the models is all the curves are up. But you'd like to have a quantitative number for 2030 or 2040. And you're not going to get it from the models. Basically extrapolating what we see now and, you know, and adding 20%, because we think maybe things are accelerating. It's not a pretty picture and trying to be quantitative rather than qualitative, that everything's falling apart, but we all know how fast for sure.

Yeah?

MAN 2 IN AUDIENCE>> I was part of a group from Marine Mammal lab that just got back from a trip to the Bering Sea for the month of April. So we saw all these changes firsthand and it was just staggering. And we were out there just trying to study the seals that use the ice and we are working, you know, 100-plus miles north of where we normally work. 

In fact I think we set a record for the NOAA Ship...furthest north the NOAA ship has ever gone in the spring. We were working up...ultimately up outside of Norton Sound, an area we've never had to work before. And, but in any case, so we did see striking changes with the quality and the coverage of ice, especially out west toward the Date Line where we'd expect to see a lot of seals and are very concerned about the reproductive success of those seals this year. We did...the further north we went, we experienced some really cold temperatures, you know, down to minus five. And even over a short period, a day or a day or two, we'd see sea ice start to set up in calm conditions. And, but of course, you know, after a wind event we'd see just massive areas of ice just disappear almost overnight. We be kind of using the satellite imagery day-to-day and we'd target an area, maybe say, south of St. Lawrence Island and then just like magic overnight the satellite would show that the ice just disappeared.

So I'm just wondering more about the feedback between the frequency of storms and the ability of ice to set up and whether there's any positive feedback loops between those two variables.

DR. OVERLAND>> Thanks for the comments. That's the characteristic of this year that...that winter is supposed to be cold and dark and so did you say when it looked like winter things happened really fast on the freeze-up but this year in late February and so on. Rather than normally northerly winds that cool everything down and drive the ice towards the south, we had this period of winds from the southwest that were warm and kept coming in and blowing the ice north but also bringing warm temperatures and melting what was already forming.

And that's a pretty...those southwest winds are not unusual but the amount of winds this year were unusual. So looking forward, by average, I wouldn't expect next year to be such an extreme as this year. You would expect winter would be...reassert itself, somewhat. 

But we're storing more heat in the Artic and if we had, you know, part of what was going on in the spring was that real delay in the fall. And I think we're going to have the delay in the fall every year. So what's going to happen, you'll probably not going to have all the warm air in the springtime but the whole cycle will be delayed because of the late fall freeze-up.

Yeah, thanks for the comment. 

Okay, any comments from the...?

Okay, thank you very much!

[APPLAUSE]