1
00:00:00,690 --> 00:00:15,920
Okay. I think I'll start. I guess it's true to say that the problem with climate change is, I would say about 90% a problem in physics.

2
00:00:18,060 --> 00:00:26,550
But when we try to apply physical theory and models to try to figure out what's going to happen, let's say in the next ten years,

3
00:00:27,690 --> 00:00:33,510
we realised there are big uncertainties and many things that we don't actually understand about our climate system.

4
00:00:34,860 --> 00:00:40,950
So we can look at the observations, but the instrumental record only goes back about 100 years.

5
00:00:42,840 --> 00:00:47,220
And so we're still left with lots of puzzling questions about how climate works.

6
00:00:47,730 --> 00:00:53,130
So a third option is to go right back into into history and look at paleoclimate data.

7
00:00:54,300 --> 00:00:58,620
And today's speak is is an expert in precisely that area.

8
00:00:59,730 --> 00:01:05,910
So we're very pleased to welcome Eric Wolff from the Earth Sciences Department in Cambridge.

9
00:01:06,750 --> 00:01:13,860
Eric actually spent much of most of his professional life at the British Antarctic Survey

10
00:01:14,730 --> 00:01:21,030
from getting his Ph.D. there through to rising to become the chief scientist of Bass.

11
00:01:21,960 --> 00:01:23,470
But about a year and a half or so ago,

12
00:01:23,470 --> 00:01:33,000
he became one of the Royal Society's research professors and moved to a more academic position in Cambridge, where he is now.

13
00:01:33,510 --> 00:01:39,720
So very happy, Eric, for you to come over today and talk to us about ice cores, climate and science.

14
00:01:41,400 --> 00:01:44,880
Okay. Thank you. Microphones working? Yes, I can hear it. Okay.

15
00:01:45,840 --> 00:01:48,540
So thank you very much for inviting me. And also, it's very nice.

16
00:01:48,540 --> 00:01:53,050
I can see because I know because I met Summers today that there are people from several departments here.

17
00:01:53,070 --> 00:01:54,330
So that's good.

18
00:01:54,570 --> 00:02:02,100
And that's important because actually ice cores themselves are a slightly well, they're not a strange subject, but they're the most important subject.

19
00:02:02,580 --> 00:02:09,270
But ice cold people in some places are in Earth sciences departments, geography departments, chemistry departments.

20
00:02:09,270 --> 00:02:13,410
But actually in quite a lot of places, they are in physics departments and there is a lot of physics in ice cores.

21
00:02:13,830 --> 00:02:17,490
So I hope you'll spot some of it as we go along and see things that might be of interest to you.

22
00:02:18,810 --> 00:02:21,360
I wasn't quite sure where to pitch this talk.

23
00:02:21,810 --> 00:02:30,240
So what I've done is to the first half is more or less a tutorial about ice cores, why they're so important, and what you can learn from them.

24
00:02:30,810 --> 00:02:32,730
If any of you have been to one of my talks before,

25
00:02:32,730 --> 00:02:36,630
and I know there are some people who have you can go to sleep during that part because you've heard it before.

26
00:02:37,590 --> 00:02:41,399
And then the second half is about what one of the things I'm working on at the moment,

27
00:02:41,400 --> 00:02:45,120
which is about the possibility of learning about sea ice from ice cores.

28
00:02:47,040 --> 00:02:52,200
And the third part is about future ice core research could, based on my practice, I very much doubt will get to that.

29
00:02:55,680 --> 00:03:03,329
So ice cores are just one of many paleoclimate sedimentary records that you can look at things like tree rings,

30
00:03:03,330 --> 00:03:10,740
marine sediments, lake sediments and so on. And ice cores are one of them and they have, like all of them, advantages and disadvantages.

31
00:03:11,550 --> 00:03:18,330
So the advantage is that they're pretty well dated, but probably not quite as well dated as you think.

32
00:03:18,870 --> 00:03:25,170
That's to say there are ice cores where we can count on your layers in the same way you can count tree rings or valves in lake sediments,

33
00:03:25,590 --> 00:03:30,690
but we can't do it perfectly. And actually in the cores that are oldest, which are the most interesting ones,

34
00:03:31,260 --> 00:03:35,580
to some extent, you can't count the layers and it becomes much more difficult to date them.

35
00:03:36,060 --> 00:03:43,040
We'll come back to that later. They have atmospheric signals which are relatively simple,

36
00:03:43,050 --> 00:03:48,930
and that's to say they haven't gone through any biology and they haven't gone through lots of water circulation before they get into the ice.

37
00:03:49,590 --> 00:03:53,190
Of course, that's where the physics comes in they turn out to be not that simple at all,

38
00:03:53,190 --> 00:03:57,560
but we like to think of them as simple and you get many variables on the same core.

39
00:03:57,570 --> 00:04:02,100
For example, the carbon dioxide concentration and the local temperature in the same core.

40
00:04:04,080 --> 00:04:08,340
I'm going to go to emphasise one problem with ice cores, which is the geographical limitation,

41
00:04:08,340 --> 00:04:12,840
because although as a polar scientist, I think the polar regions are very important.

42
00:04:13,320 --> 00:04:17,930
I do understand that trying to understand climate without knowing much about the tropics is quite difficult.

43
00:04:17,940 --> 00:04:22,920
So ice cores by themselves, or at least polar ice cores are never going to answer all our questions.

44
00:04:24,990 --> 00:04:27,270
So that's my favourite map projection.

45
00:04:27,270 --> 00:04:33,929
I'm fortunate I come after I set this to this bit of this picture up about 20 years ago and can no longer remember what projection it is.

46
00:04:33,930 --> 00:04:44,339
So it kind of loses the force of it. But with the important areas of the world in white, you can drill ice cores, elsewhere you can drill ice cores.

47
00:04:44,340 --> 00:04:51,330
And people do in the Alps, in the Andes, in the Himalayas, even in Africa, at Kilimanjaro.

48
00:04:52,590 --> 00:04:56,550
They're much more difficult to interpret, and they tend to have more local signals.

49
00:04:56,670 --> 00:05:02,730
Some of the properties that we measure don't mean the same in tropical ice cores as they do in polar ice cores.

50
00:05:03,000 --> 00:05:09,090
And I'm not going to mention them again so that they're important, they're useful, but we're not going to talk about it today.

51
00:05:11,730 --> 00:05:17,430
So the ice cold signal builds up like any sedimentary record by something getting imported into the ice at the surface.

52
00:05:17,610 --> 00:05:21,120
In the snow and then buried by the next year's layer.

53
00:05:21,120 --> 00:05:22,790
In the next year's layer. In the next year's layer.

54
00:05:22,800 --> 00:05:30,060
Again, I apologise to the Earth scientists for whom this is far too simple, but the physicists may not quite have appreciated how these records work.

55
00:05:31,020 --> 00:05:35,370
The difference with ice is the physics that ICE thins as it gets deeper.

56
00:05:35,970 --> 00:05:39,660
So if you really want to get old ice, you do need to go quite near the bottom.

57
00:05:39,660 --> 00:05:46,920
You're not going to get half the layers by drilling half thickness. You need to drill 80 or 90% of the thickness to get half the layers.

58
00:05:50,260 --> 00:05:51,800
And that's what an ice core looks like.

59
00:05:51,850 --> 00:05:57,850
Like to show that near the beginning of a talk, because I think people have a strange idea of what we might be talking about sometimes.

60
00:05:58,540 --> 00:06:00,700
So typically ten centimetre diameters.

61
00:06:01,390 --> 00:06:10,720
And when I talk about a 3000 metre ice core, what I mean is lots of two metre long pieces sort of like this, not 3000 metres laid out somewhere.

62
00:06:14,430 --> 00:06:18,659
Okay. There's it's a Friday afternoon, so we're all entitled to a bit of entertainment.

63
00:06:18,660 --> 00:06:21,840
So here's a video of drilling an ice core to give you an idea what it's like.

64
00:06:21,840 --> 00:06:24,870
This is Dome C in the centre of Antarctica. Very flat.

65
00:06:24,870 --> 00:06:32,250
There's nothing to see except this tent, which is the drilling tent, which has to be tall to accommodate the drill as the drill going into the ground.

66
00:06:32,400 --> 00:06:36,000
I know it's going to would, but that's just to keep snow from falling into the hole.

67
00:06:37,500 --> 00:06:43,290
And then at the top, there's a section that clings on to the hole you've already drilled in order to drill the next bit and then the cable.

68
00:06:43,800 --> 00:06:51,209
So it's a cable drill. There are 2000 metres to the drill, sit in the heated cabinet while the scientists get cold outside.

69
00:06:51,210 --> 00:06:53,610
It's. It's just a rule. I'm sorry. It's the way it is.

70
00:06:55,770 --> 00:07:00,989
And then the drill comes to the surface and you can put it to the horizontal to get the ice out.

71
00:07:00,990 --> 00:07:05,250
You can see the ice glistening that we can see the ice glistening there.

72
00:07:09,050 --> 00:07:12,470
You can see some liquid dripping out. Don't worry, it's not melting. That's a drilling fluid.

73
00:07:12,680 --> 00:07:14,780
You can ask me about drilling fluid later if you want.

74
00:07:17,300 --> 00:07:23,480
And although actually I've realised you don't actually see it, the person's about to do the most important thing you could ever do to an ice core,

75
00:07:23,480 --> 00:07:26,629
which is to put a pencil mark with an arrow on it pointing,

76
00:07:26,630 --> 00:07:35,240
which is up because because turning 100 metres 100 years, the wrong way up can create interesting climate records.

77
00:07:39,320 --> 00:07:44,149
That's the ice sitting, waiting. It's actually being cut at this point. You can see it's no longer got the fractured surface.

78
00:07:44,150 --> 00:07:52,370
It's never a cut surface. And then I'll just leave it for another few seconds just to show you how we handle ice in the field.

79
00:07:52,370 --> 00:07:58,620
This is just handling the ice to cut it up for sections to go back to the laboratories and you can just handle it with a band.

80
00:07:58,620 --> 00:08:03,529
So also, if you want to go into the field to do ice cold science, I strongly recommend the course in carpentry.

81
00:08:03,530 --> 00:08:13,729
It's the most important skill you could have and it's cut into lots of different sections and then packed up and sent back to labs,

82
00:08:13,730 --> 00:08:23,840
or it might be analysed in the field in some cases. I think we'll cut the cut at that point in the hope that we might get to the end of the talk.

83
00:08:25,340 --> 00:08:29,930
So signals in our schools, what do we get out of ice cores?

84
00:08:29,930 --> 00:08:33,440
So I like to think of it as having three types of information.

85
00:08:33,440 --> 00:08:39,709
Other people might think of it in a different way, but this is the way I classify it. So the first one is the isotopic content of the water itself.

86
00:08:39,710 --> 00:08:41,480
So that is exactly what you think.

87
00:08:41,480 --> 00:08:49,730
It's the ratio of 18 of water to 60, no water or deuterium to hydrogen, and it's determined mainly by the temperature at the time of snowfall.

88
00:08:50,060 --> 00:08:58,639
Now, I could give you a lecture about why that so and I could give you another lecture about why it's not really so I would get caught out.

89
00:08:58,640 --> 00:09:05,210
I did once get asked, why don't you give a lecture? But actually I've not written it, but I'm just saying that there is potentially a lecture on that.

90
00:09:06,530 --> 00:09:10,970
So you're going to have to take it as read for the rest of the talk that it gives you temperature.

91
00:09:11,180 --> 00:09:17,720
There are ways of calibrating it so that you really can back the temperature out, but they require much more complex analysis.

92
00:09:20,090 --> 00:09:24,049
Then there are lots of soluble and insoluble impurities that are trapped on the surface,

93
00:09:24,050 --> 00:09:29,570
so they're relatively straightforward to think about, but actually very difficult to interpret.

94
00:09:29,900 --> 00:09:32,959
So there are things like the sulphate that comes from volcanic eruptions.

95
00:09:32,960 --> 00:09:40,520
So when a volcano erupts, sulphur dioxide goes up into the stratosphere and you will end up for a couple of years after the big of a big volcano,

96
00:09:40,520 --> 00:09:44,450
you'll end up with an increase in sulphate in across the world,

97
00:09:44,450 --> 00:09:48,739
but in ice cores in particular, which tells you how strong,

98
00:09:48,740 --> 00:09:54,650
something about how strong the volcano was and actually helps us to date the ice cores by matching between different cores.

99
00:09:55,370 --> 00:10:02,120
But there are all sorts of other things, sea salt, which I'll be talking about later dust chemicals like nitrate,

100
00:10:02,570 --> 00:10:10,400
which I've worked on for probably 25 years and still don't understand what it means in articles and all sorts of other things.

101
00:10:11,300 --> 00:10:17,000
And we're just starting to work on various organic chemicals as well that haven't yet been much investigated and ice.

102
00:10:18,470 --> 00:10:21,709
And then finally the thing that ice cores are probably most famous for and which

103
00:10:21,710 --> 00:10:26,060
again requires a lot of physics to really understand as the snow gets deeper,

104
00:10:26,630 --> 00:10:34,670
whereas here in Oxford, when it snows, it probably melts the next day and then refreezes and you've got solid ice.

105
00:10:34,670 --> 00:10:41,540
That's not how you get ice in Antarctica. You get ice by the snow, partly sintering and partly through vapour movement.

106
00:10:42,770 --> 00:10:46,820
So sintering under pressure and vapour movements forming next between the crystals.

107
00:10:47,510 --> 00:10:56,660
It turns from loose snow eventually into a solid ice matrix, but it still has pools in it that finally close off and trap air in them.

108
00:10:57,350 --> 00:11:03,380
So that air is younger. It's only however long it takes to diffuse through a few tens of metres of firn.

109
00:11:03,620 --> 00:11:08,120
That's unconsolidated ice and it is literally a sample of air.

110
00:11:08,150 --> 00:11:16,280
You can crack the bubble open and measure the nitrogen, oxygen, argon, carbon dioxide, methane isotopes of various things and so on.

111
00:11:16,280 --> 00:11:20,960
Anything that stable, not not photochemistry unstable species, but anything else.

112
00:11:22,730 --> 00:11:30,290
However, it does require a lot of physics because first we need to understand the age of the air compared to the age of the ice.

113
00:11:30,290 --> 00:11:36,230
They're not the same because you've got young air being closed in relatively old ice, that maybe 70 metres depth.

114
00:11:37,490 --> 00:11:41,180
And that varies with time. You need to check that there isn't any diffusion.

115
00:11:41,180 --> 00:11:45,110
And eventually when we get down to 800,000 year old ice, there is some diffusion.

116
00:11:46,580 --> 00:11:49,100
You need to check that you haven't lost anything out of the core.

117
00:11:49,130 --> 00:11:53,900
And eventually when you get to ice that hasn't been stored properly, you can lose material out of the core.

118
00:11:53,900 --> 00:12:00,530
So it's it's there's a lot of work involved in these records, but that's just to show you, there really are bubbles in the ice.

119
00:12:00,530 --> 00:12:03,800
That's a finish section, a few millimetres across of ice.

120
00:12:04,550 --> 00:12:11,030
And you can see the bubbles there, typically a few tenths of a millimetre across when they're relatively near the surface.

121
00:12:11,690 --> 00:12:20,420
Actually deep down under pressure, they turn into the the air in the ice turns into collect right hydrate molecules within the ice matrix,

122
00:12:20,840 --> 00:12:28,100
but it can still be extracted more or less as if it was bubbles. So we don't need to know the distinction for at least for what I'm showing you today.

123
00:12:30,290 --> 00:12:34,910
So I'm just going to give you three examples of what we've learned from ice cores.

124
00:12:35,810 --> 00:12:39,290
The first one is from the very recent past, and then we get older.

125
00:12:40,190 --> 00:12:46,610
So from the very recent past, it can be very brief. It's a curve that you have all seen in one form or another hundreds of times.

126
00:12:46,690 --> 00:12:50,000
Probably, but maybe don't know where it came from.

127
00:12:50,020 --> 00:12:53,830
So this is carbon dioxide in ice over the last thousand years.

128
00:12:55,990 --> 00:12:59,320
What you've actually got is I'll have to walk away. But that's okay, I think.

129
00:12:59,560 --> 00:13:04,540
Yes. Well, you've got the blue line. There is measurements made in the atmosphere since the.

130
00:13:10,840 --> 00:13:16,669
All right. I went to China recently and needed to load some software to get at websites.

131
00:13:16,670 --> 00:13:19,670
The Chinese don't like to get out, so that's what that was.

132
00:13:20,840 --> 00:13:26,709
So. That's atmospheric measurements from actually from Manilla in this case.

133
00:13:26,710 --> 00:13:32,710
But it could be South Pole or any of the other sites that was started 30, 40, 50 years ago.

134
00:13:33,100 --> 00:13:36,160
Manilla happened to be the first one back in the 1950s.

135
00:13:36,340 --> 00:13:45,280
And then everything else come on here comes from high schools and shows you I like this way of showing you because it shows you well,

136
00:13:45,280 --> 00:13:53,710
firstly, it shows you that things went unusual in about 1830 and that the increase has been 40% since then.

137
00:13:54,820 --> 00:14:03,670
It shows you that there are variations before that, whether they're natural or unnatural variations, is still a cause of debate in the community.

138
00:14:06,070 --> 00:14:11,020
They're variations, though. And also there are four different ice cores here which have very different properties,

139
00:14:11,290 --> 00:14:16,060
different temperatures, different snow accumulation rates, different concentrations of other impurities.

140
00:14:16,450 --> 00:14:20,979
So the fact that they all plot on top of each other should give you a very good assurance.

141
00:14:20,980 --> 00:14:26,560
But nothing about enclosing the air in the ice gives you an artefact, because if it is an artefact, you've got to be the same artefact,

142
00:14:26,800 --> 00:14:30,370
despite the fact that all the site, all the properties that might control the artefact are different.

143
00:14:34,380 --> 00:14:38,640
That was it for the recent past, I thought. We're not going to talk about the recent past anymore.

144
00:14:40,410 --> 00:14:45,479
So now we're going to go to the longest ice cold timescale of 800,000 years with the Epica Project,

145
00:14:45,480 --> 00:14:56,640
which I was the chief scientist for for the later years. And Epica was it was a consortium of labs from ten countries in Europe.

146
00:14:57,570 --> 00:15:02,280
I'm sure you all know your European countries, so you can mentally do the exercise of working out which they are.

147
00:15:02,910 --> 00:15:12,360
They were in alphabetical order. That's the clue. And we drilled two ice cores, one of them up here at Running Mode Land,

148
00:15:12,360 --> 00:15:18,239
which was for the last climatic cycle, about 130,000 years have actually been used.

149
00:15:18,240 --> 00:15:23,730
It's actually a bit longer than that. And then the one here, it don't see, which is what you saw in the video.

150
00:15:23,790 --> 00:15:30,659
So that's the oldest ice core that we've got so far and don't see is a very dry place.

151
00:15:30,660 --> 00:15:38,250
That's why the ice is so old, because the amount of snowfall that 25 kilograms per square metre per year, that's 2.5 centimetres of ice.

152
00:15:38,250 --> 00:15:42,600
Equivalent of water equivalent. Sorry. So an inch. Yes.

153
00:15:42,600 --> 00:15:47,639
So it's a desert by any standard. That's what don't see.

154
00:15:47,640 --> 00:15:50,670
Looks like you actually saw in the video just to relate it back to that.

155
00:15:50,850 --> 00:15:56,580
You saw the drilling temples there and the sun shelter that you saw them working in was just behind it there.

156
00:15:57,090 --> 00:16:03,900
That's what's there now. The Italians and French have built a permanent wintering station there and have people there all year round.

157
00:16:07,420 --> 00:16:11,490
And not to show that I was there. If you can't tell I was there, I could have set that up in the studio.

158
00:16:11,500 --> 00:16:18,880
But anyway, I was there. So these are the some of the records.

159
00:16:18,910 --> 00:16:20,890
You've got the water isotopes at the top.

160
00:16:21,370 --> 00:16:25,870
I've actually for the cognition, too, I've added in some numbers so that they know what what they're looking at.

161
00:16:27,220 --> 00:16:31,450
You've got dust represented by calcium.

162
00:16:31,450 --> 00:16:36,100
You've got the electrical properties of the ice, which was what I was actually measuring in the field and so on.

163
00:16:36,880 --> 00:16:43,000
It's not much use to like this because the things you don't recognise on the Y axis and depth on the x axis,

164
00:16:43,000 --> 00:16:48,400
which probably isn't of any great interest. So the next thing you need to know is how we construct the time scale.

165
00:16:49,000 --> 00:16:52,320
Now here comes lots of physics and maths, but I'm not actually going to give it to you.

166
00:16:52,330 --> 00:16:58,510
I'm going to go on to explain it in qualitatively what was done to construct this age scale.

167
00:16:58,520 --> 00:17:01,180
And there are various different strategies depending on the ice core.

168
00:17:01,840 --> 00:17:06,550
What was done here was to start off with a model of how much snow fell each year.

169
00:17:06,830 --> 00:17:11,470
Okay, I should start off by saying if you want to know how old you are at a given depth in any sediment,

170
00:17:11,980 --> 00:17:15,310
what you want to know is how many, how many annual layers are above you.

171
00:17:15,700 --> 00:17:20,140
And to know that you want to know how thick each of them is. So you know how thick each of them is.

172
00:17:20,320 --> 00:17:24,459
Your first bet is I know how thick it was when it fell as snow because I know how much

173
00:17:24,460 --> 00:17:28,570
snow falls at the surface and I know how thick it is now because I know how from physics,

174
00:17:28,570 --> 00:17:30,940
I know how it thins to this depth.

175
00:17:32,200 --> 00:17:38,140
So you use models of those two things to make a first estimate of the thickness with that would give you the age with that.

176
00:17:39,190 --> 00:17:43,990
But that would be very imprecise because you don't really know those things actually.

177
00:17:44,920 --> 00:17:52,270
So you then refine it with an inverse method where you add in lots of known ages and I won't talk about all the known ages we are today.

178
00:17:52,270 --> 00:17:56,110
And some of them were the ages of volcanic eruptions, which you can see in the South record.

179
00:17:56,740 --> 00:18:06,640
Some of them were from parameters in the ice core that are controlled by milenkovic cycles, the orbital cycles of the earth around the sun,

180
00:18:08,770 --> 00:18:15,159
which again we could discuss whether it's legal to use things like that where you're not quite sure what the relationship with the ice is,

181
00:18:15,160 --> 00:18:17,110
but there are good reasons for doing it.

182
00:18:17,710 --> 00:18:22,930
The only thing I want to do is to convince you about one of the time markers that we put in so that you believe me,

183
00:18:22,930 --> 00:18:28,720
that is 800,000 years at the bottom. And that comes from this.

184
00:18:28,960 --> 00:18:37,990
The brain is much young and magnetic reversal. So that's the last time the Earth's magnetic field reversed, dated and rocks at roughly 780,000 years.

185
00:18:38,410 --> 00:18:41,319
It occasionally shifts by a thousand years or so.

186
00:18:41,320 --> 00:18:48,520
When somebody does a new dating, what happens when the magnetic field reverses is that it's weak for a while.

187
00:18:48,520 --> 00:18:53,559
And so when magnetic fields weak, you get a lot of cosmic rays, cosmic rays produced.

188
00:18:53,560 --> 00:18:59,320
Well, most familiarly they produce carbon 14, but by 17 or 2000 years, that certainly isn't any carbon 14 left.

189
00:18:59,920 --> 00:19:03,220
But they also produce beryllium ten and there's 20 brilliant ten left.

190
00:19:04,120 --> 00:19:09,760
So you go to the section of coal where you think you should see this and you find the largest beryllium ten peak in the record.

191
00:19:10,240 --> 00:19:15,040
So then you say that is 780,000 years. I'm going to fix that now and make the rest of the times go.

192
00:19:15,910 --> 00:19:18,879
So there's lots more goes into the ice scale than this.

193
00:19:18,880 --> 00:19:26,710
There was an entire special issue of this age scale, but hopefully I convinced you that it's 800,000 years at the bottom.

194
00:19:27,760 --> 00:19:33,130
There's actually a new age scale now which uses lots of ice cores rather than just one and puts them all together.

195
00:19:33,940 --> 00:19:37,450
It's a better way of doing it because it uses more information,

196
00:19:39,520 --> 00:19:44,319
but there are people who say it isn't a better way to do it because it obscures what it is you're actually doing.

197
00:19:44,320 --> 00:19:55,330
So we've got a little bit of work to do to convince people that's the best way. So now we now we can produce a record.

198
00:19:55,330 --> 00:19:58,690
This is very old. I should have said a lot of what I'm putting up here is not my work.

199
00:19:58,690 --> 00:20:02,290
I was involved in it, but it's not. I wouldn't say it's my measurements.

200
00:20:03,130 --> 00:20:08,980
So be aware that's an entire communities work. So now this is what you really want to see.

201
00:20:08,980 --> 00:20:10,060
This is well,

202
00:20:10,390 --> 00:20:19,180
this is actually deuterium of the ice on the y axis and depth on the x axis converted to age and the jitter has been converted to temperature.

203
00:20:19,600 --> 00:20:24,489
So there's an uncertainty in that conversion. But it's it's within that uncertainty.

204
00:20:24,490 --> 00:20:28,360
This is the temperature of Antarctica over the last 800,000 years.

205
00:20:28,870 --> 00:20:34,870
And seeing I've got a mixed audience, I have to say, I've done this the probably the physicists way.

206
00:20:34,930 --> 00:20:39,790
That's to say time is going from left to right. Scientists prefer time to go the other way.

207
00:20:40,030 --> 00:20:47,410
But because they're thinking of depth. But physicists and I suppose I have to admit, normal people think of time as going from left to right.

208
00:20:48,070 --> 00:20:51,160
So so time going from left to right here.

209
00:20:51,160 --> 00:20:54,250
But don't worry, it'll flip around during the rest of the talk. So.

210
00:20:55,420 --> 00:21:02,750
So what we've got here is the Holocene, the last 10,000 years, it's lost 10,000 years of relative warmth there.

211
00:21:04,120 --> 00:21:13,269
And then the last glacial so the last place you probably think of it as ice sheet down to East Anglia or Wisconsin.

212
00:21:13,270 --> 00:21:18,660
But it was also much. Colder in Antarctica during the last glacial maximum, about ten degrees.

213
00:21:20,070 --> 00:21:25,370
And then the last interglacial which was actually a bit warmer and Antarctic than the present.

214
00:21:25,380 --> 00:21:29,220
I did think of giving a talk about the interglacial here,

215
00:21:29,220 --> 00:21:35,070
which would have been quite interesting cause I think it's a very important time to look at because we know in the last interglacial

216
00:21:35,070 --> 00:21:41,340
both the Antarctic and the Arctic at some point were quite a bit warmer than today and sea level was quite a bit higher than today.

217
00:21:41,820 --> 00:21:48,120
And working out what balance of two ice sheets created a sea level that was 5 to 10 metres higher in the last interglacial.

218
00:21:48,360 --> 00:21:56,819
It's probably one of the most direct ways that we can contribute to understanding what might happen in the future under a warming climate,

219
00:21:56,820 --> 00:22:06,060
but in particular warmer polar regions. Okay, so what we've got here is roughly eight glacial cycles.

220
00:22:07,020 --> 00:22:14,430
This was not unexpected. I mean, from marine sediments, we already knew very well that there would be cycles that looked a little bit like this,

221
00:22:15,240 --> 00:22:19,830
that had roughly 100,000 year cycles of warm and cold.

222
00:22:19,830 --> 00:22:28,770
Interglacial in glacial. And there's a tendency to stronger cycles in the later part of the record.

223
00:22:28,800 --> 00:22:36,360
I've done quite a bit of work since then. On looking at other records to see whether this is true and actually this general idea that the

224
00:22:36,870 --> 00:22:41,939
interglacial the warmer in the last 400,000 years than they were before is some kind of a break.

225
00:22:41,940 --> 00:22:50,970
Here is true in Antarctic temperature and it's true in carbon dioxide and it's true in some Southern Ocean records.

226
00:22:51,300 --> 00:22:53,160
But it's not globally true, actually,

227
00:22:53,160 --> 00:23:01,830
because there's there are there are periods that are weak in the second half of the record in some records and the other way around.

228
00:23:02,850 --> 00:23:06,450
Nonetheless, there's something something slightly different after here.

229
00:23:10,170 --> 00:23:16,319
Into that. Okay. So we've got those periods and the obvious question and the one we didn't know

230
00:23:16,320 --> 00:23:20,280
about before the ice cold came out is what do CO2 do in this changing climate?

231
00:23:20,940 --> 00:23:25,110
And this is the answer. So the plot at the bottom is exactly the record you saw before.

232
00:23:25,120 --> 00:23:31,350
Temperature on the one at the top is carbon dioxide from the same ice cold over the same period.

233
00:23:32,280 --> 00:23:37,080
And I think some of you are further back in the room, so probably they look absolutely identical.

234
00:23:37,110 --> 00:23:42,809
I do know where the bodies are buried, that there are some differences, for example, there,

235
00:23:42,810 --> 00:23:47,730
where the CO2 stays high while the temperature's already gone down a little bit before you get there.

236
00:23:48,300 --> 00:23:54,750
But nonetheless, they are incredibly similar. If you went out and drilled, I won't say marine sediments kill them, might offend people in the front.

237
00:23:54,750 --> 00:24:01,670
Right. So if you drilled to lake sediments ten metres apart and measured exactly the same thing in them, they would not look this similar.

238
00:24:01,680 --> 00:24:03,630
And yet these are two completely different things.

239
00:24:03,840 --> 00:24:10,050
I mean, deuterium and an ice core and carbon dioxide in the atmosphere, they couldn't be more different, but they look incredibly similar.

240
00:24:12,540 --> 00:24:17,879
So if we look at it from one direction, we could say, well, that carbon dioxide, based on just what we know about radiator forcing,

241
00:24:17,880 --> 00:24:22,220
would be responsible for something up to a half of the glacial interglacial warming.

242
00:24:22,230 --> 00:24:27,990
So it's an important factor in causing the glacial interglacial climate change that we see worldwide.

243
00:24:29,100 --> 00:24:36,389
But also the fact that they look so similar also tells you, I think, that the carbon dioxide must be being controlled mainly in the Southern Ocean.

244
00:24:36,390 --> 00:24:40,200
And in other words, by something rather similar to what's controlling Antarctic temperature.

245
00:24:40,890 --> 00:24:44,700
This doesn't look like Greenland temperature, for example.

246
00:24:44,940 --> 00:24:48,270
Greenland temperature would have a lot of other stuff in it, as we'll see in a minute.

247
00:24:50,490 --> 00:24:54,899
And then I always have to remind you that if we want to put the last 200 years on the same plot,

248
00:24:54,900 --> 00:24:58,740
we have to shrink that down so that we can see what happened in the last 200 years.

249
00:24:59,100 --> 00:25:07,560
So there isn't an analogue in this record. CO2 was not anything like as high as it is today at any point in the last 800,000 years.

250
00:25:07,890 --> 00:25:13,470
And the chances of that being something hidden in there that's as high as that are negligible.

251
00:25:19,580 --> 00:25:27,320
Just because I like in these talks to give you a few a few killer facts to use in case you meet climate sceptics or maybe you are a climate sceptic,

252
00:25:27,830 --> 00:25:36,670
just put your hand up. It's the right stuff. So there's always this thing about what happened during the termination coming out of the last Ice Age.

253
00:25:36,680 --> 00:25:42,680
So that's this section here. And did the CO2 lead the temperature or did the temperature lead to CO2,

254
00:25:42,680 --> 00:25:48,709
which actually doesn't really matter which led, but nonetheless, it's an issue that people keep raising.

255
00:25:48,710 --> 00:25:51,230
So it's better to show you what the latest data looks like.

256
00:25:51,230 --> 00:25:55,639
So you don't have to get complicated about it because that's what the latest data look like.

257
00:25:55,640 --> 00:26:01,100
This is two different ways of doing it on the bottom. One, it's the Dome C record that you saw before.

258
00:26:02,240 --> 00:26:06,680
That's the temperature and that's the CO2. And they've been I told you,

259
00:26:06,680 --> 00:26:11,090
there's this correction you have to make for the fact that the age of the air is slightly different from the age of the ice.

260
00:26:11,570 --> 00:26:21,049
And the the correction has been made using the nitrogen isotopes in this case have been used as have been made using the nitrogen isotopes in the ice,

261
00:26:21,050 --> 00:26:28,550
which I haven't talked about, which is a very nice piece of physics about the gravitational column that the

262
00:26:28,550 --> 00:26:32,330
air is diffusing through before it becomes solid ice to make the correction.

263
00:26:33,560 --> 00:26:38,000
And you can see, well, then at the top it's completely different.

264
00:26:38,000 --> 00:26:45,260
What's been done there is to use coastal ice or ice cores closer to the coast of Antarctica where the snowfall rate is much higher.

265
00:26:45,650 --> 00:26:48,770
So that problem of synchronising the air and the ice is much smaller.

266
00:26:48,770 --> 00:26:53,390
The difference, the age difference is much smaller. The uncertainty in synchronise them is much smaller.

267
00:26:54,140 --> 00:27:03,170
And in both cases what you see is that the temperature and CO2 within any reasonable uncertainty start to rise in exactly the same moment.

268
00:27:03,170 --> 00:27:04,879
So we don't have to argue about which came first.

269
00:27:04,880 --> 00:27:09,650
Now, they both went together exactly as you would expect if they were mutually amplifying each other.

270
00:27:10,550 --> 00:27:15,200
Chickens and eggs come to mind, but that there isn't a lag.

271
00:27:15,500 --> 00:27:23,819
There isn't the problem. So that's ice cores for glacial cycles.

272
00:27:23,820 --> 00:27:29,700
But as I said before, I just want to remind you and remind myself at this point that I schools are wonderful, but they don't tell you everything.

273
00:27:29,700 --> 00:27:37,169
So you definitely need to add to them. Everything else around the world in order to understand what's happening to climate change during these cycles.

274
00:27:37,170 --> 00:27:43,170
So this is actually a marine record that, again, looks similar, has the same cycles as a terrestrial record,

275
00:27:43,170 --> 00:27:50,010
a tree pollen record from Greece with the same glacial and glacial cycles on it and so on.

276
00:27:51,900 --> 00:28:05,190
Okay, so that's the second example. And then the third example, I just want to focus on this last 130,000 years and that's Antarctica.

277
00:28:05,550 --> 00:28:07,170
And then I want to put Greenland on as well.

278
00:28:07,170 --> 00:28:12,540
So again, what you see, what you see up here is actually what you saw before so that it says deuterium rather than temperature.

279
00:28:15,250 --> 00:28:22,450
And it's still running left to right time, so we haven't switched it around yet as the Holocene left to last close to maximum and then hit Greenland.

280
00:28:23,080 --> 00:28:26,739
So Antarctica looks sort of a bit boring.

281
00:28:26,740 --> 00:28:31,900
It's it's moving quite slowly. It's doing some interesting things, but it's not doing that much.

282
00:28:32,320 --> 00:28:38,170
Greenland, on the other hand, even at this scale, is going crazy throughout the last glacial.

283
00:28:41,590 --> 00:28:44,950
Just look at Greenland now. These are things called dance. Go to rescue events.

284
00:28:44,950 --> 00:28:47,650
That's really done. Scott and Huntzberger off to whom they're named.

285
00:28:49,570 --> 00:28:56,620
And then if we just focus in on the last 40,000 years so that you can see what we're looking at in a bit more detail.

286
00:28:59,850 --> 00:29:04,290
These are what they look like. It can be very cold in Greenland.

287
00:29:04,560 --> 00:29:08,230
And then suddenly the temperature jumps by something of the order, ten degrees.

288
00:29:08,250 --> 00:29:13,470
Some of these are actually 15 degrees, but ten degrees is a typical number for the jump here.

289
00:29:14,010 --> 00:29:18,570
And that looks instantaneous. It's actually about 40 years when you look in detail in the Greenland record,

290
00:29:19,440 --> 00:29:23,700
and then it sort of slowly cools and then jumps back down to the bottom again.

291
00:29:24,270 --> 00:29:28,770
It stays cold for a little while, for a few hundred, maybe a thousand years, and then jumps up again.

292
00:29:29,190 --> 00:29:34,799
And it keeps doing this. We've got there are 23 numbered events, but there are actually more events than that,

293
00:29:34,800 --> 00:29:39,709
depending on how you count them, where Greenland is doing this strange thing.

294
00:29:39,710 --> 00:29:42,180
Now, I should really emphasise this is Greenland doing this.

295
00:29:42,180 --> 00:29:46,110
So it's not that the world is doing this and it's not even that the Northern Hemisphere is doing this.

296
00:29:46,650 --> 00:29:52,620
You certainly do get similar records from the North Atlantic, from European records,

297
00:29:53,010 --> 00:30:00,100
and you get similar periodicity in other records like Cave Records in China that are recording something different,

298
00:30:00,150 --> 00:30:02,100
not not necessarily recording temperature.

299
00:30:02,760 --> 00:30:07,920
So there's something that's spread around the whole hemisphere and there's a warming in the northern hemisphere.

300
00:30:08,850 --> 00:30:11,159
But it's not a global change in that.

301
00:30:11,160 --> 00:30:15,450
Well, it's a global change, but it's not a global it doesn't globally look like there's nothing I'm trying to say.

302
00:30:18,000 --> 00:30:20,879
So these are very interesting things they've been much written about,

303
00:30:20,880 --> 00:30:25,740
including by other people in the audience far more than me, so I shouldn't label them.

304
00:30:26,130 --> 00:30:31,320
One particular characteristic of them is something called the bipolar seesaw you just saw flashing

305
00:30:31,320 --> 00:30:35,400
up something there to warn you that this this plot does go from right to left just to confuse you.

306
00:30:36,720 --> 00:30:43,050
So what's happening? That's a similar record to what you saw before? It's a different Greenland core, actually, but it shows you the same thing.

307
00:30:43,320 --> 00:30:48,510
And what happens is that when Greenland is cold here, Antarctica is warming.

308
00:30:49,230 --> 00:30:55,500
And when Greenland is warm, Antarctica is cooling. And that can actually be written down as an equation, a rather simple equation.

309
00:30:55,500 --> 00:30:58,860
It's just a one once just the integral of the other effectively.

310
00:30:59,700 --> 00:31:08,790
And it works surprisingly well, just as that equation, which I think is cool, but it's known as the bipolar seesaw equation.

311
00:31:08,790 --> 00:31:13,739
It definitely works. Whatever you think the mechanism is, and most of us think we know what the mechanism is.

312
00:31:13,740 --> 00:31:16,110
But whatever you think it is, the equation works.

313
00:31:18,380 --> 00:31:24,950
And what's suggested is and I don't actually know whether because I think I know where this plot came from.

314
00:31:24,950 --> 00:31:30,110
So I don't know whether it replaced. I put it up or sorry, I put it up in an oceanography department where you hate cartoons like this.

315
00:31:30,530 --> 00:31:39,280
So so as you know, I think in this department is a lot of heat travel through the ocean into the North Atlantic.

316
00:31:39,290 --> 00:31:43,610
And the idea behind this, which I'm sure many of you know about already,

317
00:31:44,300 --> 00:31:53,000
is that this thinking up here gets disrupted by fresh water off the ice sheets at some points during the last glacial.

318
00:31:53,360 --> 00:31:57,379
And that disrupts this whole flow of heat to the north, either takes it further south,

319
00:31:57,380 --> 00:32:03,410
weakens it or makes it shallower, or some combination of the three, and then it switches on again later.

320
00:32:04,250 --> 00:32:07,430
And there are all kinds of variants on this. There are various variants on this.

321
00:32:07,430 --> 00:32:12,010
It certainly requires sea ice to be doing something quite strong.

322
00:32:12,020 --> 00:32:16,580
In other words, advancing during the cold periods and retreating during the warm periods to amplify the effect.

323
00:32:17,300 --> 00:32:22,070
Some people have suggested it may involve the growth and shrinkage of ice shelves somewhere.

324
00:32:23,780 --> 00:32:30,200
I'm working with some people in this room actually on an idea about how runoff into the Arctic may affect this process.

325
00:32:30,530 --> 00:32:39,080
But anyway, the basic principle seems to be seems to be the and the significance of these events are that rapid change that occurred into the past.

326
00:32:39,380 --> 00:32:43,880
In the past, as far as we know, only when there are large ice sheets. But nonetheless,

327
00:32:43,880 --> 00:32:50,690
they should make us just a little bit worried that the climate system isn't going to behave in the regular way that we might think it's going to.

328
00:32:54,020 --> 00:32:58,370
So that was all the general stuff. So if you already knew all about ice cold, you can wake up now.

329
00:32:59,510 --> 00:33:03,090
And if you really didn't know anything about it, I'm probably going to get too complicated now.

330
00:33:03,620 --> 00:33:05,630
So half of you can leave now.

331
00:33:05,930 --> 00:33:12,139
But so now I'm going to talk about something I'm working on at the moment, and it's something where we haven't got a final answer.

332
00:33:12,140 --> 00:33:15,470
So this isn't going to come to some brilliant conclusion that, yes, we've solved everything.

333
00:33:16,880 --> 00:33:20,780
The question is, is sea salt in ice cores a proxy for past sea ice extent?

334
00:33:21,440 --> 00:33:27,170
So what we do in paleoclimate is to measure something as a proxy of something else, water isotopes as a proxy of temperature.

335
00:33:27,800 --> 00:33:31,370
And we've got this idea that sea salt may be a proxy for past the ice extent.

336
00:33:31,820 --> 00:33:38,570
And it's work led by my colleague James Levine. So why are we so interested in sea ice?

337
00:33:38,870 --> 00:33:43,219
Well, I hope I don't have to convince anyone here that it's a very critical component

338
00:33:43,220 --> 00:33:46,950
of climate for various reasons of which I've only written down a few here.

339
00:33:48,950 --> 00:33:51,440
If you don't get the sea ice right, you won't get anything else right.

340
00:33:53,570 --> 00:33:59,330
But unfortunately, we've got very short time series available given the sort of oscillations there are in the atmosphere there.

341
00:33:59,620 --> 00:34:03,620
You know, we've got two or three cycles of some of them and less than one cycle of some others.

342
00:34:04,640 --> 00:34:06,710
In satellite data, it's been 40 years.

343
00:34:07,310 --> 00:34:13,100
There are ship borne and airborne observations that have been used to try and extend back the Arctic record to the 19th century.

344
00:34:13,430 --> 00:34:15,499
For the Antarctic record, you can't do anything like that.

345
00:34:15,500 --> 00:34:22,729
There are some whaling records which are highly disputed and as well it's not something where we

346
00:34:22,730 --> 00:34:27,290
can just rely on the models to tell us the answer because the models have some serious deficiency,

347
00:34:27,320 --> 00:34:30,500
even in the simplest things, the current extent and trends.

348
00:34:30,500 --> 00:34:35,450
And I put this up not in order to embarrass the ice modellers, but just because it's the it's the truth.

349
00:34:36,410 --> 00:34:41,989
The black line is the data for Antarctic sea ice extent in February.

350
00:34:41,990 --> 00:34:45,950
So that's the minimum. February's the minimum because it's summer in Antarctica.

351
00:34:46,580 --> 00:34:54,680
The black line is the data from the satellites and these are the IPCC class models, what they say is happening.

352
00:34:55,280 --> 00:34:59,419
So firstly they're all over the place in the extent and secondly they all say

353
00:34:59,420 --> 00:35:03,440
that sea ice extent is decreasing and it hasn't decreased in Antarctica yet.

354
00:35:05,180 --> 00:35:09,020
So they're clearly doing badly now. It's not this is not a criticism of sea ice modellers.

355
00:35:09,020 --> 00:35:11,540
It's probably not the sea ice that's wrong in these models, actually.

356
00:35:11,540 --> 00:35:16,580
It's the fact that there are biases in the in the temperature of the ocean is one is one issue.

357
00:35:17,030 --> 00:35:20,030
And there are other things to do with the winds that are driving the sea ice around.

358
00:35:20,030 --> 00:35:24,830
So it's probably not actually the sea ice that's wrong. But the end result if the sea ice is wrong.

359
00:35:29,340 --> 00:35:38,489
What was the solid, right? Uh, the solid red is I think the MULTI-MODEL mean, I think.

360
00:35:38,490 --> 00:35:46,740
I'm not absolutely certain now. I'm sorry. I'd have to look that up. Yes, it must be the Multi-model mean, because it declines.

361
00:35:46,740 --> 00:35:50,610
I'm pretty sure it's the Multi-model mean, sorry, the models are terrible, but on average they're good.

362
00:35:51,930 --> 00:35:56,440
I don't know. I don't know what that tells us, though.

363
00:35:57,030 --> 00:36:01,049
They're good. They're good in extent. But they do. They do have the decline, which hasn't been seen.

364
00:36:01,050 --> 00:36:12,660
So. So there would be a huge benefit to getting a sea ice proxy, a really good sized or a really good set of sea ice proxies.

365
00:36:12,780 --> 00:36:18,870
Apart from anything else, when you're doing paleoclimate modelling, trying to look at the climate of the last glacial,

366
00:36:18,870 --> 00:36:22,080
it would be really good to be able to check that we hadn't sea ice roughly, right?

367
00:36:22,650 --> 00:36:26,549
Because if we had the sea ice from, we'll certainly have everything else wrong on this issue.

368
00:36:26,550 --> 00:36:29,850
With the sea ice across the dance go to rescue events is one example of that.

369
00:36:32,400 --> 00:36:38,760
So the obvious way to do this is for marine sediments and they do there are sea ice proxies in marine sediments and they're very, very nice.

370
00:36:38,760 --> 00:36:46,319
They give time sequences of sea ice presence or absence at particular sites, mainly through biomarkers or biological proxies.

371
00:36:46,320 --> 00:36:51,180
And you can combine the sites to give a spatial picture. I'll show you that in a minute, but it would be very nice.

372
00:36:51,330 --> 00:36:58,530
That's very painstaking. You have to measure at every site and then put it together to get a spatial picture.

373
00:36:58,530 --> 00:37:02,100
And it'd be very nice to complement that with a regional picture from an ice core proxy.

374
00:37:02,940 --> 00:37:05,159
And there are two candidates of which I'm going to talk about.

375
00:37:05,160 --> 00:37:12,149
One, but just to give you an idea what I mean by that, so this is the best it's not quite the best now because it's is a few years out of date.

376
00:37:12,150 --> 00:37:19,410
But this is the the compilation most people show for sea ice at the last glacial maximum in and around Antarctica.

377
00:37:21,990 --> 00:37:30,990
And what's been done here is to draw lines on with the last glacial maximum winter stretch of maximum winter.

378
00:37:30,990 --> 00:37:33,800
That's the WAC. You can see where the data points are.

379
00:37:33,810 --> 00:37:40,290
So you can see why there are question marks there where there's no data and then the summer weather, just a few odd bits of line here and there.

380
00:37:42,930 --> 00:37:46,920
What we'd like to complement that with, I mean, people can go and get more cores and that would be great.

381
00:37:48,090 --> 00:37:54,629
But what we'd like to complement that with is an ice core that would tell you not how much ice there is as an individual place,

382
00:37:54,630 --> 00:37:58,950
but how far did it extend in a sector extending away from the continent?

383
00:37:58,950 --> 00:38:04,250
Away from my site. So why sea salt?

384
00:38:05,180 --> 00:38:09,740
Well, sea salt over most of the world comes from the ocean spray and bubble bursting.

385
00:38:12,470 --> 00:38:17,180
And so if if that was the case, then what you would imagine. I'll give you one of the punch lines first.

386
00:38:17,540 --> 00:38:22,580
What you would imagine would happen is that when you had more sea ice, the ocean's getting further away from you.

387
00:38:22,610 --> 00:38:26,690
So you would think that the amount of sea sea salt reaching your core decreased.

388
00:38:27,080 --> 00:38:33,080
But actually, every time we look, if we look at winter against summer or at the last glacial maximum against the present,

389
00:38:33,380 --> 00:38:36,610
what you find is the other way around, that when there's more sea ice,

390
00:38:36,620 --> 00:38:40,040
when we know there was more sea ice, we actually get more sea salt in the ice cores.

391
00:38:40,670 --> 00:38:48,110
And models can't explain that at the moment. So what we think is happening and we've done a lot of work to make us think that,

392
00:38:48,110 --> 00:38:55,940
is that a number of observations suggest that there's a source from the sea ice itself, from the surface of the sea ice itself, a source of sea salt.

393
00:38:56,390 --> 00:39:02,270
And there are various things, too, to two of them are what I just said, this counterintuitive idea that when you've got more sea ice,

394
00:39:02,270 --> 00:39:07,760
you actually get more sea ice, despite the fact that what you thought was the source is now further away from you, a lot further away from you.

395
00:39:09,200 --> 00:39:15,110
The second one is, is a chemical a very distinct chemical signature, which is what we call negative non sea salt sulphate.

396
00:39:15,710 --> 00:39:24,830
And I'll show you what that is. What it is is that if you plot the amount of sulphate in snow or aerosol against the amount of sodium,

397
00:39:25,310 --> 00:39:35,090
what should happen is that it should have a ratio of the point 25 to in weight from the seawater itself and then

398
00:39:35,090 --> 00:39:40,610
it should have a bit more that comes from other sources like volcanoes and biogenic emissions from the ocean.

399
00:39:41,810 --> 00:39:50,360
And so if you calculate this quantity here, it should always be positive because if there was only seawater and nothing else, it would be zero.

400
00:39:52,490 --> 00:39:58,910
What you actually find is that as you get more and more sodium, especially in the winter, it becomes more and more negative.

401
00:39:58,910 --> 00:40:03,110
This quantity, which means some of the sulphate that was in the seawater, has disappeared somewhere.

402
00:40:05,120 --> 00:40:08,830
And you find that everywhere you measure it around and around coastal Antarctica.

403
00:40:08,840 --> 00:40:13,999
Actually, if you go into central Antarctica it's harder to spot it because there's a very big background of other things,

404
00:40:14,000 --> 00:40:17,090
but you can spot it by doing size segregated aerosol sampling.

405
00:40:19,670 --> 00:40:23,600
So what we what is happening,

406
00:40:23,600 --> 00:40:27,930
what we're sure is happening is that the sulphate is being depleted because mirabilis

407
00:40:27,950 --> 00:40:32,900
it's being precipitated out of rubber like sodium sulphate precipitates at minus eight.

408
00:40:33,650 --> 00:40:36,770
So it's not unreasonable, it's being precipitated out.

409
00:40:36,770 --> 00:40:45,979
And of course the amount of sulphate when it freezes out precipitates out is a much higher proportion of the sulphate in the water than sodium is.

410
00:40:45,980 --> 00:40:52,850
So it gives you the impression of a depletion and the only mechanism that anyone can think of for how that happens,

411
00:40:53,210 --> 00:41:00,170
where you can freeze out metabolite and then separate the sea salt aerosol from the light left

412
00:41:00,170 --> 00:41:06,079
behind somewhere else is by that happening on the sea ice surface before the uplift of the aerosols.

413
00:41:06,080 --> 00:41:11,870
So the idea is ice freezes out on the sea ice surface, but then the meribel precipitates out as well.

414
00:41:12,590 --> 00:41:21,590
So calcium carbonate actually precipitates out first, but then meribel like and then walks uplifted as aerosol one way or another is missing.

415
00:41:21,590 --> 00:41:23,240
The meribel item therefore is depleted.

416
00:41:23,930 --> 00:41:29,480
And so we think that the negative non sea salt sulphate is a really clear signature of a source coming from the sea ice.

417
00:41:30,800 --> 00:41:37,640
We do know we've been to sea ice surfaces and measured it and it is depleted on the sea ice surface because of this metabolite precipitation.

418
00:41:42,780 --> 00:41:49,940
So what's the source? I'll come to that in a minute. I actually I've only put this slide in because it's such a nice picture.

419
00:41:49,950 --> 00:41:54,479
I shouldn't have covered it up with the writing. But so this actually came from Shackleton's expedition.

420
00:41:54,480 --> 00:42:00,570
And although his ship, he thought was about to sink at this point, he still took the time to take pictures of frost flowers for people like me to use.

421
00:42:01,020 --> 00:42:07,650
These are frost flowers in Antarctica on Antarctic sea ice, which are one of the possible sources of this depleted sea salt.

422
00:42:09,720 --> 00:42:17,730
We're also interested in this depletion because of the impact it has on a process called called the bromine explosion,

423
00:42:17,730 --> 00:42:21,570
which leads to depletion of the ozone in the boundary layer.

424
00:42:21,600 --> 00:42:31,110
But I'm not going to talk about that today. So the idea was that if the sea ice is actually the source of much of the sea salt going into Antarctica,

425
00:42:31,440 --> 00:42:38,640
then maybe the sea salt is actually a proxy for sea ice, but in the opposite sense to the one we would first have thought of.

426
00:42:39,290 --> 00:42:43,650
So when it's cold, there's more sea salt because there's more sea ice.

427
00:42:45,060 --> 00:42:50,760
So now let's explore it. What we're actually doing here in about the next 10 minutes is to do what we

428
00:42:50,760 --> 00:42:55,200
normally do over five or ten years in a proxy development where we have an idea.

429
00:42:55,590 --> 00:42:58,740
We think it's brilliant because it fits everything that we're thinking about.

430
00:42:59,130 --> 00:43:03,780
We then start to think what the mechanisms are and realise it's pretty damn complicated and it probably shouldn't work.

431
00:43:04,020 --> 00:43:06,720
And finally, we come out the other end with something that partly works,

432
00:43:07,140 --> 00:43:10,440
and we're sort of somewhere in between stages two and three at the moment, I would say.

433
00:43:13,510 --> 00:43:17,510
So what is the sauce on the sea ice? Well, it could be. There are four possible sources.

434
00:43:17,530 --> 00:43:23,290
I think the first one is the brine layer on the sea ice surface, but certainly the brine layer on the sea ice surface.

435
00:43:23,650 --> 00:43:32,350
And it certainly is depleted in metabolites, but it's extremely unlikely that just blowing wind across the ice can lift up lots of lots of brine,

436
00:43:32,380 --> 00:43:36,700
lots of this very thin layer of brine into the atmosphere in order to make sea salt aerosol.

437
00:43:37,990 --> 00:43:42,910
So we're going to grade that out because PowerPoint is so clever that it can do that.

438
00:43:45,010 --> 00:43:49,530
Then there are leads within the sea ice phone, so the little bits of open water within the sea ice zone,

439
00:43:49,540 --> 00:43:52,720
but they won't work because they wouldn't be fractionated. They are just seawater.

440
00:43:54,760 --> 00:44:01,180
They've gone. Then there are frost stars, which are what you saw on the photo they form on new thin sea ice.

441
00:44:01,510 --> 00:44:06,670
They're actually ice that grows on the surface and then whips up salt.

442
00:44:09,250 --> 00:44:13,270
I've written a lot of papers in my time about Frost Flower suggesting that they are really important.

443
00:44:14,020 --> 00:44:17,770
I'm not quite sure that they are. I'm not sure they're ubiquitous enough.

444
00:44:18,550 --> 00:44:22,780
And we've also written has done some work which shows that they're not very easy to mobilise with wind.

445
00:44:23,650 --> 00:44:30,430
I'm not sure that works right because I'm not because Antarctic frost that we haven't done that Antarctic frost we've done it on Hudson Bay.

446
00:44:30,430 --> 00:44:36,460
Frost clouds and laboratory frost clouds. And I'm not sure Antarctic frost flowers are the same in terms of how strong they are.

447
00:44:37,270 --> 00:44:38,800
But I don't think they're ubiquitous enough.

448
00:44:41,470 --> 00:44:53,780
So the fourth proposal is what is causing this salt to become aerosol less fractionated salts become aerosol is actually salty snow on the sea ice.

449
00:44:53,780 --> 00:44:58,180
So on the sea ice, the snowpack and that gets mobilised as blowing snow.

450
00:44:58,270 --> 00:45:04,450
So these things that definitely happens there is definitely snow on sea ice and it definitely gets mobilised of blowing snow.

451
00:45:05,200 --> 00:45:12,340
And then in the atmosphere that would sublime, leaving not naked but nearly naked aerosol particles.

452
00:45:13,360 --> 00:45:20,919
And this was suggested by Shin Young in 2008. So the idea now is that this snow on sea ice is salty.

453
00:45:20,920 --> 00:45:28,690
It's salty partly because of flooding, but covers that puts a lot of ocean water onto the top of the sea ice surface.

454
00:45:28,690 --> 00:45:34,280
But within the snow pack, which then gets whipped up, it becomes blowing snow, which the blind skip, see.

455
00:45:34,300 --> 00:45:41,650
So terrorists will be transported to ice cold sites. So what Shenyang did was to make a model which had various parameters in it,

456
00:45:41,650 --> 00:45:51,400
and the most important ones are probably the salinity of the snow that can be blown, the threshold wind speed that mobilises it into blowing snow.

457
00:45:52,360 --> 00:45:56,230
And then the sites break from which which determines the sublimation rate.

458
00:45:56,680 --> 00:46:00,700
Oh, it's one of the things that determines the sublimation rate. And what we've done is two approaches.

459
00:46:00,700 --> 00:46:06,820
One is to make measurements and the other one is to do some modelling using P Tomcat, which is a chemical transport model.

460
00:46:11,230 --> 00:46:18,580
I think you can probably already see from the parameters that it's going to be tricky to make this into a quantitative sea ice proxy.

461
00:46:18,760 --> 00:46:22,900
So I already said this is going to be tricky. Now it seems to work.

462
00:46:23,260 --> 00:46:28,600
But why it works. That's what exactly what I was just saying. Why it why it works now is a bit of a mystery to me.

463
00:46:29,770 --> 00:46:34,580
But we'll explore it a bit anyway. So here's the cruise where we took some data.

464
00:46:34,610 --> 00:46:42,760
My colleague Marcus Frey went on this cruise. It was in the sea ice from here out to here somewhere.

465
00:46:45,530 --> 00:46:53,569
And we had lots of instruments on the ship. We had instruments for measuring the snow particles or the amount of blowing snow,

466
00:46:53,570 --> 00:46:57,850
the size distribution of blowing snow instruments for measuring the aerosol forms.

467
00:46:57,860 --> 00:47:00,980
Afterwards, we had instruments on the ship.

468
00:47:01,260 --> 00:47:08,150
So on the crow's nest of the ship. And then we also had instruments on the ice itself.

469
00:47:08,690 --> 00:47:12,790
And you might say, why do you take a photo when it was dark? It was always dark on this cruise.

470
00:47:12,800 --> 00:47:22,540
So it was winter. There are some interferences when you're doing this kind of field work.

471
00:47:23,150 --> 00:47:27,600
I'm assured that the penguins didn't get anywhere near the samples, but, uh.

472
00:47:32,150 --> 00:47:35,280
So we measured various things under control.

473
00:47:35,360 --> 00:47:40,069
You can so we measured the salinity of the snow that was on the in the snowpack and we

474
00:47:40,070 --> 00:47:47,479
measured the salinity of the snow that was in the air and we measured the snow particles.

475
00:47:47,480 --> 00:47:55,340
So this is a blowing snow event here, starting here, when the wind speed reached a certain point, the wind speed there is in green.

476
00:47:56,170 --> 00:47:59,360
When it reached a certain point, suddenly the snow starts blowing.

477
00:47:59,690 --> 00:48:00,980
And then we measured the aerosol.

478
00:48:05,260 --> 00:48:10,059
So one thing you can already see on here, actually, is that at least in this blowing snow event and it was generally true,

479
00:48:10,060 --> 00:48:13,510
although we didn't have a huge number of blowing snow events, I don't want to overemphasise this.

480
00:48:13,820 --> 00:48:17,660
Sometime after the blowing snow started, we started to see lots of aerosols.

481
00:48:17,680 --> 00:48:22,989
So there is actually little qualitative evidence that the blowing snow does lead to aerosol formation,

482
00:48:22,990 --> 00:48:25,090
which was one of the first order questions for the grant.

483
00:48:27,910 --> 00:48:35,530
And we measured various spectra of the sizes of the blowing snow and the size of the aerosols, which are what are needed to parameterised this model.

484
00:48:38,490 --> 00:48:46,530
So for him, any conclusion from the field or actually the salinities were lower than we expected, which is sort of a problem.

485
00:48:47,400 --> 00:48:53,670
But we think they may have been atypical. So we're going on another cruise in the Arctic soon to try and get some more data.

486
00:48:54,000 --> 00:49:01,440
The threshold wind speed was similar to what we used before, so we got that right and we did see increased aerosol concentrations during blowing snow.

487
00:49:02,130 --> 00:49:06,930
I think I'm running out of time for when you're expecting me to finish, actually, but I think in the next 5 minutes probably.

488
00:49:06,970 --> 00:49:11,400
Okay. So just very quickly, the Tomcat modelling.

489
00:49:11,820 --> 00:49:16,559
What we did was to include an open ocean source similar to what other authors who've

490
00:49:16,560 --> 00:49:22,080
used both higher order and lower order models have used on this blowing snow source.

491
00:49:23,790 --> 00:49:26,210
And it gives good agreement with air.

492
00:49:26,700 --> 00:49:31,680
Once you've changed it a little bit, it gives good agreement with aerosol concentrations at various sites around the world,

493
00:49:32,400 --> 00:49:34,350
including in Antarctica for the present day.

494
00:49:35,250 --> 00:49:44,579
What we found, and this may sound again negative, is that what controls the amounts of sea salt that reached central Antarctica.

495
00:49:44,580 --> 00:49:52,200
This is how much sea salt reach don't see in this model from year to year was not actually the amount of sea ice as we had hoped.

496
00:49:52,800 --> 00:49:57,150
That was the meteorology, the climatology. And actually it wasn't even just the meteorology.

497
00:49:57,150 --> 00:50:02,520
It was very specific. It was the got to make sure we get this right.

498
00:50:04,440 --> 00:50:08,519
It was the blue there. It was the transporting meteorology.

499
00:50:08,520 --> 00:50:11,640
So it wasn't the meteorology that was causing the blowing snow to be lifted up.

500
00:50:11,910 --> 00:50:16,020
It was the meteorology that transported it in. And what we think is going on.

501
00:50:16,320 --> 00:50:23,790
What we did, we discovered that by running the model with a constant year of meteorology unchanged and a different year of sea ice each year,

502
00:50:24,060 --> 00:50:26,820
or a constant year of sea ice and a different meteorology each year.

503
00:50:28,050 --> 00:50:34,920
What we think is going on is that there are only a very small number of events that transport sea salt to Antarctica each year,

504
00:50:35,340 --> 00:50:39,090
probably of the order, half a dozen or so looking into that at the moment.

505
00:50:40,140 --> 00:50:44,460
And of course, that can vary from year to year. If it's for this year and next year, you'll get a big difference.

506
00:50:45,180 --> 00:50:50,190
But it won't be anything to do with sea ice extent. So this property is not going to work from year to year,

507
00:50:51,570 --> 00:50:56,610
but nonetheless that should average out over a period that I think is typically of the order of decade.

508
00:50:57,240 --> 00:51:03,540
And then maybe the sea ice comes in. And so cutting to the chase, what we did was going to jump that.

509
00:51:04,440 --> 00:51:08,879
What we did was to run the model for present day sea ice and last go to maximum sea ice,

510
00:51:08,880 --> 00:51:13,520
but keeping the same climate which you'll instantly see is mad.

511
00:51:13,530 --> 00:51:14,850
But nonetheless, that's what we did.

512
00:51:16,080 --> 00:51:22,320
And it gives you a 50% increase in the amount of sea salt reaching Antarctica, which is actually exactly what you see.

513
00:51:22,590 --> 00:51:25,890
You see a 50% increase, the amount of sea salt reaching don't see.

514
00:51:27,570 --> 00:51:37,469
So the model says that the sea ice increase could due to the sea sea due to the sea ice or the sea salt increase caused by the sea ice.

515
00:51:37,470 --> 00:51:44,310
Change between the Holocene and the last glacial would give us the result we see for the amount of sea salt in Antarctica.

516
00:51:44,880 --> 00:51:48,510
What we haven't done is the converse experiment of what would the changing climate do?

517
00:51:49,080 --> 00:51:54,360
And that's a very difficult experiment to do, because we know that models don't get the last glacial maximum climate in the Southern Ocean.

518
00:51:54,360 --> 00:51:59,490
Right. We know that people think that the winds moved, but the models say they didn't, for example.

519
00:52:00,900 --> 00:52:04,620
So it's not obvious what climate we should use in order to test this.

520
00:52:04,620 --> 00:52:07,680
But nonetheless, we're trying to do some work to do that,

521
00:52:08,070 --> 00:52:15,120
and that will be a future piece of work if you're willing to suspend disbelief and think that qualitatively

522
00:52:15,120 --> 00:52:19,530
this probably is a sea ice property and I really like the way I presented it makes me think probably it isn't.

523
00:52:20,190 --> 00:52:28,259
But nonetheless, if you did, then what this would be saying is that there was less sea ice, there was more sea ice in the last glacial.

524
00:52:28,260 --> 00:52:33,300
I don't think we're very surprised at that. There was less sea ice in the early Holocene and there was less sea ice,

525
00:52:33,510 --> 00:52:38,270
quite a lot less sea ice than we could quantify it with this model in the last interglacial.

526
00:52:38,280 --> 00:52:44,400
And that would be quite important in understanding why the last interglacial behaved as it did in Antarctica,

527
00:52:44,970 --> 00:52:50,520
and especially if we could get ice cores from all around Antarctica to get each sector separately, which is what would be the ultimate aim.

528
00:52:50,940 --> 00:52:58,650
So I'm just exposing to you the workings of how we construct a proxy or possibly deconstructive proxy after several years,

529
00:52:59,790 --> 00:53:02,070
because I think that's the most interesting physics actually.

530
00:53:03,480 --> 00:53:10,980
I'm going to jump both and give the conclusions that we've now got of the second half of the talk.

531
00:53:10,980 --> 00:53:16,500
That is, we've now got a modelling tool with which we can explore the controls on sea salt reaching ice cold sites.

532
00:53:17,530 --> 00:53:21,179
We've we're improving the parameter ization because of field experiments.

533
00:53:21,180 --> 00:53:27,240
It's this interface between field experiments and modelling that I think is the the only way to deal with this kind of property.

534
00:53:28,650 --> 00:53:36,630
We know that it doesn't work for Interannual sea ice because that's controlled by the meteorology and.

535
00:53:37,260 --> 00:53:41,390
We can explain what happened in the last case to maximum, but we don't yet.

536
00:53:41,400 --> 00:53:45,000
We can't yet put our hand on our heart and say this is the unique explanation for it.

537
00:53:45,480 --> 00:53:51,180
We did some modelling that I jumped over which shows that we can define what the region of influence is for each high school site.

538
00:53:52,140 --> 00:53:56,820
It's probably safe to use this under a climate similar for day, which is why emphasise the interglacial.

539
00:53:56,880 --> 00:54:03,180
I think the changing climate between different interglacial is not likely to be so large as to have a strong influence on our results.

540
00:54:04,740 --> 00:54:09,120
So the only other thing I wanted to say because I know conclusions might you think thank God is about to start.

541
00:54:09,120 --> 00:54:16,500
But actually I've got one more. One more thing I'd like to say, which is in the future, we would very much like to go back further than 800,000 years.

542
00:54:16,860 --> 00:54:21,810
And the reason for that is because if you just look at marine sediments climate for the last one and a half million years,

543
00:54:22,830 --> 00:54:32,400
then there's a very obvious change, which is that instead of having these 100,000 year ish cycles here, we had 40,000 year cycles back here.

544
00:54:33,240 --> 00:54:40,440
And these correspond to different components of the Earth's of the astronomical cycles that control the Earth's orbit and its tilt.

545
00:54:41,400 --> 00:54:43,170
And if we've got an ice core that covered this,

546
00:54:43,170 --> 00:54:49,440
we think we could really make progress in understanding what was going on across this transition from these two different types of cycles.

547
00:54:49,780 --> 00:54:53,940
Now, the problem is we don't know where to drill the ice core. We're sure that it's there, but we don't know where.

548
00:54:54,270 --> 00:54:56,639
And so we've got to do a lot of geophysics first to find out.

549
00:54:56,640 --> 00:55:00,540
And that's a really big international effort involving people on four different continents.

550
00:55:02,480 --> 00:55:04,100
And with that, I will stop.