[No audio until 0:55 (mishap during recording)]

...based oceanographic transects, right? So these are called repeat oceanographic transects, what we call hydrographic sections. They come from the World Ocean Circulation Experiment in the 1990s, the CLIVAR CO2 program in the 2000s, and now from GO-SHIP, the Global Ocean Ship-Based Hydrographic Investigation Program in the 2010s. And so this is some work that Sarah Purkey did. It's actually an update of it, she first did it in 2010 when she was a PhD student working with me, and we decided to use all these spooky hydrographic sections and just look at the temperature trends along the sections. So they're sampled like once a decade, which isn't very often, but we did see some consistent pattern. We see warming in the below 4000 meters, and you see in this plot, it's strongest around the Antarctica. It grows a little weaker as you go to the South Pacific. The North Pacific, you can't really see it on the projector, but that is a very light shade of yellow. It's actually all warming up there just very slowly. It's warming in the eastern Indian Ocean and the Western Atlantic. The stipple regions are where we don't have enough data to predict a trend with any confidence.
We can't distinguish the trend significantly from zero, right? So there's a lot of places like that with these for peak hydrographic sections. So this is what I'll talk about a little later. So this is sort of a decadal trend. It arises from a major reduction in the formation of Antarctic bottom water, the coldest densest water to the planet that happens around Antarctica. I'll talk a little bit more about why that is at the end of the talk, but the uncertainties are about 50% of the signal. So this is about 8% of the global energy imbalance, but 54%, it could be 12%, right? It's not, it's actually fairly uncertain. And it's also always about a decade out of date, because these are based on repeat hydrographic sections. This is a big international effort, but people only get to these sections once every 10 years. So even if you update it every year, you know, it's 10 years out of date, basically. So we're not really monitoring this interesting and changing and important situation in real time. So this is just one transect. And this is actually, we looked at this in, I first looked at this in about 2005, I think, just after the second transect was completed. And this really started me on my journey of looking at deep and abyssal warming. I have a confession to make, back in like 1999, some colleagues published a study in general science that showed this very small warming, was about five millidegrees over a couple decades and I got called by Richard Harris at NPR, the science reporter at the time, and he said, do you think this is significant? I was like, ah, I don't know. It's pretty small. You know, the measurement error is about two millidegrees, so this is five millidegrees. It's just over the measurement error. I don't know. I haven't seen it, but this section made me a believer, after that. So we looked at this, this is a section of mean temperatures are the black lines. The trends are in colors. The scale is off to the right, again, it's sort of plus or minus what, eight millidegrees C a year. But what you can here and this goes from the Scotia Sea at 60 S the equator at 60 N it goes from the surface down to 6,000 meters close to the bottom of the ocean. So this crosses the Scotia Sea, the Argentine Basin, and the Brazil Basin. And so there's a few things that we've got to do. These are contorted thick lines at 1 degree [indistinct]. This is a thick line down at the bottom that's sort of at 3,000 meters in the Argentine Basin and then deepens down to 4,000 meters and then stays here. Is the, I think that's the zero degree C contour, this is the one degree C contour and the two degrees. You can see this large gradient between one and two degrees that's down at about 4,000 meters. That's actually a contrast between the cold Antarctic bottom water, which is warmer than in the south and going north in this, in the west, in the Brazil Basin and the warmer North Atlantic Deep Water, which is warm up in the north and going south. So these things are sort of moving in opposite directions. There's a big strong thermal line between them. But you can see everywhere below that strong thermal line is warm, right? You see this really strong warming signal down in the Brazil Basin. It's a couple millidegrees C per year on average. Same with the Argentine Basin. And in the Scotia Sea, you also see warming. Actually at mid-depth in the southern ocean in this section. So this is from pre-occupation. They were done in 1989, 2005 to 2014. So we'll look back and revisit it in 2014. So what the Argentine Basin the state farming rate is 1.9 ±2.2, right? There are these eddies. These are basically the vertical strengths and the temperature are basically mesoscale eddies. Those have a scale of about 170 kilometers on average in the ocean. And they're the primary noise choice source for the uncertainty. And so this short section of Argentine Basin actually doesn't give us statistically significant warming assessment. The Brazil Basin were a little better. If you go back and look at that map, that we showed previously. You can see the Argentine Basin is actually what we were just looking at. The Argentine Basin, the stipple, the Brazil Basin is actually statistically significant. All right. But we're going to go on and try to quantify these using a Deep SOLO flow data as well as hydrographic data and see if we can do a little better than ±50%. So this is in terms of trends. This is a Deep SOLO float. This was developed by the Instrument Development Group at Scripps. The PMEL units that we deploy are built by MRV systems. Scripps builds and deploys their own. They are capable of 200 plus full depth CTD profiles over five plus years lifetime. In fact, the first one that a [indistinct] deployed back in 2018 has now done 250 deep profiles, and it's still going strong. So these are robust and actually capable instruments. They are the primary pressure cases a 13 inch diameter set of glass mated glass hemispheres. So that's amazing that something can cycle you know down to 6,000 meters depth which is basically what 600 atmospheres of pressure, right, and back. But the glass is pretty strong that spherical straight shape of course is best for compression. And it's actually closer to the compressibility of seawater that will be aluminum cylinders and that makes them more efficient to run up and down because you don't have to change the volume of water. So they have an internal oil lab here, they have an internal oil reservoir here, a high pressure pump, batteries to power it, as well as the electronics and the power communications. A CTD that measures temperature, salinity, and actually temperature productivity and pressure for once you get some quality, density, and depth, all kinds of things. And then they basically cycle up and down by pumping oil from the internal reservoir to the external reservoir if they want to rise because that increases their volume, but their mass stays the same, so their density is lower. And by allowing them to go back into the internal reservoir, they want to sink. So pretty cool. I love this technology. Here's how many we have had out there as of this spring. There were about 200 floats in regional pilot arrays. About 50 of these are PMEL's, a bunch are Scripps'. We also have international collaborators in the UK, Australia, Japan, Europe, France, a whole number of countries are putting out the floats. I'm gonna look at, we're gonna look at Antarctic bottom water warming in a few of these regions. So the South Pacific, the South [indistinct] Basin here. We're gonna go back to the Argentine Basin first. We'll look at the Brazil Basin. So these are all places where we have arrays and we can see using the Deep Argo data, know if we can detect anything more than we were able to with the repeat hydrographic data. Right? So this is actually the location of that repeat hydrographic section from the Argentine Basin that I was saying gave us 1.9 ±2.2 millidegrees C in here, we're at below about 4,000 millibars off to the side. But we actually have a whole bunch more data. So what I can do is I took Deep Argo data, this is just from about a year and a half of deployments, and those are the pluses. I took all the historical support data that was available from 1972 to 1998. So that's through the World Ocean Circulation Experiment, and those are the x's. And I compared nearby historical data, the orange x's, to nearby Argo profiles which are the magenta pluses. And I just said if they're within 165 kilometers of each other, I'm going to compute a trend, right? So a whole bunch of trends and then we can look at the statistics of them and average them in a basic wide sense. You do that, this is what you get. So we're looking at the trend versus depth from 2,000 meters down to the bottom and this is the 5 to 95% confidence intervals and I basically said that any any trend that has a time difference that's, well, basically in 165 kilometer by 165 kilometer by 60-day bin and I use 60 days because that's an eddy time scale, I use 165 kilometers because that's a dominant eddy time scale. This statistically is independent. So you get a warming trend that's very similar to the warming trend that we've got using the repeat hydrographic section, about two millidegrees C here. But the uncertainties are much much lower. They are a factor of 10 lower. Now we're using a lot more data. Part of that because we can use a lot of historical data. Part of this because these are both zip around this basin like crazy. It's very energetic but the end result is that we're getting you know, we're getting that instead of a 2 ±2 value we're getting 2 ±0.2.
So we've increased the certainties 10-fold. The other thing is if you think about I talked about this reduction in Antarctic bottom water formation. A typical definition of pure Antarctic bottom water is it's below, colder than 0 degree C. It's according to a deepening of the 0 degree ice to sea isotherm by about 200 decibars over this time scale, 34.5 degrees for the meantime of the temperature trend that we're computing. So big change in depth of the isotherm wave reduction in our bottom water. You can do a little better than just looking at the basin-wide trend though. So this is where sort of improvements come in. We can see if we plot the transverse latitude. The overall trend, this is at 5300 decibars, so this is just one pressure level, a very deep one, that's down in the Antarctic bottom water. The trend is two millidegrees on average. I just showed you that  in

last plot. But there is some spatial variability. See, as you get down towards 49 S here, the trend tends to be a little higher, actually quite a bit higher at the southern edge of the Argentine Basin. And this pattern actually holds for three decades. If the Arnold flows are all from, what, what'd I say, 2021 to 2022? But if we look at the 1970s, which are the red pluses, the blue pluses, the x's, black x's from the 1980s or the 1990s. We see this higher values near the southern boundary persistent across all those decades. So this is kind of an interesting, you know, it's not, we're already getting beyond basin averages here, right? And learning some new stuff. So if you look at the warming pattern and you look at the Deep Argo temperatures, they are at -0.01 in the interior of the basin and a little colder, 
-0.015 and maybe -0.017 near about 50 S of the southern part of the basin. If you look at all the historical data together, you know there's going to be some scatter through it because it didn't work over time, but if you look at all of it together you can see that it's [indistinct]. So it's
warming much faster here. And that's because the coldest,
densest water that are entering the basin, the ones that are closest to the source that are warming the fastest. So that's consistent with what we might expect. You've got this cold, dense water entering and it's warming fastest and then as it mixes into the interior, that forming is diluted a little bit. And if we look at another section now, we're going to look at the A23 section from the World Ocean Circulation Experiment, just to sort of show you, to plan you what that looks like. [indistinct] But this is probably about six kilometers depth at the bottom, and you can see, and it's going from 50 S, the left side of the 30 S, right? So I would say that near 50 S, this cold, dense water here is warming, right? And this is actually a signal of the Deep Western Boundary Current entering this basin. It enters from the southeast third force of the basin flows west along this topography at the south, this ridge at the south, at about 50 S, and then work along the boundary. So just a deep shear, the coldest water, is actually warming up much faster. It's this temperature gradient that actually supports that northward bottom intensified northward flow. So the bottom intensified northward flow is actually weakening over time, consistent with the shutdown of Antarctic bottom water circulation. You can assess the amplitude of that and if you estimate the transport below 4500 decibar assuming zero velocity there, it drops from 10 sieverts to about 9 sieverts is roughly consistent with a 200 decibar isotherm drop over three and a half years that I talked about earlier. So we actually have less water coming into the basin, the amount of Antarctic bottom water is shrinking as a result.
So the overall spatial pattern is actually consistent with what we're seeing from the trend at the basin. Alright, stepping on to the Brazil Basin, we see a similar result. The Brazil Basin floats don't move around as much. It's wider up there. So these are, again, the x's are [indistinct] sections from the 1990s, the pluses in magenta and blue are Deep Argo float data, and we're just comparing the magenta with the yellow. So magenta floats with the yellow or the orange hydrographic data, and you see about a two millidegrees C trend, and its uncertainties are actually a little bit bigger 
than the ones in the Argentine Basin even though we have more float data here and that's because the floats don't move around as much so we don't get as many independent estimates, right, they're not all spatially but we still have a much better estimate of the warming rates than we did with repeat hydrographic section. Now I'm going to move to the southwest Pacific Basin and we can do a couple of different calculations here. There are a lot of floats. This was actually the first place where Argo floats were deployed. They were deployed back, the first two were deployed in 2014 by Scripps. They deployed the first real part of the array in around 2016, and have been deploying floats ever since. So we compare those, Deep Argo [indistinct] CTD data, this is 
just the Deep Argo data. You can see a very clear warming signal again in the abyssal warming. In the abyssal waters below four kilometers, a little bit above a millidegrees C here. You can look at the trends for the float data alone. So that's what the right-hand side figure is. I've been showing you long-term trends, right? Just decadal average trends up until now. This is actually a trend from 2014 to 2023. So it has a little bit bigger uncertainty than the long-term trend. But in the bottom water, the trend is also bigger, right? It's reaching up at around three. So again, we're always saying maybe a little bit of acceleration here, right That's the first time I've talked about that for the bottom water. But here it is. We are perhaps seeing that in our little tour. I'm going to skip around to another ocean basin. I'm going to look now at the Bellingshausen Basin. So this is some work that I did remotely with a summer student fellow at Woods Hole and Allison McDonald here at the mood. So we looked at trends. We had section data here. And we said, well, we have data from the 1990s. We have data from 2005 to 2011, data from 2017 to 2018 and then we have the float data, which was 2023 and '24, right? So they're color-coded on this map, from blue to orange to red to black by year. And if we, again, look at nearby float data compared to historic data from those three different epochs, from the '90s, mid-2000s, then the late 2010s, what do we see in terms of trends? So here, the trend is close to 3 millidegrees seen per year in the '90s. That's bigger than we've seen in the Argentine or the Brazil or the South Pacific Basin. That makes sense. This is closer to the Antarctic bottom water source. If you remember that original plot of the long-term trends from Sarah's work I showed you, you know, it's larger than the sources. It gets closer to 4 millidegrees C as you look from 2005 to 2023, and then if you look at 2017 to 2018, the 2003 is off the top of each of the degrees C a year. So again, like the Southwest Pacific Basin, we're starting to see some evidence for acceleration of this warming. This is substantial warming, and actually if you look at the last one, that's half the 2020 global average SST warming rate of 17 millidegrees C a year. So this is really actually a pretty big signal, and it's not, it's over a couple kilometers, and that's why when you go to the aggregate, it amounts to about 8% of the global signal. Right. It's, in places, it's quite large and it's persistent [indistinct]. All right. So.... Or this is like hot off the presses haven't even submitted yet. But I'm going to talk about it anyways. So this is... Sarah and I talked about well, why don't we, why don't we try to combine the repeat hydrographic section data, you know. But not just the repeats, but all the historical CTD data that we have, with the Deep Argo data. And try to estimate some trends. So these are the results. This one on the left is for the Abyssal layer. That's 4,000 to 6,000 kilometers. Again, these are temperature trends. Max we got is 
±5 millidegrees C a year. Both of negative being blue, positive being red, and this is a Deep layer, the average trends in the Deep layer from 2,000 to 4,000 decibars. So what I've done is I've made all available historical ship-based CTD data, the Deep Argo data, at 2,100 decibars, that's 66,000 historical data and 30,000 Deep. I'm using it from everything that's available and flagged as good from 1970 to May 2024. I decided to bin these in 167 by 167 kilometer bins, no matter how many data I've had in that bin, and just compute a trend. So long as the first point and the last point were separated by more than 10 years. And we're doing that at 10-dbar intervals from the surface to 6,000 decibars. That results in using about 70,731 profiles in 3,338 bins spatially after discarding some extreme outliers. There is junk in the World Ocean database. I mean, honest to God, [indistinct], you have to like, actually [indistinct]. But that goes, and again, that's at 2100 decibars, and all these counts that I'm talking about are defined as you go deeper. There's fewer and fewer data deeper. But the ocean area also shrinks [indistinct]. And so you'll see in a minute that the average is up to about 0.5% of the ocean [indistinct]. So I then had trend estimates that are colored in all the little squares, inside all those little squares that are on there. So the left-hand side is any square that had a value somewhere between 2,000-4,000 decibars included. Sorry, left-hand side is 4,000-6,000 decibars. Right-hand side is 2,000-4,000. And then I used just MATLAB natural labor interpolation to map data from where I had the samples to unsampled areas that I could do a full integral and that's basically a local weighted average. So it's just a it's just a way to smooth it takes the nearest nearest neighbors within some Moroni Hull and averages them by area. So it's a little bit complicated to explain but it's basically a weighted average, ok? So we can then integrate these rates globally with the 4,000 decibars of Abyssal layers and I'll do that in a little bit and I'll actually spend a little more time on these plots too. So let's move on. Just so you can see this is again at 2,100 decibars on the left-hand side. It's just sorry about the lost plot.This is the best way I could think of doing it. Most of the values, this is profile cap. So in those 3,000 bins, you might have as few as two profiles. 120 of the bins of those 3,000 bins only have two profiles. They're separated by 10 years of power, but if you're a trend, the mode here is six profiles within a bin. There's about 230 of those at 2,100 decibars. And then there are some bins that have hundreds and hundreds. There's actually one bin off of Hawaii at the Hawaiian Ocean time series that has over 500 profiles. So that's a combination of the Hawaiian Ocean time series and we put two Deep Argo floats
there. So we have a whole bunch of profiles. But on average I should know how many there are on average [indistinct]. You can see the distribution. And then on the right, this is basically all the bins looking at what the year of the first profile is in blue and the year of the last profile is in red. All right. So you can see sometimes, I was actually computing a trend from 1970 to like 1980, right, or from, you know, but for the most part, I'm on average computing a trend from about 2015 or '16 to I think, 1987. The answer here is 1987, right? So this is a long term sort of trend.

So with those 167 by 167
kilometer bins that are sort of, you know, on the eddy scale what percent of the ocean volume of my sampling versus pressure, and it starts at about 30% to 2,000 decibars, gets to about 29% at 2,100 decibars, and decreases a little bit toward the bottom. On average, if you look at what the volume-averaged sampling rate flow of 2,000 decibars is, it's about 27% of the ocean volume. And if you look on the right-hand side here, you can see the North Atlantic, what sample? We sampled, you know, almost everywhere in the Northern, North Atlantic, up in the Greenland Islands, you can see the, and that's because historically this is just super well sampled by ships. Same for the North Pacific, around 10, US, even in the deep ocean it's pretty well sampled by research vessels. You can actually see people took CTDs on the TAO stations back through the 1990s. You can see those. That's a PMEL work there. And then in the south, the Atlantic is actually pretty well sampled, too. That's a combination of good historical sampling by ships combined with the Deep Argo array, right? So in the Pacific, even though there's a lot of Deep Argo data, the array is not 10 decades old yet, or a decade old yet. So I don't have

a lot of [indistinct], you  [garbled]

I need another year of shipwork data to
make an estimate. So there, it's only a slight

know, it's not as well, you know, there's a lot of problems. There's other decades, though, five or six years, I'll have a decade's worth of Deep
Argo data and start to calculate [indistinct]. We could probably try to find trends now, actually, we just haven't done it. So then, as I said, North Pacific, North Pacific are pretty well sampled. Southwest Atlantic is also good. Southwest Pacific is still not sampled, everywhere else is still a little under sampled, right? But that's why we want a global Deep Argo array. Let's just take a look a little bit more in a little more detail to deep warming rates. I know those other figures were small, so this one's a little bigger. So this again is from 2000 to 4000 decibars. And there's a few patterns that leap out at me. One is that south of about 50, that's 50 S, the Southern Ocean is warming in that layer, right? There's a little bit of high amplitude noise just to the north of that. That's because the Antarctic Circle flow of current, which has a big contrast in temperature between the warm waters to the north and the south waters, the cold waters to the south, is gandering around there. And so, you know, we might be aliasing some of that, that's a noise source to us, right? North of that, you see kind of blobiness in Pacific and the Atlantic, or [indistinct] at least I do.
These are very old waters that's been it's between two and four kilometers water in the North Pacific and North Indian. It's been for millennia since they last saw the ocean so we don't necessarily expect a big warming trend or blue trend there. And then you go up into the North Atlantic, you can see a lot of warming up at the GIN Seas, north and east of Greenland, that's because the deep convection there has actually shut down and that those waters are warming as actually warmer waters are to flow back into that place the waters that used to be cool by deep convection. Similarly, it looks like in the Labrador Sea, we've had a lot of warming between since like late '90s to the 2000s and you can see that up in the western sub-polar North Atlantic there's some cooling and the rest of the subtropical western North Atlantic and the eastern sub-polar I'm not sure what that is. It could probably again be related to decadal variability in North Atlantic deepwater formation. But that probably could use some more work to find out. If you look at the Abyssal warming rates, you can see again warming that's sometimes up to five millidegrees C a year. In the Southern
Ocean here, all the basins of Jason Antarctica.

You can see this cool pattern of that
warming, uh, spreading up 
into the Pacific diminishing as it goes. Most of the inner bottom water flows up along the western boundaries here, uh off east or off east of New Zealand at the top of the that ridge that part is warming the most, right? The interior part of that base is not warming much again that's consistent with the coldest densest waters actually warming and that reduces the that's consistent with a reduction the northward flow of the pan on the bottom water in that basin. Just like I talked about in the Argentine Basin in some detail. You can see warming extending into the eastern Indian Oceans Basin, the Maroon-Chile Basin, the Bauer Basin in the eastern Pacific. You can see warming extending into the western Atlantic, the Argentine Basin, the [indistinct] Basin, and actually even into the North Atlantic along the western flank of the Mid-Atlantic Ridge there. The only places where you really see cooling are in the rest of the North Atlantic and in Angola Basin in the eastern part of the South Atlantic. Those are actually basins that are not primarily fed by Antarctic bottom water formation. So the bottom water there is actually of North Atlantic origin. So it's North Atlantic and [distinct]. So we have a pretty consistent pattern of warming anywhere that there's Antarctic bottom water. And I said I would talk a little bit more about, let's look at the global averages first. So this is just, I'm just probably a temperature trend. Remember I said that the bottom water was warming sort of, the deep water was warming south of 50 S, so I just plotted these averages north and south of 50 degrees south. So the blue line is the warming rates south of 50 degrees south, the average is about 2 millidegrees C in here. In the bottom water, 4,000 meters to the bottom, and then a little less in the deep water. The rest of the global ocean, north of 50 S, you see a warming of maybe half a millidegree C a year in the bottom water on average. Some places that are higher than that, some places that are lower, but the average is about half a millidegree.
And then you see less warming, especially between about 3,100 and 3,500 decibars of artifacts. The really old waters in the Pacific and Indian Oceans that are not warming at present. And then you can see this trend start to increase. So that's going to meet pretty well with the trend from Argo, the long-term trend from Argo that I showed you at the beginning of the talk was still not zero, was above zero at 2,000 decibars. So we sort of met that. Um...

So...it's interesting. We proposed,
Sarah and I, back in 
the early 2010s that this warming is probably due to a reduction in the Antarctic bottom water rate, formation rate. There were already papers out that were proposing that there was freshening on the shelves of the Antarctica. So basically, bottom water is formed around Antarctica as very cold winds from continent blow across polynyas, holes in the ice, that extracts heat as ice forms that makes the water saltier. So making water colder and saltier, both of those things make the water denser. That dense water spills off the shelves into the Abyss and then spreads north in all the basins that we've been looking, right?
Everything but the North Atlantic and the Eastern Southwest. If you melt marine terminating ice sheets and inject fresh water above the shelves, that has to inhibit that very dense formation and less bottom water forms. This is a figure of a lab from Fogwill et al. 2015 where they did a numerical modeling experiment where they enhanced melt water flux around the Antarctica. And so the black lines are control runs and the colored lines here are various enhanced fresh water melts. They put fresh water to the surface and you can see that this is so great of Antarctic bottom water page. About seven meters a second on average of the control runs. As soon as you introduce that meltwater, bang, the Antarctic bottom water formation is drastically reduced, right? And so that's consistent with what we're seeing. If you look at temperature changes on the top panel, you can see this warming around Antarctica, south of about 50, 60 S, and then in the bottom water. More subtly, smaller, but still there, extending north. And so this is just a map from another recent paper by Li et al. that looked at a similar problem where they had wind, they looked at both wind changes, thermal warming, and meltwater. And the meltwater is really the dominant factor here is what this key context in all the accumulation map is. But I just wanted to show you the pattern that they show here for warming in the events from 1991 to 2023 and 2030, you see some resemblance to that in our observed plot that I showed you. So the Abyssal recurring rates are actually expected to accelerate for model results. So deep warming, there's a part for sea level rise, there's energy imbalance, carbon update, global ocean circulation, etc. So I'm just going to make a pitch that we should be monitoring. So this is Deep Argo in January 2024, color-coded by nation, you can see there are 11 nations contributing floats. There are about 200, but we're really far from a global array. What OneArgo is proposing, oh, and this is just the profile of patients. So these are all the Deep Argo floats that I talked about as of January. This is our target, 1,228 floats global, so we're at sort of about 15% of the target. But if we do this, we can monitor what is clearly a changing deep and abyssal ocean. So conclusions, ocean warming is 90% of global warming. Ocean warming really is the global warming signal. Sorry, I did this slide, like, a few minutes before the talk, sorry I [indistinct]. Argo has allowed us to better quantify the warming and its acceleration.

Oh god! [laughter]
That is awful!

Sorry, I never do that.
[Indistinct] Never. Yeah.

Um. Auto... damn, autocorrect! [laughter]
Repeat hydrographic sections have detected global abyssal warming. Deep Argo regional pilot arrays, and I've shown you results from three of those that improved our regional abyssal warming estimates, sometimes as much as tenfold. And in a couple places, detected acceleration of that warming.

Okay.
And finally, I should do global analysis of ship-based and Deep Argo CTD data. That actually improved the spatial resolution before we were just doing based on averages. Now we're doing maps. And reduce uncertainties by about half. They've gone from ±50% to about ±25%. We got, by the way, the exact same answer that Sarah got back in 2010. But like I said, we know better what the patterns are, and we know, you know, instead of 50% of the survey we're down to 25% already. So this warming is caused by melting Antarctic Ice Sheets, and is expected to accelerate in the coming decades. Global implementation, as we already know, will allow us to better understand and quantify this phenomenon. Thank you.

[Applause]

All right, I put everyone to sleep.
I see a few hands.

John, I think your hand was up first.

[John] Yeah. [inaudible]

[Greg] So there is a freshening signal that I didn't talk about today.

55 minutes already. [laughs]
Right, but there is there is a freshening signal that is associated with that warming, especially close to the source around Antarctica.

People have done that with the peak hydrographic section data.

We did actually look at that using the Argo data in the Bellingshausen Basin recently, and we think we can see that freshening as well with the Argo data.

It's not easy because the freshening signal is pretty small, but we think we can do it.

I think Chidong has his hand up next.
[Chidong] Yeah, I think that is the thought. So.

The warming is caused by melting, I understand. Melting is [indistinct] got a sinking motion.

But on the other hand, I'm thinking if the abyssal warming is connected to the global warming, [indistinct] and before we should start the surveys, right?
And so my thinking is, on water, maybe [garbled]

larger, the surface water is [garbled] about just completely opposite from what you think.

I don't know where I'm at.
[Greg] So it's actually a little bit complicated to explain, but basically if you think about the abyssal temperature budget in a steady state, it's a balance between inflow of cold abyssal water that's formed around Antarctica, and mixing of heat from above as that water upwells, right? 
And so that's what, it's that combination of mixing from above and cold water coming in, being constantly renewed from below.

And that's been going on for thousands of years, right?

If you slow down the flow, it's gonna warm down there, basically.

So just think of it like a pipe model.
The other way to think of it, people also ask, how does the signal get all the way up into the North Pacific, right?

So quickly, right? The effective
timescales are hundreds of years.

Well, the answer is is that it's not. It's near the source, it's partly an advective signal, but everywhere it's also a heave signal.
It's communicated by planetary waves, which have a much faster speed than the advected time scale.
So basically you reduce the amount of Antarctic bottom water coming in in the basin.
You set up a Kelvin wave that propagates along the weft.
The West Coast gets the equator, propagates east as a Kelvin wave, goes forward as velocity waves and radiates into the chair.

That takes about 40 years, that signal takes about 40 years to get up to the North Pacific,

and the freshening has actually been happening in Antarctica since the 1960s.

So it's all fairly consistent.
If you don't like thinking about frosty and Kelvin waves, I also think about it in terms of a balloon analogy, where imagine you're just, you have a balloon that has tiny holes in it.
Those holes are the mixing, right? You're blowing into the balloon at a constant rate.
The balloon gets bigger and you get basically the air that's going out of those tiny little holes is equivalent to the air that you're putting into the balloon, right?
It takes a while for that air from your lungs to get to the edge of the balloon and escape.
But if you slow down the blowing, the wall of the balloon shrinks instantly.
What's bringing that?

Pressure waves, right?
It's pressure waves in this case.

They've traveled with the speed of sound, but those are what communicates the signal of that change in the inflow to the wall of the balloon, right?

It's similar to the Kelvin waves communicating the signal forward.

That was way more than you asked for,
but I like to sort of try to tell that story.  
[Trish] All right, how about one more? 
[Greg] Yeah, is there anyone online?

[Sandra] Uh...No, no questions online yet.

[Man] I'm curious, what would it take to get your 1,300 Deep Argo floats, like what is the number?

[Greg] Budgetary?
[Man] Like a budgetary number, like what would that cost?

How would you pull that off?

[Greg] So it's basically, Argo right now is about $12 million a year that the U.S. puts in.

We want to actually go deep, we want to expand into the polls, and we want to do biogeochemical.
So I'm going to give you a number for all of that.
The deep probably doubles the cost.

Biogeochemical more than doubles it.  
End number is about 50 million a year from the U.S.

That gets you full depth coverage, you know, temperature, salinity, and five or six biogeochemical variables. 1,000 of those floats. 1,200 floats going to deep.

Pure Core floats, you know, you've added a lot of heat than HEC.

But yeah, the overall price increase is roughly tripling their electrical efficiency.

[Trish] All right, let's thank Greg again.
[Greg] Thank you.

[Applause]