 OK. I'm going to start again. So I'm switching away from DO events and go to the glacial interglacial cycle. But do you have any questions on the DO events? So there seem to be two mechanisms, right? One is related to CIs and they can be trans important. I'm going to do CIs for one, and one is more about the shutoff of the fresh water input and shutting down the order of simulation. Yes. They're somehow related to each other. Do you know what the leading idea is? Like if you need both or? Do you need? Exactly. But I guess you can have them in the end. So if you look at the literature, people, definitely the AMOC is the leading. It's the leading idea in the literature. If you look at those two examples I showed at the end, the Duncan paper and the Vittorietti and Peltier. In the Duncan paper, you could invoke a small change in the AMOC, but it's not fundamental to the mechanism. In the Peltier, there is a big change in the AMOC. But again, it's not necessarily a building element. It may be more a symptom of what's going on rather than a, and now I think the debate is very much open. They are not orthogonal. That's one point to make. They are not exclusive. I don't think at this point we have all the data for people making their mind. I understood your message. The rapid warming, as a result of the shutdown of the AMOC, that was a suggestion, yes, the analogy was that the shutdown of the AMOC might lead to the rapid warming at the end of the events. We talked about the asymmetry, and you were making the analogy, and you were saying that is that the right confusion? The AMOC shutdown, then, because you have the asymmetry, instead of where it shuts down? Oh, you mean in the more classic way people think about it? Yeah, when you got the end there, you were talking about the dimensions, instead of when you go from cooling or warning. OK, this one doesn't really need the AMOC to change in strength. Yeah, the asymmetry comes from the depths of the ocean, which is involved into the transition. So in one case, you have large intense convection, which makes the effective heat capacity in contact with the atmosphere to be large. And the other case, you don't. But in both cases, you have an AMOC, which is substantial. It's not collapsed. Well, you can still have some water mass transformation. Yes, yeah, it's weaker. But maybe, I don't know if, Brian, you have the numbers in mind? It wasn't a clear case of one is there, and the other one, there is not an AMOC. In both cases, you had. In the literature, there are two main mechanisms that have been involved that support a multiple equilibrium. One is the threshold of flux idea of Stommel, which Broca latched on to. And Stommel and Broca are such hugely influential scientists that that's been the predominant paradigm, actually, in the paleo literature. But then there's an older piece of work used to decode and sell us, which is the ice albedo feedback, which was going on in the 60s. And that is the dominant mechanism that we are finding in these models. In those, yeah. The AMOC is, so that's the switch. The AMOC is responding, and playing a role in the time scales, but it's not the trigger of the, it's not the mechanism. No, it's not the mechanism. You don't need a shutdown. But I would say that in the literature, there are very few examples. I'm not even sure I can find an example, except for those two recent papers where these sea ice albedo feedbacks, at least the coupling between a big ice cap and the ocean, has been invoked for due events. I think it's a fairly recent idea. Any other questions? Glacial, interglacial cycle? OK, so now I've shown you two examples of a multiple state we had in a very idealized configuration, which had only, well, no continents or just a very narrow continent. And that's an example of a simulation which has, in this case, two big continents. So the continents are now 40 degrees wide. They extend over the full depths of the ocean, so those yellow patches. We call that boomerang, because it sort of look like a boomerang if you look at it from the North Pole. But we haven't found a really six year name than that. If you have any suggestions, you'll get credited for that. You won't get rich, but. And so we have a small basin and a large basin that looks like an Atlantic and a Pacific. And there is a southern ocean, which is zonally re-entrant. So there is an ACC going around. There is no Antarctica. And something really interesting in that simulation is that you have an amok, meaning that there is an overturning circulation in the small basin, but there is none in the large basin. And we've analyzed why this is the case. It has to do with the precipitation pattern. I'm not going to go there here. The important point is that that system has some very interesting similarities with the present day, the climate, as we know it. Just a question, do you have rivers? Yes. Well, runoff. If it rains over the ocean, over the land, it's carried to the ocean. Yes, yeah. And there is overland. You have a very simple scheme. So if it snows, it builds up a snow cover. And there is a water storage. And when the buckets are full, the water runs out to the ocean. It's very simple. So the albedo is prescribed. Except when there is snow that falls on top, then it's the snow of the albedo that's the sensitivity of it. Oh, actually, that's how we found those multiple states is playing with the albedo. Yeah. So the system is sensitive to that, actually, quite a lot. So we get two states. One is a cold state and one is a warm state. I haven't tried the snowball because in that context, it's not really interesting. What you see is that the warm state has an ice cap to the south pole. It's quite warm. There is very little sea ice at the north pole. And then you have a cold state which has two big ice caps at the two poles. And again, they are stable. They exist for the same forcing, same parameter. They are just two realizations of the same dynamical system. You can, for example, the sea surface temperature between globally average between the two simulation differs by 10 degrees, the SST by 8 degrees. And I'm going to come back to the atmospheric PCO2 in a minute. So why do we have those stable states? It's the same story as before. If you look at the ocean, it transports in the warm and the cold state. So the warm is in red and the cold state is in blue. You get the same picture that there's lots of heat being pushed away from the equator into the mid-latitude. So it released to the atmosphere around 40, 50 degree of latitude. And so if you look at the North Atlantic, the Northern basin, you have a state with very little sea ice and then a state which extends all the way at the point where you have maximum convergence of heat, which is released to the atmosphere and baseball to stop the progression of the sea ice. If you look at the southern hemisphere, it doesn't change much between the two states. So we think that the multiple state in that case is emerging from the northern hemisphere. And that's quite interesting because if you compare to the previous case where we had a symmetry between the two hemispheres, we have a system which is now not symmetrying between the two hemispheres. And yet we still get multiple states. It's actually not zonally symmetric either. So the stable state can survive a slightly more complicated geometry than almost an aqua planet, which is interesting. So a lot of what's going on here is I would like to make the case that those two states look a lot like the present climate and the last glacial maximum. So the last glacial maximum is the state about 21,000 years ago where we get the coldest of the glacial state. Just before we left the last glacial maximum and drifted into the warm climate we know now. So this is a plot showing the overturning circulation in those two states. That's the red curves. And the arrows gives you an indication of the direction of the flow. And the shading in the background is showing the temperature. So not surprisingly, in the cold state, the most of the ocean is near freezing minus 0.5 minus 1 degree. In the warm state, you get actually a quite warm deep ocean, which is 8 degrees. And you can recognize the North Atlantic cell here, which is carrying heat between the southern and the northern hemisphere, which associated the deep water formation in the small basin, upwelling into the southern ocean. There's a bit of Antarctic bottom water. It's quite weak in the warm state. And as you switch to the cold state, you see that that cell is decreasing in intensity. And this cell is increasing in intensity. There are some interesting points here. If you look at the bottom water and the cold state, it's near freezing. And we have observation that at the last glacial maximum, deep water, bottom water, were near the freezing point. The bottom water is saltier in the cold state than in warm state. And that's simply because in the cold state, that cell is driven by a brine rejection due to the expansion of the sea ice. There is a bit more sea ice in the southern hemisphere. And the growth of the sea ice means that there is more brine rejection and they get slightly saltier water being produced, more and slightly saltier water. There's an interesting point here. You can't really see it well on that plot. But if you look at how much water is coming up into the southern ocean up to the surface, it's actually not changing between the two simulations. The total amount rising to the surface is about the same. What's happening is that in this state, most of the water rising to the surface is being warmed up. It's gaining buoyancy. And in a steady state, if you gain buoyancy, the water are transforming into lighter water and moving a greater world. In that state, most of the water coming up to the surface is rising in a place where it's losing buoyancy because it's now near the ice edge. And so it's transformed into denser water and turning into southward into the Antarctic bottom cell. So it's really a shift of how much the water going up goes north or south in between the two simulations. And that comes also with it's tough to see, but on this plot, the boundary between the two cells is moving upward in that cold state. And again, it's something to do with which is related to observation. And just to point out that what's happening in there seems to be extremely similar to what has been described in that what's on the paper. Essentially, what happens in that coupled system is that the buoyancy flux at the surface in the southern ocean are changed in between the warm and the cold state. The winds are very similar, but the buoyancy forcing, which changes. And really, it's about water coming up and either being transformed into lighter or denser water. And the change in the buoyancy flux in that system has led to buoyancy loss, which is largely increased in the southern hemisphere. So pumping up that bottom cell. Just to make the point that the winds here are not a big player in the story. So this is the zonal means, zonal stress between the warm, sorry, I call it interglacial and glacial, the warm and the cold state. And you see that there is very little changes between the two states. So in the glacial state, you get slightly weaker winds at the peak and slightly shifted to the north. But really, it's not a big deal. Ultimately, if you compute the Ekman pumping, it's the same nearly in the two simulation. And that's interesting because it's in contrast to a lot of the literature which has been thinking about how the wind stress over the southern ocean might have changed and be the driver of how much the ocean overturning circulation has changed. And we are typically in the case here where the winds are not the player. It's the buoyancy change at the surface, the fact that you have a big ice cap which is changing the brine rejection. That's what is driving the ocean circulation in that simulation to first order. I have a question here. Oh, sorry. Here? In the warm state, I was thinking it should be open less, I mean, less denser. So it should not have this stronger, but I mean, you are showing that in the cold state, it's weaker, like they want to stay in the cold. It's really the buoyancy loss that matters. It's not whether it's denser in the first place or warmer. It's how much when the water is coming from the tropics into the high latitude, how much buoyancy does it lose? And so I can't remember. You have one of the quotas here. I can't remember what they've done with the buoyancy change in here, but I can tell you about this one because I checked. And the buoyancy loss is indeed a bit larger here than it is here. We did nothing there. You did nothing? OK. So in your case, there wasn't any buoyancy loss. I mean, in both cases, in the North, it was the same buoyancy. Yeah, yeah. In our case, it does change. The buoyancy loss does change a bit. But don't we have a strong state, as you're suggesting? Yeah. No, it's not because the warmer denser, no. Actually, if the warmer would, oh, OK, I see where you're going. So in that case, it's driven by precipitation. So the buoyancy loss here, though ultimately it has to do with how much heat you release to the atmosphere. And the buoyancy loss here, if I remember correctly, is thermally driven. So it's losing more heat here to the atmosphere than it is losing heat here. I can't tell you. I don't know. Is there any change? I don't know the answer to that. Does it copy figure that out? Yeah. Yeah. I'm not sleeping tonight. And you'll get an answer tomorrow. In the sake of time, I'm going to skip that. I'm just going to carry on with trying to make the case that those two states have similarities with the present day climate and the last glacial maximum climate. So here, what I've plotted is the overturning circulation in the small basin. So that's the one where you have this deep overturning circulation. So we have something like 15, 20 zverdraps. And when we move in the cold climate, we get 10 zverdraps. So there is a weakening of the overturning circulation in the small basin. And also, as you move to the cold climate, you get more water coming from the south, coming into the basin and coming out. So the small cell, which has gone up from, it was barely as zverdrap here, it gone up by factor 3. So now it's 3 zverdraps. The point here is that on this side, you have what panoramic people are able to extract from observations. So this is a map of delta C13 for the present day climate and for the last glacial maximum. And so what that map is saying is that in the present day climate, you see the trace of high delta C13 water being entrained into the Nordic seas and then coming out of the Atlantic at mid-depth. So you're reading in this the overturning circulation like that. And there are many caveats who is doing that. I should be careful. This is just a tracer. It's not really exactly a transport. But one way to interpret that is that you're reading the overturning circulation in that plot. Now you have the same plot for the last glacial maximum. And one way to interpret it is that now you see that the water, which have high delta C13 are coming in and leaving maybe at 2,000 meters. So the depths at which those water are coming out has risen by 1,000 meters. And then you get more water, which are extremely depleted in delta C13. So the hypothesis here is that those water are coming from the southern hemisphere. They haven't been exposed much to the atmosphere for a while. So it's literally water coming in here and then being entrained in the lower cell and coming in again into the Atlantic in very depleted quantity of delta C13. So in the model, we don't have delta C13, but we have phosphate. And because we have phosphate, we can turn phosphate into a proxy for delta C13. So what I'm plotting here is the delta C13 inferred from our simulation and try to compare it to the delta C13 from observations. Obviously, it's not perfect. We have a square basin. We have a flat bottom. But if you look at the numbers, it's pretty amazing. So in our warm state, we have high levels of delta C13 with that peak at the surface being entrained into the north at deep water. And if you look into the cold state, you see first an increase in delta C13. You see that water coming out at lower depths, and you see those very depleted delta C13 water coming from the southern hemisphere. So whatever is happening in our model matches very well, at least at the level that we can afford to make that comparison. And you have to be careful with those. This is not overturning. It's just a tracer which is carried by the overturning mixed by eddies and turbulent mixing and so on. So there is no one-to-one. But it's pretty interesting to see how these two match in that simple model. And so the two states looks like the present day climate and the last glacial maximum climate. So an interesting other comparison is if we take those two states and we drive a carbon cycle model with those. So what I mean by that is we take the dynamical state, the currents, the atmosphere, the winds, mixing of the ocean, et cetera. And we drive the carbon cycle in a passive way. So meaning the atmospheric PCO2 is not feeding back on the climate. So it's a very important caveat here. Nonetheless, if we do that, those two systems have a difference in atmospheric PCO2 of about 100 ppm, which almost exactly matches the observed change in atmospheric CO2 between a glacial and interglacial cycle. So it's quite interesting to think about how CO2 in our couple model has been stored in the deep ocean. So we don't have a land model, which is very fancy. And most people would agree that the difference between PCO2 between the, sorry, the difference in PCO2 between the last glacial maximum and now is due to the fact that CO2 was stored in the ocean. CO2 has been pumped up into the deep ocean. So how can we analyze that in our simulations? So the change in CO2 in the atmosphere can be decomposed in three mechanisms to pump it out of the atmosphere into the ocean. So the first one is the saturation pump. Essentially, what it is that if you have cold water, it can hold more CO2. So as you move from a warm to a cold climate, you just dissolve more CO2 into the ocean. And that pumps CO2 out of the atmosphere into the ocean. That's just chemistry. There's a biological pump. And that just reflects the fact that at the surface of the ocean, you have phytoplankton growing. And as it grows, it's taking CO2 to form its body, its cells, and sucking CO2 out of the atmosphere. Then those phytoplankton, they die. And they fall at the bottom of the ocean, where they can be re-dissolved into the water. But the net effect of that is that that pump is taking CO2 from the top of the ocean down into the bottom of the ocean. So away from a place where it can be g-solved back, sorry, outgassed into the atmosphere. And when you do that, there is a third pump, which is the disequilibrium pump, which is very important for our case. What that pump is doing is it's a measure of how much water are equilibrated with the atmospheric PCO2. So you think of water coming at the surface. Maybe it's cold. It has some amount of DIC, but it's so cold, it could suck up way more DIC out of the atmosphere when it's at the surface. But that parcel will need to remain near the surface for long enough to equilibrate with the atmosphere. And the equilibrium timescale for a parcel of water at the surface with the atmosphere is about a year. So many parcels of water in the ocean, they don't stay for a year at the surface, happily equilibrating with the atmosphere. They're going to get sucked back into the ocean on a faster timescale than that. So the water is not always in a chemically equilibrium with the atmosphere and temperature and salinity. And that's what this is measuring. So if you do that, what you find is that the solubility pump is sucking about 60 ppm out of the atmosphere into the ocean. The biological pump is actually the change between the two states, I should insist, is equivalent to 36 ppm back into the atmosphere. And the air seed equilibrium is the big term here. It's accounting for 85 ppm out of the atmosphere into the ocean. So that's the main driver of why the CO2 is lowering in our simulation when you go from a warm to a cold climate. So let me explain why, how this is happening. So this is a map of the disequilibrium pump. So it's showing how much waters are not in equilibrium with chemical equilibrium with temperature, salinity, and the PCO2 of the atmosphere. So that's for the warm climate and that's for the cold climate. So for the warm climate, what you find is that the disequilibrium pump is very small everywhere. And if you go to measure the disequilibrium pump in the present day climate, it is small everywhere. It's a small term. That's the disequilibrium pump in the cold state. And what you see are big, large values in the Antarctic bottom water cell. So that cell is corresponding to a flux which is going around like that. How this happens is that there is an ice extending, ice cap extending almost across the upwelling region of upwelling. So what happens is that water coming to the surface because they are trapped under sea ice cannot equilibrate with the atmosphere. So you have cold water which holds a lot of CO2. They come up to the surface. They are under ice. They can't out gas to the atmosphere before being entrained back into the deep ocean. At the same time, you have biological production is still pumping CO2 into the deep ocean. And it doesn't matter whether it changes between the two states. It's always doing that. You always have phytoplankton growing, dying. So slowly, the CO2 content of that water is just increasing and storing a lot of CO2 in the deep ocean. That mechanism is actually not new. Stevens and Killing suggested something along those lines in 2000. But they used a boxed model to do that. So they had a boxed model which so they split the ocean into four or five boxes and just accounted for the flux maybe in a stone mold way. In our case, this is actually happening in the couple model. So it suggests that that mechanism could survive a more complex type of dynamics than has been suggested. I should say there are caveats to that. We do not reproduce all observed features of the last glacial maximum. And there is one which is a big issue for us is that there are suggestions that the oxygen content of deep waters was actually lower at the last glacial maximum than it is now. And we have exactly the opposite. And that's probably because in our model, we are missing some of the process. And possibly we are missing something like iron cycling and so on. But nonetheless, if you start to make a list of what do we know about differences between the last glacial maximum and now. And you put that against a list of the differences between the warm and the cold climate we have in the model. There is a list which is so that I made the list. So if you think, for example, of deep ocean temperature, clearly the system was colder at the last glacial maximum. The range is 2 to 4 degrees Celsius. Estimates we have 7.7. So it's a bit too large. But it's in the bulk number. It's the right order of magnitude. The salinity estimates are that at the last glacial maximum, deep water were probably one or two PSU saltier. And there is a reason for that that we can't reproduce that number. Most of the reason why the salinity of the ocean was larger at the last glacial maximum is that we took water and we put it over land. And so just the whole ocean got saltier on average. And we don't have that effect. An interesting one is that the sea ice extent in the southern hemisphere is estimated to be shifting by about 10, 5 to 10 degrees of latitude in observations. And we get about 13 degrees of latitude. And the reason why that number is important is that it's what is driving the PCO2 storage into the deep ocean. The sea ice extends in the cold state right above the regions of a pooling. That's why water coming from the deep ocean to the surface can't outgas. It's because you have that sea ice cap, which is preventing water to equilibrate with the atmosphere. So PCO2 is in the right range with the caveat that we don't have an active, radiatively active PCO2. And we have interesting things like the depth of the MOC is also changing in the right direction. And as I mentioned, the deep oxygen is a big. We just don't get the right sign there. The point of that is they look kind of similar. And when you think about the lack of complexity of our coupled climate simulation, it's quite amazing that we can reproduce so many features of the last glacial maximum. And so we are trying to make the case that maybe we should think about glacial-integral cycle as being representative of multiple states of climate. There are many tick marks there that seems to suggest that those two states could be multiple states. And if you go that way, I'm not going to go a lot into the stochastic resonance. I just want to make the point that when people look at the paleoclimatic record, so that's various paleoproxies that people have, and those are time series of the orbital parameters, the mean and cubic cycles. And we know that the precession is varying with a 20,000 timescale periodicity. Obliquity is 40,000, and eccentricity is 100,000. And those three periodicities, we find them in the paleoclimatic record. That's been established for a long time by this haze paper in the 70s. A few years ago, there was a small review article in Nature celebrating this achievement of being able to make the link between the paleoclimate oscillations of the glacial-integral cycle and the link with the orbital parameters. But the interesting fact is that despite those 30 plus 40 years, whatever happens between those oscillations and the response of the climate system we still don't have an answer to that. And that's a very neat symbol here of having wheels which are turning in various directions. And we still haven't figured out the mechanism where we go from there to there. We do have statistical robust evidence that the glacial, sorry, the Milenkovich cycles are seen in the glacial-integral in the paleoclimate record. We don't know why. And I just want to point out that, again, if you look at the literature, there are a bit of a two-family type of view of the problem. There is a linear view, which is we know the oscillations of the incoming solar radiation at the top of the atmosphere from Milenkovich. And from there, we have a response of the climate and maybe some very strong feedback because we know that the forcing at the top of the atmosphere is due to Milenkovich is not that large. Especially if you start to integrate across a hemisphere, it's actually almost zero. Effectively, in many cases, it's exactly zero if you integrate globally. So why is it that the climate system will respond to the incoming solar radiation at 65 north and would not care about the exact opposite forcing at 10 north? But people have been focusing on this. So usually the way to go around that is to think about feedbacks. Why is it that the forcing at 65 north is so important, way more important than the one in the tropics? It's not obvious. And CO2, ice sheet, and CIS are usually the top of the list feedbacks you think about, at least what proposed in the literature is. The high latitudes are really important because that's where all those feedbacks are acting. But essentially, it's a linear view. There is a forcing, and then it changes something in the system. And then the system responds by a positive feedback, making the response to the forcing even larger. And then you drag the system from point A to point B. And then there is the non-linear view, which is the one we are trying to push forward, is that you think about that double well, and you have those coexisting states. And now the question is not how you go from one well to the other, but how would you go from one well to above the hump? And so it's a very different type of dynamics. And clearly, this is, I would say, the prominent view in the literature. There has been lots of work along those lines. But really, a lot of that was done with a very simple model. And we are slowly getting onto the trail of thinking that maybe we should be pushing a bit harder in that direction with more complex model. Yes? Your two states seem to be further apart than in reality. And when you're at minus 7 cold or something like that, your cold state is 7 degrees colder or something like that. Now, in fact, that is with the radiation state, which is not responding to the CO2. So if you actually reduce the CO2 down to 1,9, then your state would probably be about 12 degrees colder. So it would be a lot colder. Actually, we did that experiment. It's just a few degrees colder. The issue is that we know why the separation is too large. And it's actually a bias in the warm state. The warm state is too warm. It's not the cold state, which is too cold. And it's too warm because the North Atlantic deep water is too warm. And it's too warm because we don't have Canada. Canada is the cause of Brian is looking at me weirdly. What? No, we don't have, I think what we are missing is that we don't have those cold, dry outbreaks over deep convection zones of deep air. You don't sink at the right temperature. Yeah. And so we don't produce those 3 degrees Celsius deep water that we should be producing. But it is an issue. Yeah, it is an issue. OK, almost done. So I'm going to skip that. I just want a brief plug, which has nothing to do with multiple states. But we've been doing some work with high-obliquity playing with those aqua planets and looking at the high-obliquity case. And looking at the reason I'm just bragging about it is that we just got a citation in the New York Times. Two days ago, but people were interested in why could we have life being sustained in a high-obliquity world. And the answer is yes, because there is an ocean. And I'm not going to go further into that. And just a fun movie. That's actually another case of a tidally locked planet. So that's a planet which is turning on itself at the same rate as it's turning around the sun. So it's always showing the same face to the sun. So there's a face which is always day and a face which is always night. And so we did a coupled simulation of that with the same system, coupled ocean atmosphere. So the fun stuff is that you can change the rotation rate in that model. And obviously, if you have a one-day rotation rate, which is our present day type of rotation rate, the ocean is not ideal resolving. Because the Rasby radius is about 30 kilometers, and we have a grid cell of 300. But you can be doing some cheating there. You just increase the rotation rate to 20 days. So because the Rasby radius is NH over F, and F now is way smaller, the Rasby radius goes up. So now in that simulation, we have a Rasby radius of a few thousand kilometers. And we have an ideal resolving ocean at a three degree resolution. But what's really fun is that, if you look at, whoops, is it going, is that the ideal resolving ocean is able to carry a lot of heat from the warm side to the cold side. So there is way less sea ice on the cold side of that ocean, which is ideal resolving. But now, if you look at the atmosphere, when you have a one-day rotation rate, what you see is that you can see synoptic scale ad's in the atmosphere. Normal. The atmosphere is ideal resolving at that resolution. If you go really slow, now the ad seems to disappear, and we only have convection. Again, if you put numbers, the Rasby radius is 1,000 here. So now it's 20,000. It doesn't fit anymore into the hemisphere. So we just have a convective atmosphere sitting on top of an ideal resolving ocean. And here we have an ideal resolving atmosphere sitting over a parameterized ocean. It's just fun. And conclusion? So really. Your plan has to all be the starting 20 days. Yeah, yeah, yeah. And so effectively, because of K player's low, it should be further away. No, you have closer. Sorry, it should be closer. But we did not change the solar constant to make the comparison. But in theory, the rotation around should go with the distance, yeah. OK, just to summarize, this here, it's just the currents. You have jets in the ocean which are going one way or the other. So there's a system of alternating jets. And at the equator, it's westward. No, sorry, eastward. And off the equator, they are westward. And so that's just bringing heat, yeah. And here it's strong enough to just melt away the whole. OK, summary, really what we just want to point out is that when we think about multiple states, I think there is obviously a huge literature which has been emphasizing the AMOG base stability. And what we are thinking is that maybe we think more about ocean, ice, multiple states and instabilities. And as pointed out yesterday at the end of Brian's talk is that there has been lots of efforts into studying the AMOG base stability. There hasn't been anything near which has been done for the multiple states of the sea ice albedo type. And maybe we should be doing that more often. Have you watched the sunset last week? Last night. Yeah, Kelvin Helmholtz instabilities. Really? Questions? Sorry. Several years ago, with Ricardo and with my student, we did a similar experimentation with large basin versus small basin with constant. And we did a project to reproduce your, without constant, large basin versus small basin. We had difficulty to reproduce your results. And also with constant, we made many experimentation with river discharge, how river goes this way and this way. And we made so many experimentation. And finally, the river discharge doesn't make much change. But the albedo, this simulation was very sensitive to albedo, particularly in tropical content albedo. So we don't know how to control this. So then we gave up two years ago. You shouldn't? You should not? That's why I said, we need to exchange. And we have to, I mean, your simulation should be by rock. And my simulation? No, I'm kidding. No, no, no. Bad luck? Yeah, bad luck. Good luck. Good luck. Really, yes. So, I mean, I don't know, how you prescribe albedo in your case? Continental albedo? I put A number. A number. One number, yeah. I think it's 0.1, which is something like a forest. OK, and what is it? 0.06, I don't know. I would have to check. Yeah. Yeah, I think, I mean, simulation. Our system is sensitive to the albedo choice. It's definitely sensitive to the albedo choice, yeah. Yeah. I don't know if it's good or bad, but it is sensitive, yeah. Do you see the same atmospheric physics? I don't know. I think we are both using multistrike physics, aren't we? The physics, yes. Physics? Our case? No. Physics 6.50. Oh, your model? Yeah, our model. Oh, OK, OK, OK. So it's the, OK. We were complete. OK. Oh, yeah. Yeah, that's. Completely different. OK. So we should call it something. Yeah, yeah, we should, yeah. We should figure out if there is some obvious way to go, yeah. Sorry. There is no El Nino. I can't, we had that same, yeah. I think there is no El Nino. I remember checking, but there is not, no. Actually, there is probably too little variability in the system, if anything, because theory would tell you that there should be some spontaneous transition from time to time, and we don't see any. So maybe it's a sign, but no, we don't have El Nino. Well, so I skipped the stochastic resonance part, but for any noise into the system, you could compute a probability of transition. Hours is clearly larger than multi-thousand years. I don't know what should be the real number. So we could add noise, and we could add noise to the point where we see transitions, definitely. But I don't know what's a good, we could use observation, I guess, to gauge how far we are off the mark, yeah.