09:02:15 Anyways, let's, let's get started with our working group today. 09:02:20 Whoops. That's not what I wanted to do. 09:02:23 Yeah, so I just want again give a big picture of what's the climate application working group as a whole is doing. 09:02:29 Last week, we focused on PD staircases on Jupiter. And today's discussion will be focused on staircases in the poor oceans due to interactions with ice. 09:02:40 And you can see the sessions that we have laid up for next Monday and the Monday there after. 09:02:47 As for how things will proceed. 09:02:49 We're going to start off some short talks and we'll, we'll be recording those, including some of the overview discussions that follow. 09:02:57 Once we get to about the hour mark, we will take a break and decided that point if we will go into breakout rooms are just stay as a whole. But the discussion at that stage will not be recorded. 09:03:08 Unless we do go into breakout rooms then when people come back after a half hour, so at the one hour 30 minute mark. Then we will record the summary of what was discussed in the rooms. 09:03:23 And also, just to keep a focus on what we're trying to do here we're not trying to solve any problems. We're just trying to ask what are the pressing outstanding problems, what is needed to make progress, and what skills from the working group can be 09:03:34 combined to help us address these issues. That's really what we're trying to accomplish in the next couple of hours. And so with that, I'm going to pass the shoe over to Leo, who will give an overview of the specific topic on social interactions. 09:03:50 Go ahead, Leo. 09:04:05 Lee I don't think you have your microphone on. Yeah, I think I was muted. 09:04:11 There we go, and thanks so much for. Yeah, inviting this, and this sub working group. 09:04:19 I'm very, I've been very excited about having this trip. Because, you know, I think it's really cool area that a lot of people going to get a lot out of. 09:04:27 So I'm going to give you a brief and incomplete introduction to ice adjacent layering. And please do ask any questions or, but in saying that you know, I've missed that the most important part, or whatever. 09:04:38 And that's all very welcome. 09:04:40 So, I'm Leo, and I'm a PhD student dumped, and also at the British Antarctic Survey supervised by john Taylor, to keep Nichols, and just to sort of whet your appetite for this is adjacent there. 09:04:53 There's this plot from Jacob's a towel 1981 of these lovely observations of some very clear sort of 70 temperature steps adjacent to Airbus glacier tongue. 09:05:06 In the west of Antarctica. And so, well, we'll get to them but first you have to understand about isolation and directions before we can get to the layers. 09:05:15 And so we go. 09:05:18 So, there we go. So, the basis of a lot of isolation interaction and what I'm sure many of our speakers will be talking about is the three equation model. And so this that abusive version of this kind of introduced in, in 1950, and it, the fundamental 09:05:34 comes down to sort of an equation for the heat flux and equation for the salt flux equation relating that kind of the temperature of melting breathing, at which that happens, and the pressure as well. 09:05:45 And so this first one that he blocks you can imagine the heat coming in from the ocean, it's got a balance that he coming in from the ice and the latent heat, and from the face change. 09:05:55 Likewise with salt. You've got the salt coming in from the ocean, not very much goes into the ice it might be sort of important in some specific scenarios, but then you've also got your fresh water flux and sort of coming in. 09:06:08 And that could be melting or freezing but mostly say will probably be talking about melting. 09:06:12 And then finally yeah this liquid is condition basically saying that if you've got salty water, it decreases the, the freezing temperature, and likewise you've got this pressure dependence as well. 09:06:26 So the context for some of this the reason why it's kind of such an important topic to be talking about is obviously you know the geophysical context. 09:06:34 We're living in a time of great change with a lot of changes to the, the cryosphere as it exists and a lot of those links changes in the oceans, so for instance for ice shelves, you know this is this lovely picture of kind of melt rates around the ice 09:06:58 shows and Antarctica and you can see this elevated and melting on the west of Antarctica which is particularly worrying. 09:06:56 And a lot of this is thought to be due to changes in the ocean circulation. 09:07:00 And you also have changes on like Greenland ice sheet, and particularly sort of amplified near the edges here, and here you have a lot of kind of these marine terminating glaciers, that you're sort of some kind of over the top ice but then. 09:07:15 And then finally see it it's another pressing issue with the changes in CS extent. And this kind of gives you a realm of different possibilities for social interaction so you could have, for instance on a shelf maybe the ISIS overlaying your top the ocean, 09:07:29 but potentially you know with a marine terminating glacier you might have sort of a lateral and, you know, horizontal or vertical space, which are worrying about. 09:07:40 So, because of that mass loss, we want to be able to model, those different situations so that we can, you know, understand how it's going to change into the future. 09:07:51 So the thing that's usually done in ocean models is people use a three equation prioritization, basically where we parameter eyes that input of heat flux and salt flux, and in the top quintile of your simulation, or what have you. 09:08:04 And usually that's done, basically using kind of a nozzle number, which we put into this exchange velocity, lowercase gala tea and gamma s. 09:08:15 And, which then are kind of thought to be proportional to friction velocity, which can relate to a Fairfield last to using the drag coefficient, and these exchange coefficient began it began. 09:08:27 So you end up with these three unknowns in your free equation prioritization, which are, you know, they're a mixture of of observation the constrained and tuned to make the models work. 09:08:40 And, but, you know, there's a lot of physics trapped inside this drag coefficient, and this committee and cameras and, and particularly becomes a problem when you start thinking about layers. 09:08:51 So, just to start with the vertical case. And, you know, there's some very classic results so this would be hot button turn in 1980, where you have a salinity gradient, and you've got a block of ice in there, and you get these very impressive intrusions 09:09:07 coming off the side which, you know, dominate the contribution to the melt rate of this ice block and change the circulation, etc. 09:09:15 and. And recently, this has become, you know, much more prominent and well studied because of various observations, but particularly these recent ones in 2019 from Jackson I fell on the comp glacier in Alaska. 09:09:30 And these are, these fantastic observations really close to the glacier where you can see these are temperature profiles, you get these really large peaks in temperature basically these intrusions are coming off of off the ice face. 09:09:44 And you can kind of see in the density as well you get sort of flat layering. 09:09:48 So it's obviously very important because I'm not obviously but the. 09:09:55 The key thing about the glacier. 09:09:58 The other thing that has been happening is that it's got much elevated not rates from what is expected. So, and there's a standard way of evaluating these things which is kind of a chapter blue model with some subglacial discharge to that three equation 09:10:13 model to predict, you know how much money you'd expect. And, and essentially the standard values have used in, in those models give you a very low amount rate observed. 09:10:26 You know orders of magnitude larger. 09:10:28 So in this paper they kind of play games where they try and change all the different parameters and see if they can tune this three equation model to give you the right answer. 09:10:36 And basically, they find that it requires really kind of large changes in all of your parameters to to make this three equation model work, 09:10:46 and in particular they found that a horizontal velocity was very important so velocity going along the, the face, which is kind of not included in the, the classic kind of intrusion, ideas, 09:11:00 but to move on to think about the horizontal case, which may be more applicable to say and I shall. 09:11:10 There is again some classic work so this is Turner instrumental This is a celebrity gradient in a box and it's heated from below and you get these very cool and kind of double abusive layers that are just forming stacked on top of each other. 09:11:18 And this was linked recently in 2015, do some observations on George the six. So, this is on the western toxic peninsula. 09:11:27 And you've got this borehole observations taken at school point here. 09:11:31 And in those Ts profiles you get these very distinctive layers, intelligence and it just a sort of, and they're very very adjacent to the ice, you know, few meters away. 09:11:43 And also, there's some cool CS observations which will be talking about later. 09:11:47 So I won't go over them too much now. 09:11:52 And, yeah, so the argument made in them in tomorrow towel, was also based on another piece of classic work so this is Martin and Cashman in 1977. And this is a really basic experiment, but it was kind of very interesting, basically you've got a little 09:12:05 box of water that's it about zero degrees, salty water, you know, ocean salinity. 09:12:12 And then you put a block of ice on top and you just see what happens. And you so immediately you get these grind Banjul is away from the ice and because of this difference in will actually give you 70, the temperature boundary layer is growing faster 09:12:33 layer, which actually enables us to get this peak and density of drawn in here, and I kind of discuss it in the paper. And, which can force convection. 09:12:36 And what they found over there today bank to experiment is that conviction was very long lift and it continued throughout the experiment. 09:12:43 So they didn't get any layering. But they have no initial stratification so you might not expect to see something similar to the Turner install, where they had a stratification that was being heated. 09:12:55 And so we've got to have them talks about this, so I won't go into too much. Now, one thing just to say is that there's no convective steady state here, so if you balance, if you take a steady state assumption, and you balance the flux of heat and salt. 09:13:07 You can kind of show with the three equation model that a positive turban or buoyancy flux, a potential energy, generating turbulence, as in conviction requires a boundaries lunatic that's very very low, which is sort of unrealistic in that it's not observed. 09:13:26 But yeah, as I said, this will be a topic of discussion correct. 09:13:31 And so another thing I just wanted to say about those Georgia the six observations, one of the things that I've been doing in my PhD is looking at some more in data from under George the sixth where we've got these turbulence clusters, and four and 14 09:13:44 meters away from the ice. 09:13:46 And one of the results that we found is, we were basically trying to see if we could apply the three equation model as best we could to these observations. 09:13:55 And what we found is, even by tuning those parameters in the way that they did for the vertical eyes shut off political aspect, sorry. 09:14:03 And this was the best we could do so. The yellow and purple here all of the different observations. And we've been them by me inflow speed and going past the, the ice and actually the, the three equation predictions are incompatible with the observed 09:14:19 developments in that they increase with means bees, which the observations and density. 09:14:26 So, that is to say that there's some problems with the three equation. 09:14:31 Melt privatization, and that they are linked to this layering in various observations that have happened. So, it's kind of a really important application for layering. 09:14:41 That is kind of yet interesting but very complicated, so little summary of the things I've said about the three equation, and why it's important. 09:14:51 And so I'll just quickly go over my first ideas for those big questions that Bruce was talking about. 09:14:58 So the big motivating ones which are not, you know, I don't really get anywhere near these but the big one for me is, what is the effect of this layering, fundamentally, what are the feedbacks on loss of ice mass, and ocean circulation. 09:15:14 And so, you know, working towards that point, you know, how can we get to understanding that, and it's likely that we need to be able to include these effects into ocean, OCEAN models which have ice interaction. 09:15:26 And to get there. We need to understand how prevalent these layers are these observations that we have from George the sixth say or the CIS observations and you know we'll talk about, Are they the norm of A, you know, very special cases, it's not clear. 09:15:44 And when those layers are present, what exactly is the effect when I smell, which maybe we'll get up with some, some simulations later in the session and kind of linking to these questions is yeah what triggers. 09:15:54 And what maintains these, these layers. 09:15:58 Some slightly more specific questions maybe with the vertical eyes, what's that role of the along along ice velocity, that was a thing that was mentioned in in this paper where it has quite a large effects and having a long along nice philosophy and so 09:16:11 you know the general question of how these intrusions of a vertical space interact with them via turbulence is is a big open question. 09:16:19 And here's the three equation proposition on fit for purpose here. 09:16:23 Same question you could ask the horizontal space, and. 09:16:28 And then also, can we explain the George the six observations with these this boundary layer, sort of idea that we talked about with these Martin and cavemen experiments, or is that, you know, a non starter. 09:16:42 Basically, you're, you're faced with the same problems that you face in the open ocean in terms of explaining why layers are there. 09:16:48 But now with some extra problems of this is so any, any questions about you know oh well maybe this is the mechanism for layering. It's also possible. 09:17:00 That's true. Underneath ice as well. And finally discounts and methods. Before we start on to our talks. 09:17:05 Obviously, observations are going to be vital for us understanding this problem, and that are expensive and tricky lab experiments, there are a lot done in the past, but reasons, leave have been much fewer, I mean there are some good examples but I think 09:17:19 there's a lot of scope for a lot of experiments so these processes, and then simulations have really come into their own recently, so these larger the simulations. 09:17:26 We probably some this year. and then also improved on a hill at all. 09:17:38 Last year, and then we're going to hear from Maddie Rosevear who did some, some cool simulations, a second multi based simulations which are really neat kind of new advancement in thinking about these problems where you can evolve the ice as well as the 09:17:46 as the ocean. So Louise going to talk to us about that later. 09:17:49 And then, finally, this is something that is maybe a bit more for this audience is maybe thinking about a 1d model of layering. You know we've been talking about various those through the workshop, it'd be interesting to add one of those two or three 09:18:01 equation model in some sort of interesting way. 09:18:05 Yeah, so that's all I've got to say by way of an intro. If anyone has any questions about that, else we can move on to hearing from Maddie. 09:18:22 I think it's that was a course overview and we're going to see more details later Leo we should just go on talk right away, that'd be great. Great. So, Maddie unfortunately couldn't be with us today. 09:18:35 But she is centers recorded to do you want to show that young. Yes, absolutely. So, I'm going to send the link to the chat to Maddie stock. 09:18:47 And let's plan to be reconvene and discuss that presentation in in eight minutes so the topic is eight minutes. 09:18:59 And yeah, I'll see you. I'll see you in eight minutes then enjoy. 09:19:06 You know yeah I might also just instead of having silence for eight minutes, how would I try to play the movie as well, if it crashes and burns crashes and burns. 09:19:17 Alright. That sounds good. Some people might find that playing a movie over zoom doesn't work too well so we encourage you to follow the link. 09:19:25 But I'll put it up here and we'll see if this works. Besides, that's good. Thank you, Bruce, you're right with that. Yeah. Okay. 09:19:39 Hi everyone I'm sorry not to be there in person, but I'm feeling a bit under the weather. Sorry person I should say live. 09:19:47 I guess none of this is happening in person. 09:19:51 But thanks very much first Leo and Jada for the invitation. For those who don't know me and Maddie rosier and base at the University of Western Australia, and I'm talking about work I did during my PhD at the University of Tasmania. 09:20:06 With that game from the University of Melbourne, and then getting plenty from the university, sorry from the Australian anti-collision. 09:20:14 So this is work on the role of diffuse of conviction in the base of emotion event tactic I shells. And I was investigating this season, large Eddie simulation, and this is my model domain. 09:20:27 So it has sort of a spatial scale for that 10 meters by 10 meters, although it varies depending on the current speed. 09:20:35 And what we're looking at. 09:20:37 on the horizontal plane is the instantaneous note rate for this session interface, and beneath it is the ocean and we've got the slow solution so here we're looking at the vertical velocity, and you can see this small scale fluctuations associated with 09:20:55 calculated turbulence. 09:20:57 And on the right hand side is the temperature so you can see the warm water here is that depth and we have this cool actively making may have been API's. 09:21:05 And so we have a steady sort of flow in the interior of this domain so set up a geographic balance and so we get this equity value there, which you can see green, and the conditions for these simulations so the the temperature is sort of between point 09:21:23 two 1.4 degrees 09:21:27 point 2.4 degrees warmer than a local freezing temperature and the velocities are in the range of zero to two centimeters per second so these conditions are loosely based on observations from the grounding line of the research shows that were published 09:21:42 by Carolyn Beckman and others in 2018. 09:21:46 Alright so straight onto the guess three key results from my work and so I have a recent paper in PNS about this stuff, and it's in the slack if you'd like to look at it. 09:21:58 So firstly, that the guests that we get the fish convection. 09:22:13 During mixing and the formation of them Aaron staircases nice I shows. And this happens even in the presence of sort of sheer from an ambience Alright, We found that in the diffuser connective regime melting is inherently time dependent. 09:22:20 And this is because the associated face and mix later so flexors are not in balance. But more than that, soon. 09:22:29 And because melting is transient decreases over time. As soon as the strength of diffuse of conviction. And so, what we see in the simulations is, we get a transition from defeated invective dominated shear dominated mixing at some point, provided we 09:22:44 have some shear in the simulation. 09:22:48 So firstly the hill and staircase. 09:22:51 We can go straight to see here so looking at the stratification of the water column and how it evolves in time, we form a mix layer beneath the ice with a defined benefit plan sharpens over time. 09:23:06 And then we get the formation of a second, then maybe Sir, if you like. 09:23:21 Mix I have a nice, so this is out there Mahela and stickers. And you can also see negative and squid values associated with this convention. I'm going to skip that for time. 09:23:23 So I said, Nothing is time dependent, in the face of collective regime so we'll have a look at that. So here we'll get melting and how to goals in time, these simulations sort of fire this purple line that sits above the others, and all have the same 09:23:40 that we're driving but different current forcing a different friction velocity, we find that the section velocity doesn't really affect them outbreak, very much in a singer in the simulations. 09:23:53 So what is driving this, I guess evolution. So, in panel be, we're looking at the thickness of the thermal and say land diffuse is suddenly so these are the regions of the flow where heat and sort of being transported by molecular fusion, pretty much 09:24:12 exclusively. So, and this dashed line is the diffuse evangelist. 09:24:22 And the solid line is the equivalent. 09:24:23 And what we can see is the sale and if he says somewhere sort of grows with this 09:24:30 sort of efficient link scale, simple scaling given here, that's the gray line. 09:24:36 And what's happening is the system is behaving sort of like the high stability ratio diffuse of interface, sort of, you know, as if it were between two layers. 09:24:45 So in this case, convective overturns and not removing salt from the edge of the deficient region next to the ice. So another way of looking at this is by looking at the Boise flexes over depth. 09:25:02 So we sort of have a have a diffuser contribution to the ice. And if you look at the blue contributions for salt. 09:25:12 And then we have a turbulent contribution. 09:25:17 Within the next layer down here and you can see, these two are not balance. This is sort of a manifestation of that accumulation of salt, next to the isolation interface. 09:25:26 And that's driving at time dependent memory. 09:25:30 So next, I was transition from the piece of connective tissue dominated mixing so the quantity we're looking at here is the way it flips or, you know, turbulence production from ynT over depth and time for two different cases with different, different 09:25:42 temperatures so different, elevation of the temperature above the local freezing condition. So on the bottom this is the case that we've looked at before where we saw the formation of the staircase. 09:26:00 So you can see that initially the buoyancy flexes positive. 09:26:04 And then we sort of had this competition, and goes quite positive negative. 09:26:10 And eventually, you find that, you know, she had dominated mixing and it's occurring, next to the ice. Interestingly, that interface beneath it still is still associated with positive points is that so you still have the physical reaction going on down 09:26:23 here. And that's consistent with the those layers that we still growing at a later time. 09:26:31 But above this case of hide them driving and we see that the positive points effects of the diffuse of collective dominant mixing sort of persist for much longer. 09:26:42 All right, those are my three key results. 09:26:46 And then I have a couple of questions for the discussion so I showed that the net rate is changing and super helpful for you know trying to prioritize melting. 09:26:59 And in sort of large scale. Ocean and climate models, so you know what mine is high average my rate look like the next raging. And it seems unlikely that we can keep growing this defuses sublet If ever I imagine be interrupted by transit processes like 09:27:20 current variability or breaking in tone waves. You know, so what what role does the frequency of those events play. 09:27:25 And what does that mean for at a time average noting, it's probably a pretty nasty question but also an important one. And when and where would the physical mentioned be important for basal melting. 09:27:38 So how prevalent use it, I guess. 09:27:40 And then finally, you know, what's the effect of a basal slope and when do we, when does this problem turn into natural production. Instead, and callers my questions that so you guys. 09:27:54 If you have questions for me. 09:27:58 Put them in slack and I'll try and answer them for. 09:27:59 Okay, so sorry again those be there in person, and thanks to be done by. 09:28:07 Well hopefully that worked out and Evernote, see the video in here. It was great. 09:28:14 Okay, great to know. So back in your hands you know and Leo. 09:28:19 Thank you. Thank you, Bruce. 09:28:22 Well I suppose if people have questions for Marty please put them on slack. 09:28:31 And unless you want to discuss something right now. 09:28:36 Any, any comments on the talk. 09:28:42 One thing I'd say. 09:28:49 I just thinks very interesting about mother's work, which is very very cool. Is the simulation she's using is a larger the simulation, which has heightened resolution near the AI space, so that it can resolve the diffuse of subletting, which is, you know, 09:29:01 sort of what the similar thing to, to what other people have been doing in their larger the simulations. 09:29:07 But then you see these layers actually forming, despite the resolution in the backfield. Not being very high so you're getting layers. 09:29:16 Despite that, which I think is really interesting it opens a lot of questions for how we can use unresolved kind of larger the simulations to potentially resolve kind of layer like processes so brilliant. 09:29:30 Anyway, 09:29:33 right. So, Leah. Do you mean that these deeper layers are just a larger scale much thicker. 09:29:43 I'm just that I would have expected for print a few Civ layer that you need to have a lot of resolution there in the simulation. 09:29:50 But actually, I think that's relatively little. 09:29:54 So on that point what I would say hi to some here on that point what I would say is one of the main things are largely simulation as most of you might be knowing does this take out the energy from the highest mode. 09:30:09 Right, so. So, if the physics, which is driving these layer formation is not dependent or, essentially, what matters is what. So essentially, if it doesn't matter how you take out the energy. 09:30:28 Right. 09:30:28 Then it's okay if it's only about a so if the physics is driven by other skills which is not really small skills which is unresolved in this case. 09:30:39 Then I don't see an issue with it not forming the layers. 09:30:44 Of course, if there are, if there are some physics in which you have energy going upscale. 09:30:53 Oftentimes we see in stability stratified environment that is you can have some, especially in a when you have some sort of Coriolis forces, and you can have energy go upscale from really high wave numbers to lower wave number. 09:31:08 Yes, in that case a largely simulation might be a problematic because if the earliest what it is doing is only taking out energy then it might be problematic and it might not be accounting for all the physics. 09:31:21 But if, if, what is only happening is taking out energy from the land tiles with our highest modes. 09:31:28 Then I think it should be fine. 09:31:31 Yeah, it's kind of. 09:31:32 Yes, good point. 09:31:34 Thank you so. Yeah. 09:31:38 All right, anything else. Can I ask a question to. 09:31:49 What is 09:31:54 is or any provisions of the week for volunteers, they can go into laminar values. 09:32:02 I remember years ago when lady simulation of backward facing step, it was very important to have this continuous transition in Poland in our values. So, what is the situation. 09:32:13 Yes, I'm not totally sure about exactly the, the setup that Maddie has, but I know, certainly in the setup that a lot of other people doing these bloody logic simulations, but doing. 09:32:26 Yeah, it's very key that they do have that change to lambda. As you approach the debate and I think Maddie does actually. Yeah, because he has to capture those, those boundary layers that all kind of probably lambda, very well resolved. 09:32:47 All right, maybe you could transition to, I mean, I think. 09:32:53 One more question from Martin. 09:32:55 Yeah, for modern Yeah, exactly. 09:33:03 Obviously audio is coming through when the video is okay. Yes. Okay. 09:33:04 Um, I'm not very much familiar with DNS and Ellie so earlier studies that you mentioned. Can you comment on the boundary conditions that you are using conventionally so immediately I surface and would be very much interested in what you, what you prescribe 09:33:21 etc surface. Yeah, so it's a similar story between them. 09:33:25 I think next I'll just talk briefly about the simulations that I have that are very similar to Mattie's. That's why I'm sort of answered lots of questions, because I've had my head in this a lot. 09:33:34 And so the way the Badger conditions work it's, you know, Everybody's boxes periodic in in the horizontal directions and at the surface, it's got a three equation model, like, similar to the way that I can the introduction, but it's the diffuse of the 09:33:49 equation. And so, I'm not entirely sure how it works in, in her code but in our code, we have like a go sell above the interface, and you calculate the gradients across that goes so, and then you update each sale at each time. 09:34:03 The, the new kind of outputs from the three equation model that you can then have as your boundary conditions that sign up. 09:34:12 In order to treat you is right i mean you have some melting of the ISO, 20 and simulation that you're conducting and you're kind of flux is the key that you don't you don't change the volume. 09:34:24 Blocks is the key but you don't you don't change the volume. So you say but the actual amount of fresh water isn't going to change the volume, but it just acts as a boundary condition in temperature internet because it's very I mean I just started to 09:34:35 work into connection between this workshop. And this was probably the most puzzling point to sort out what boundary conditions are you guys using and like what is conventionally accepted and what like boundary models may be existing in order to come up 09:34:47 with for example the eyes to the ocean and things like that so I'm just I'm very much interested in that. So what did you say you pull up on this in the next minute said I'll just keep saying and then wait, wait, wait. 09:35:01 Hey, well, okay thank you very much. 09:35:04 Well done, perfect time for you, Leo to present to present UK. Right, I won't go on too long because I realize I've spoken a lot so far. And, you know, just a common big schedule I sent on Slack, it's just a tentative schedule. 09:35:18 So we of course want a lively discussion so yeah just use it as a tentative schedule. We don't have to rush. 09:35:30 Take it. It's all yours Leo. 09:35:33 Great. So this is just a bit of an overview of some of the work that I've been doing with Guttenberg until bullhorn john Taylor. 09:35:41 It's just come out this year in JPO. 09:35:46 And, and it's some similar simulations, essentially to the ones that Maddie was talking about, but with a slightly different setup so slightly different dimensions we've again got this, this melting face on the top in the same way. 09:35:58 And, but now instead of sheer driven turbulence, we're doing sort of a body forcing of tournaments, so we're forcing the high wavelengths, the light wavelengths. 09:36:05 And Pharaoh. 09:36:07 And then letting the turbulence develop from that and we're sort of using a random, random forcing there and. 09:36:14 And the way the way it's, it's set up is to achieve a specific dissipation so we plug in the desired dissipation rate and the debits will adjust to to force at that rate. 09:36:25 This was sort of motivated by the observations of George the six that I talked about in the introduction, and where we have measurements of dissipation and we didn't want to sort of assume that it was sheer force because it didn't sort of match up. 09:36:37 In terms of the male behavior with info. 09:36:42 And then we relaxing the temperature insularity in the field so that's sort of linked to those boundary conditions that Martin was talking about. 09:36:50 And you've got to see the nice thing about this boundary condition on the top is you get to find out about the actual Mel patterning, which actually, you know will be relevant for them, Louis, talking later but these simulations. 09:37:03 So we got a similar a similar story to the one that Maddie maybe didn't quite have time to talk about but basically where you have these peaks and density. 09:37:11 This button and Calvin classic thing. 09:37:13 But then you can get some transition to like a fully kind of theories ratified and Angela it, I should say that that doesn't necessarily mean that double the music connection isn't happening in FX, you know, some would say that it's only double the physical 09:37:26 connection once it's got that stable profile and but that there is a transition, because it is time dependent as somebody said, and the same way this is a turbulent buoyancy flux for three different simulations that we ran and you can see this positive 09:37:39 buoyancy flux associated with that diffuse of conviction. And then some transition from a positive to a negative as you prefer. 09:37:47 The fourth turbulence and as you decrease the, the temperatures. 09:37:51 And the thing that we did in this paper was kind of draw on some theoretical work that we published last year in JFM to do with, and available potential energy and double the future flows Jovanka internet because it's kind of a whole, whole thing. 09:38:06 But, essentially we found that the important factor in generating event or potential energy was this upgrading and buoyancy flux in a, in a three dimensional sense, so it's kind of this three dimensional criterion for your points erupts going upgrading. 09:38:22 But of course you, you can come up with a simple one the formulation for that same statement. So you look at the buoyancy gradient in terms of density ratio and a temperature gradient and similarly with the buoyancy flux. 09:38:32 And you can kind of see that the signs of these two values is going to differ, particularly when the density ratio exactly when the density ratio is between one and compressor of a cup of tea. 09:38:45 These molecular facilities. So there's kind of this golden region where you can get an upgrade and points of flux. And really this is 3d but you know for illustrative purposes. 09:38:53 ratio has to be below cap price of a cup of tea so it's really strong gradients right at the base. 09:39:08 And then you can imagine you've got some turbulence in the far field, which could can mix up your, your flow. 09:39:14 And what we did is we quantify this using a blind see Reynolds number. 09:39:18 And, and basically we found that you've got convection, you've got this double the feast of convection, when there was a non turbulent region, where there was an upgraded and buoyancy flux that essentially was our criterion for allowing you to generate, 09:39:33 essentially, you know these blobs of of dense water that would sink and initiate connection. 09:39:56 was kind of interesting that we use a Boise Reynolds number here, and not some measure of the viscous Angela you know this isn't about how far you are from the wall, it's about how strong the stratification is in terms of dumping the turtles and. And the nice thing there is that you can 09:39:58 the nice thing there is that you can then apply that criterion to some sort of theoretical profiles beneath ice, so we took some very basic diffuse of profiles from this sort of classic paper and then plugged in our criterion, and we could create this 09:40:14 sort of regime space, which tells you how long these profiles have to defuse for before the theoretical convection ends, so this number won't be at all accurate, but the because it means that a few Civ profiles and obviously the actual profiles won't 09:40:27 be defensive. 09:40:29 And, but it works well as a way of distinguishing between convective and non conductive states. So, I won't go into that much but there are these observations and simulations of cold water I shelves which don't really show any diffuser conviction and 09:40:41 and they match up with our simulations much with crosses that she had driven and ones that have consistent conviction in circles there as are these observations of Georgia. 09:40:51 So kind of matches up quite well. 09:40:55 So yeah, some similar simulations to Maddie with maybe a potential edition of this and this regime diagram, but yeah, we're missing the layering from our simulations because we didn't start with the buoyancy gradient, that's all I'll say. 09:41:14 Thank you, Leo. 09:41:17 Are there any questions. 09:41:22 I have a question again 09:41:28 under very stable conditions. 09:41:33 You know about it. 09:41:36 experiments. 09:41:39 As a matter of formatting. 09:41:40 And the so on the very stable condition. 09:41:46 over bursting terrible and bursting. 09:41:50 Have you seen anything like this happening in your simulations in very stable conditions. 09:41:57 Yeah, so I think this gets to a big question, and I think it's one of the questions that we should be really thinking about this session is whether that model for diffuse have conducted tournaments driven by the ice, which both Mandy and I are pushing 09:42:13 that model relies on the idea, you've triggered that conviction. 09:42:18 And actually, if you, if you haven't, if you don't have that sort of conviction and very stable that as you say, simulations by virgin Island Taylor so are our co authors on on our recent paper. 09:42:30 And they did find that turbulent bursting with really strong stratification, but that's definitely I think that they fade. And it's kind of an extra part of this regimes may be considered. 09:42:39 And yeah, we're not together. 09:42:42 This is not a convection This is burst in. Yes. 09:42:59 Yeah, it's not sort of fear driven bursting but it could also have elements of convention as part of 09:42:54 this awesome given see any specter calculated showing this event also have a big effect on intermittency, if you look what happens, you know what the factors in there you can see 09:43:09 my, my simulations are really not geared for looking at the intimate and turbulent machine so, Yeah. 09:43:19 Thank you, Maurice. 09:43:30 Actually Sonia had her hand up first. Oh, I'm sorry. Okay. Sorry, go ahead, get mine is very quick question. I was just wondering Leo in yours, and in Maddie simulations, if you know you when you have the transition to a shear driven mixing regime. 09:43:46 How are you forcing the shear and your simulations, Where did the share come from. 09:43:51 Yeah, so I think in Maddie simulations it's a pressure drop across the across the simulation domain, and in my simulations it's this body forcing term in the Navy Stokes, which is sort of adjust based on the, the dissipation way. 09:44:06 Thanks. 09:44:11 Like this. Yeah, I think my questions probably pretty quick too I was curious about you have this some criterion for convection that's comparing deaths of our row and your buoyancy Reynolds number buoyancy Reynolds number is really prone to how you define 09:44:27 it especially the n squared if you have a varying varying profiles. So I'm just curious about sort of like you've got what looks like a really like nice criterion, but how sensitive is it to how you're actually defining the terms in it. 09:44:44 Um, So it didn't appear to be particularly sensitive, so I did try playing around with like a local value of the points you Reynolds or you know an average value. 09:44:52 The thing is that the stratification really gets very extreme next to the ice. So it's, it's so heavily driven by that that you know it doesn't, it's not particularly sensitive, we find them, but it's yet to be tested on observations which I think would 09:45:06 be the real test. 09:45:08 the real test. Yeah, yeah, it would be really interesting to see how well it works when you start like if you could, if you can apply this to like your George six data. 09:45:15 Yeah, yeah. 09:45:22 I heard the question you put in your trouble and regimes. 09:45:27 What was it difficult difficulty that you used in this emotions. 09:45:34 As in molecular facilities. 09:45:37 So we used accurate values of the molecular the facilities, and I didn't really explain so as as a larger the simulation, in the sense that it's kind of an implicit larger the simulation. 09:45:47 So we resolve the Common Core of scale everywhere. 09:45:51 And, but we don't resolve the bachelor scale facility everywhere we only resolve it in that boundary layer by using the strips grid. 09:45:58 So we use accurate, the facilities, which I think is important for getting the magnitude of connection. 09:46:04 And, and then sort of we're hoping that I was hoping we you know we ran a sort of a post hoc convergence tests to show you know if you include, which wave numbers, does your salt budget does your soapbox converge, and we did find that it did the same 09:46:18 with our heat flux and, which is sort of linked I think to Matty's appearance of layers, is that actually it's not particularly sensitive to those really small motions, and that you can capture the bulk motion if you can kind of capture those those blooms. 09:46:36 Okay, thank you. 09:46:39 Other additional questions for Leon before we won 09:46:47 a quick question. So when you said you were putting a body force. 09:46:52 So it was it is it, which, like, which mode where you targeting is like, was it at the highest modes or is there. Yeah we forced the largest wave numbers so we just have a cut off that a few different authors have used, you know just the first kind of 09:47:13 couple of ways number. 09:47:22 Okay. Okay. 09:47:24 Thanks, guys. 09:47:28 Right. 09:47:31 Okay, well let's move on. 09:47:34 So our next speaker is Louie this. 09:47:42 It's all yours. 09:47:44 Great, thank you. can you hear me. 09:47:47 Yes. Okay. Excellent. Well thank you all, and thanks to Leo for bringing this this discussion to my attention and suggesting a talk contribution and Facebook method. 09:47:57 So, I'm going to try share my screen 09:48:04 is 09:48:08 all assuming you can see all right. 09:48:11 But yeah, so my name is Luke we saw, I just started as an assistant professor at ns to you in France, and today I'd like to just briefly talk about the possibility to simulate simultaneously. 09:48:32 The few dynamics and the evolution of the ice ocean boundary because of nothing and freeze it, because we can get ablation or we can get accumulation of ice or water based on face centers. 09:48:40 And so I'm going to talk about the face and method to do this but obviously there are other methods, and it's still unclear whether the face method is better than others, probably depends on the situation. 09:48:52 Yeah. And so the title just reflect the fact that you're doing these simulations and you can kind of question. The natural topography of isolation boundaries. 09:49:04 So I'm going to show you how far we've come. And I hope you won't be too disappointed to know that we haven't yet investigated cases where we have layering with topographical features, appearing at the isolation interface, but hopefully we'll get there, 09:49:21 so it's been very much a collaborative effort, and the most recent progress you can find in those three papers. 09:49:29 And it's been a contribution from a lot of people in particular by Eric Hester who is finishing off his PhD at the University of Sydney, and he will be moving to the west coast I believe very soon. 09:49:43 Alright so let's get started so. 09:49:47 So the first observation is that the isolation boundaries free okay so it's free to move because we have faced challenges. So one kind of natural question is what should the isolation binary look like why it's melting and I'm going to be interested, mostly 09:50:00 in boundaries when they're melting, on average, and I'll be looking at cases of ourselves, which was, I think, brought up by Leo, So you basically have ice overlaying water. 09:50:14 So there are free cases right you can have a sort of a flat ice water interface is this is shown on the left, and I'm using blue colors for both the ice in the water just to reflect the fact that it's, it's basically h2o molecules in both cases, it's 09:50:28 kind of the same component. 09:50:31 So there is the first option where you do have a flat ice water interface, then you have a second option where it's very rough interface, meaning that the landscape, typical took graphical lens skills are very short compared to say the depth of freedom. 09:50:45 And then you can have just took a graphical features so these are sort of more coherent, they can have larger scales, such as maybe on the scale of water depth, or 10s of meters kilometers. 09:51:00 Land scales. 09:51:01 And we've observed these type of graphical features that isolation boundaries, and they're very important from a fundamental standpoint because if you have topography, you can impact heat flux extraction from from from the fleet by Zi spice layer of ice 09:51:17 and so you can impact the ice, the mean I smell trade. And so you can see these through the sort of parameter eyes heat flux is q super script, l, which I wrote was like a bunch of parameters and then it's times, so it's time so the you, which is say 09:51:36 the mean feat velocity at a few meters from the boundary, and then times the temperature difference between the temperature in the bulk of the water, and the temperature at the bottom right so this is a parameter ization and you need to kind of understand 09:51:50 how you should change your heat transfer coefficient, such as gamma t, and also your drag coefficient, and these two coefficients, they depend on things such as Are you having layering, but also on the topography of the ice water boundary. 09:52:06 So, so here's the question, what should the ice water boundary look like when it is melting. 09:52:14 So, if you want to kind of look at the evolution of the ice water boundary. 09:52:19 The kind of natural way to look at this problem is to think that you have two domains. One is eyes, so it's kind of like solid mechanics. One is water. 09:52:28 So it's kind of three mechanics, and then you have a discontinue space James Bond or between the two. But this is really challenging numerically to have this kind of two domains evolving differently following different equations and was a coupling through 09:52:45 boundary condition which is itself moving in time and space. So one alternative to this is to say, Wait, this is actually water throughout safe assault maybe but we can talk about this later. 09:52:53 So how about we just say that this is just like water everywhere, so we can define the temperature for the domain, but the ice domain but and in the in the liquid domain. 09:53:04 But then we use a marker right so it's kind of we augment the systems, it's all of the variables of interest are defined throughout, but then you have this marker sigh or feet pending or your front, which takes the value of say zero issue in the eyes 09:53:28 and one during the liquid. 09:53:20 And so, so this marker What does it look like well the bottom left panel shows, it's a it's a vertical slice of assimilation, and so it's showing temperature and the blue colors, highlight temperature in the eyes and the red colors highlight temperature 09:53:38 in the industry in the liquid. 09:53:40 And so the temperature field is the depth field is defined everywhere but where you can kind of see these dotted line in the middle, we actually have a transition from the liquid layer at the bottom to the solid layer at the top. 09:53:56 And so the marker will transition from 123 to zero in the solid for very thin diffuse transition layer for this market right so it's kind of a smooth so we smooth. 09:54:12 The problem we don't have a discontinuous boundary, but we have a smooth transition boundary from the solid, liquid. Otherwise we're just solving the kind of manual steps equations, we some terms to represent the Latin heat source in the temperature equation 09:54:22 so this is terrible doing here. But all of the black trans or just kind of dimension is not your stocks equations. 09:54:28 So we have temperature heat so we have heat sink or sources depending on face changes this is when the face and viable phase changing. 09:54:37 And we also have damping of philosophy and the face variable evolves according to an equation of its own. I'll just say that this face field formulation equations, they actually converge to the true to domain equations method which is kind of a natural 09:54:54 way to think about it, and to the true Stephens boundary condition at the interface. When this transition layer epsilon becomes infinitely small right so this is kind of a, of a, of a consistent from an aesthetic standpoint, formulation of the ice water 09:55:09 boundary problem. 09:55:11 Okay now I just want to show you some, some results, which we've had which we've obtained. So this is a kind of sort of similar to Matty's paper where you get these share driven flow it's a pressure driven flow, it's coming from the far back of the figure 09:55:28 toward the front of the figure. And so we have a pressure driven flow, however, it doesn't have any salt. 09:55:35 So, so yeah so it's just temperature flow, but we've tested, because we can do this in simulations different kinds of stratification because we're imposing a temperature difference across the domain. 09:55:47 So we can have a temperature stratification that leads to a stable density stratification natural certification so temperature doesn't have any effect and buoyancy or unstable. 09:55:57 So we're playing around with density fuel basically. And we've looked at what kind of took a graphical features naturally emerged from these interactions between turbulence 3d turbulence and, and heterogeneous space changes that the binary. 09:56:10 And so in all of these cases we've got these two paragraphs features so they all kind of channels. 09:56:17 So these channels and keels are always aligned with the main direction of the flow. So they're actually not really changing the drag coefficient of the ice water boundary, but you can see that in these cases, regardless of the stratification you're considering, 09:56:33 you're going to get naturally some topographical features, which in this case, do not affect the meanness right, but in other cases. 09:56:41 This was very limited to small children's parameters, you could actually get some cases where you're changing the mean matter right by allowing the interface to change. 09:56:52 Okay, so the last case i just want to show is taken from a paper by led by Eric Hester. So this is a case of an iceberg melting in salt water. 09:57:03 So these are two sets of figures showing temperature on the left, and salt on the right. 09:57:09 But there's one, and so it's salt water right so you get sold in your temperature, but we've used a limited number, which is a ratio of the thermal dp CVT to sort the facility, which is equal to one, and we've used initial conditions such that the temperature 09:57:27 is uniform 20 degrees throughout the domain and the salt is also the concentrations field is equal to 30 grand per kilogram, also for the domain. So basically you have two components, but they actually have some same to see VCs, and the same initial distribution. 09:57:42 So it's actually just a single device using simulation, right. So, but we can still look at both fields, they look exactly the same. And what we're seeing in this case is that the the meltwater the iceberg is in black. 09:57:56 The meltwater, which is kind of seen by the fact that it's cold so it's kind of purple on the left column is rising so it's pulling at the top. So if you're looking to the side of the iceberg. 09:58:06 You can see that there is a thickening of thermal boundary layer at the top. Of course we do have some plumes plunging to the bottom, because we have a reset escalation. 09:58:18 So meltwater, which is generated at the bottom of the iceberg is rising on the side of the iceberg and then putting at the top, and then we get this recirculation. 09:58:27 And so the concentration, just the same. 09:58:30 Okay, now let's look at the case where the teachers are different. This is when we can maybe think about, well done with the physical interaction surely but possibly layering although not in this case. 09:58:41 So this is a case where we have a limited number which is close to 10 right so so temperatures defusing factors and salt, but it's not as massive as what you obtained the real ocean, and the results are qualitatively different from the, from the previous 09:58:55 case, because you still do get these kind of pooling of meltwater at the top, but now we're also getting plumes going sinking to the bottom of the domain. 09:59:05 So, we were speculating so I mean we haven't. We've done some analysis but I don't know how to what extent you can prove that this is why the conduction, but the plumes you don't get when you get when you have the Louis number equal to one. 09:59:19 And you get sinking plumes beneath the iceberg when you increase the deliveries number different to one right so I think the singing blooms our signatures of Wi Fi connection and they do change significantly as a metric of the iceberg. 09:59:32 So these are things to take into account. And what's interesting is that the does sinking plumes were also observed in experiments because Eric's work was a combination of simulations and experiments. 09:59:48 And so for this case of Louis greater or equal to 10 something like this. 09:59:52 So concentration field looks a bit different from the temperature field. 09:59:56 It's essentially the scale, discuss or smaller and that's because salty qzV Solar's and temperature diffusing the team. 10:00:05 Alright, so just to conclude we some wrap up with some questions. So, in terms of what the face field simulations can bring to the community and the kind of question that could help address, I think our well first due to progressive physical features 10:00:20 spontaneously emerging issues. First, I'm not sure if this has ever been observed, and this has not been seen in simulations because there are no simulations, attempting to look at WC flows with face changes resolved and going along with this question 10:00:37 there's a question of can the natural typography, if it's not flat of ice water interfaces modifies the WCC dynamics and possibly layering. 10:00:48 There are some, some questions myself, which is why is it so hard to do experiments of WCC convection, because it's so hard to do these things in America that would just be created experiments. 10:00:59 And then, this is something that Leo Burnett but do we really need an initial certification for layering somehow was wondering. 10:01:06 Yeah, but thank you for your attention. 10:01:12 Thank you. Thank you. 10:01:15 Questions for Louis. 10:01:22 I'll ask, I'll start with one. 10:01:26 So when you showed the sinking plumes. 10:01:30 You mentioned that it might be related to double dishes have conduction Can you elaborate a little bit on that. 10:01:38 Sure yeah so I think that the best way to look at this, is that, in both cases we do expect that the thermal expansion of the thermal expansion coefficient is going to be is going to be positive. 10:01:55 So, the. 10:02:10 The fresh water will always be more buoyant. 10:02:10 Then, ambient water, which is satellite. So we should get in both cases, water just putting at the top, and big enter this is what we're seeing in the simulations, where is the lives number is equal to one. 10:02:22 But when we change the feasibility of salt and temperature, this is when we can get another kind of dynamics, where we get the plume sinking, but to get this really we need to have the temperature to choose and faster so that then we can have the cold 10:02:38 water, which is becoming sideline. So the sideline water rights is kind of, you have a layer of meltwater and beneath the ambient water is cooling fosters and it's freshening. 10:02:52 So this is giving you a second layer of cooling ambient water, which this times is dense and this is why you can get sinking, but to get this you really need to have to temperature to choosing faster, the temperature normally diffusing faster into the 10:03:18 than the salt because otherwise and you're stuck with just all of these kind of fresh mouth water going up. Okay, i think i think my. What I meant is just looking at this blooms can we can be characterized as well I assume that later on, we would see 10:03:27 the layers right originating from this mixing, and then looking at the blooms Can we just speculate on the characteristics of the layers like thickness, or. 10:03:39 Yeah, for example thickness. 10:03:40 Yeah, so I mean this is a great question and I think we haven't gone this far, so I'll be unable to answer your question, but yeah it's a question of, just by running these simulations for longer time. 10:03:57 I don't think we will get the layering, but then Leo brought up this idea that if you have you kind of need, initial stratification. 10:04:13 We will using ambient temperature uniform ambient temperature and uniform salt concentration. So it's an option to maybe just look at what happens if we get some sort of a leaner stratification but I would also suggest people want to try this, to not 10:04:23 have an iceberg, but rather have a just a uniform layer of slab because the side effects of the iceberg can induce some sort of Alaska circulation, which would probably be entirely disrupt the layering. 10:04:35 I don't know if this is, I guess, maybe some other people in the room knows this better than I do, but be interesting to test. 10:04:52 Thank you. 10:04:47 Um, I see two more questions. One is from Chris. 10:04:49 Hold on. 10:04:50 Yeah, thanks very great talk. At first, I had a question about sort of the double defensive implementation of the faith field because I, I know it works quite well for the melting dissolving side of things, but can we expected to represent the solidification 10:05:13 the freezing as well. 10:05:16 or is it, unable to 10:05:20 resolve these sort of mushy layers that you get from the small scale Brian rejection. 10:05:26 and then a follow up to that would be like, Is there a hope that we could sort of incorporate a model for that into the phase field and have sort of a wonder code that could do ice melting and the solidification and dissolving and everything. 10:05:41 Yeah, so for for freezing when you just have a temperature flow so a fresh water flow it's not really an issue you can do freezing just as well. When you do get so there is a question of do you get a National Air, which is, in other words, can you get 10:05:59 actually salt trapped in the ice while you're freezing, and then you need to well have a nice day or dancing in your model. And so we haven't done this so these simulations have salt and temperature but obviously we're not trying to do anything mushy 10:06:15 like so we are basically rejecting all of the salt. Anytime we're freezing the validity of this, I suppose, should be investigated early and we haven't done it with just assume this. 10:06:32 I think it's possible that if you have a fairly trivial and slow over or you will be you're flowing be efficient at removing salt, so it would be easier to reject salt from the freezing medium. 10:06:49 I think these kind of measure layers dynamics, and all of the roughness, you can get right under sea ice for instance. All of these things are very much related to measure dynamics. 10:07:01 I wonder whether maybe they're happening and they're so prevalent because the water underneath is otherwise kind of question so comes right so no external flow. 10:07:14 We are of the opinion that if you get kind of trouble and dynamics and you don't have to worry so much about salt being trapped because it's just rejected very efficiently but yeah I think this is an open question and we have not addressed it, but it's 10:07:28 probably it's just to to your to your last point I think it's possible i mean i think there is reason to hope that we can also have a national aerodynamics because this, we could just have another marker right so we have five for ice and water and liquid 10:07:45 water but then we could have another marker, based on the salt concentration which, if you allow it not to be entirely zero in the social space. 10:07:55 I think it'd be possible. 10:08:01 Alright. And the next question is from David. 10:08:06 Thank you, which is kind of similar question in a way. 10:08:20 Very nice method that you have it reminded me a bit of the talks we were having on con Elliot now via scopes, where, you know, you've got a phase equation. But it's a different phase equation of course but it's coupled in in just the same way. 10:08:26 So I wondered how much freedom you had in your foot, you kind of wrote it down in a fairly general form, I wondered how much freedom you had within that phase equation and how 10:08:39 you know susceptible your results are to how dependent your results are on that on that equation. 10:08:45 When did I wonder really how you chose your face equation, I guess. Just a nice way of doing things. Yes, so right so going back here yes as a key question here is kind of Queen equation for the face you about viable so there have been different formulation 10:09:02 so I think you can have. And this is what we're using initially you can have firmer dynamically consistent equations for the evolution of this face your variable which means that you start from the info p or and the entropy of the three then, and you 10:09:19 do right the volume conservation equations and. 10:09:24 And so everything can be not very neat from a from a dynamic principle but, and then in this case you're very much constrained about the kind of faithful equation you end up with the, the faithful equation we've been using. 10:09:40 Over the past two years, is slightly more gentle so we don't require to start from, from a dynamic principles. 10:09:58 We're using some more. 10:09:52 Some more free arbitrary formulations. But then we're thinking about the awesome topic so it means that we're looking at the terms from the, so it's kind of, we're looking at arrayed in a way of much esoteric expansions We are the four fields from the 10:10:08 boundary. And then we have the, the new boundary domain, and in the far field we're requiring the equations to be as close as possible to the wx equations for the free, and to just heat equation for some solid, if that's all you care about. 10:10:24 And then we do have some topics, based on these epsilon parameter, and then we look at the leading other terms in the near near boundary domain, and then we we try to cancel as much as possible. 10:10:40 And basically, you can cancel more terms, using general formulation of the facial method. 10:10:47 Then you should use it from a dynamic equations because it from a dynamic equations will have leading other terms which are afforded epsilon. If you do, if you use a more general formulation, then you can remove also second order terms. 10:11:01 So you go to second order. 10:11:05 And so I think you know where you could consider is that you can use pretty much anything provided that you're checking the accuracy of your method against the true Stephens boundary conditions, and bulk equations for the freedom to solid. 10:11:21 And this is discussing the proceedings of the Royal Society paper. These are some topics methods. 10:11:31 Okay, thank you. 10:11:33 Alright, so we have, I think four more questions. I'm Sam had one but I don't see the icon anymore. 10:11:45 I did have one, if there's time. 10:11:49 I was, I was just curious in that iceberg problem. 10:11:56 At the end, you have this diffuse of W diffuser convection happening, but is that a significant effect. 10:12:04 I'm not aware of, sort of people thinking about iceberg melting with that as a significant effect because you've got you've got sort of a certain amount of flux of fluid flowing along horizontally and vertically, but does that tend to dominate in the 10:12:20 problem, or is that is that fingering actually quite a large flux. 10:12:27 So are you asking compared to Susan meltwater dynamics compared to excel forcing or are you asking specifically about 10:12:38 WTQZ effects compared to just upwelling on the sides. Yeah, exactly. Yeah, so just the problem that you've got here. 10:12:49 Yeah, so for this problem it's it's fairly different I didn't collect quantitative data for this, which compares melting. But also, it's, it's, it may be because we have a finite domain right so we, so it does affect eventually is a large scale circulation 10:13:04 within the box. 10:13:09 the box. I would imagine that otherwise just the upwelling of the meltwater would probably dominate the now trait of the iceberg. 10:13:22 Yeah, the effect of double diffuse effects here the impact of it on the overall melting of the iceberg in in real life, I don't know, but in this box it does matter. 10:13:33 Okay, thanks. 10:13:36 Alright. 10:13:37 So we have three more questions we're way behind schedule. 10:13:40 But I suppose it's okay because we already have a discussion. So let's take these two questions and then maybe go for a short break. 10:13:47 And then, yeah, I'll, I'll talk quickly on the observational work so the next question is from David. 10:13:58 Please God know if that's me it's just because I haven't taken my hand down. 10:14:03 Oh, sorry. Okay, I'm at Gar. 10:14:08 Um, yeah so I just had a quick question about the topography that you showed on the underside of the, of the ice. So there was a very regular pattern of cups cup like structures, and I was wondering whether you know what determines the length scale of 10:14:26 that typography and whether it's related to the some fresh water plumes that you were describing in your simulations. 10:14:36 Right okay so so you're talking about this topography which I put on this phone. 10:14:41 Well it was an actually an observation, I was an observation right okay so you're talking about this one. 10:14:47 Yeah, right. Okay, yes so yes this this is taken from an experiment done by Mitch Kaushik. 10:14:55 So these are scholars. So, these are three dimensional features and we have not yet observed them in numerical simulations. 10:15:05 I think the lens skills of these features have not yet been related to a particular. 10:15:14 Well I've not been clearly defined. What I can tell you about this is that it's been speculated, but it's been only speculated because it was really hard to do observation of experiments so we don't have all of the data, we want. 10:15:27 And there is no simulations yet of it, but it's been speculated that basically vortices, which are prevalent in trouble and share flows so fairly small scale vortices can be trapped within the trucks of the scallops. 10:15:42 And then they can try to enhance melting in the truck so it's kind of a self reinforcing mechanism when if you start having a very small topographical feature just because even more. 10:15:55 By default, this is being trapped. 10:15:58 So this is a landscape would then be related to the typical landscape of India was vortices three dimensional vortices, but I could not tell you more about what set their mythology yet. 10:16:14 Yeah. It's something we're pushing we're trying to understand better. Yeah. 10:16:19 Alright. 10:16:21 And the last question is from Martin. 10:16:24 Yeah, okay, just switching on. 10:16:27 So maybe more technical question regarding the face field is the audio coming through. Yeah. 10:16:35 OK, so the face field I mean you can go back to the equation, this equation that you that 10:16:43 we had the epsilon parameter right and this is kind of the width of the success of this jumped out and see excellent do you maintain the spelling of epsilon or is it kind of also very in in the concept of a simulation Can you comment on it. 10:16:57 Yes, excellent parameters of six parameter and the goal is to take it as small as possible because like you've said this is the final week of the transition so when we're changing from liquid for two solid, and and and you realize this should be like 10:17:10 molecular scale. 10:17:14 But we're constrained by numerical simulations right by the resolution so basically our attempt. What we try to do is to have it so small that it is smaller basic fees and the smallest natural physical landscape system so when you close as a boundary 10:17:30 you're thinking, this could separate sickness or discuss, or form of severe fitness, or sorts of layer thickness. And because because of the physicality so small, to solve sadly or fitness is extremely small six, it's really challenging than to have epsilon 10:17:48 epsilon smaller than this, because you still need to resolve this epsilon right the width of a which fees changing. You still need to resolve this landscape. 10:17:58 So we're talking about yeah it's it's really it's computationally very costly So, so we're, we're definitely not going to try doing simulations at real parameters values. 10:18:21 But I was wondering for everyone I've seen it said you can derive it from first principles for me to creation and to ever go in the direction of aesthetics. What was my point so this is kind of follows porous boundaries was I was kind of thinking I mean 10:18:25 if you say it's small enough, then it's maybe it's yes yeah you're absolutely right, this is kind of a porous Darcy Lowes equivalent kind of formulation for ice water systems absolutely yes this is a great intuition to have Yeah. 10:18:41 Okay, thank you very much. 10:18:43 Thanks so much for all the Christians everybody. 10:18:48 So as we move on to the with our show breaking. 10:18:51 Yes, I would suggest to take a five minute break and then we can move on and I'll briefly talk about observational work and for the next for the rest of the discussion let's session, let's just focus on these three questions that Bruce posted on slack. 10:19:12 And just formulate those questions, discuss what skills. The group has and but how we can move forward, forward with them. 10:19:21 Sounds good. 10:26:30 So I suppose everybody's back. 10:26:46 I'm back. Yeah, we do. Okay, great. 10:26:39 So, um, well, as I said before, before the break we are not following the schedule but I think it's good we were asked to be flexible, so we're following that suggestion, and so far we had great talks and great questions. 10:26:56 So, I suppose, unless there are other announcements will start sharing my screen, and we'll talk about observations in the Southern Ocean. 10:27:10 All good, Leo. Okay. 10:27:14 Okay, fantastic. 10:27:27 Oh, okay. 10:27:31 I suppose it's different. Sorry, just one moment. 10:27:45 Okay, here we go. 10:27:46 Do you see my screen. 10:27:50 Yeah, I think so, um, yeah the regulation of ci sickness by double digits of processes and the Ross Shire. 10:27:59 So here you see the. 10:28:02 This to schematics the right schematic shows the general circulation of the Southern Ocean highlighting to gyro structure to gyrus, there was char and Wardell gyro together with ACC goes around Antarctica. 10:28:21 And on the left, you see a three d schematic of the overturning circulation. 10:28:26 So by callers, the authors showed different What am I says Antarctic bottled water. 10:28:33 Lower and upper circumpolar deepwater and intermediate water. 10:28:51 So the area of research for us was the Ross gyre, and these region the central central rose giant and in the region where we have see is the main question was to figure out what controls, she is thickness. 10:28:56 In the Ross gyre. 10:28:58 The typical winter profiles from the Ross gyre are shown here. So this is potential temperature, this is salinity. 10:29:11 And what you see right away is that in winter months. 10:29:16 Mixed layer is very deep almost hundred meters, followed by stronger certification, and then we enter upper circumpolar deep water that is warmer and sold here, then they make slam. 10:29:32 The study in this study we used mainly Argo observations and some data delivered by custom build floats we call them MRV floats, and they operate similarly to Argo, but have much finer resolution, and the temporal resolution can be sad according to different 10:29:56 missions, we had today temporal resolution, and they were deployed in the central it was January so on this map, you see Argo coverage and Emery floats were deployed in three locations in the central portion of the, of the gyrus here, the jar is marked 10:30:14 by yellow contour. 10:30:17 So what we see right away is that right below the surface mixed layer. 10:30:23 We have staircase structures. 10:30:40 These staircase structures are robust in winter months. And in summer months, they are regularly disrupted by supposedly troubled and mixing. 10:30:38 So these are some characteristics of the layers, layers thickness varies from three to seven meters. 10:30:46 Temperature jobs are points three degrees Celsius to put and salinity point or five lateral coherence of the straight case mix layers is not observed and like these staircases in in the Arctic Ocean. 10:31:03 And as I mentioned before in winter, the structures are more robust. 10:31:08 So in our study, we examined, two different scenarios doubled issues of scenario in the presence of week, share, and troubling scenario. 10:31:21 So, in double defusing of scenario, we assumed that DVD release for heat and salt are not equal in trouble and scenario. 10:31:32 We assumed that sold and heated transported at the same rate that is equal to travel and the futility. 10:31:41 So, for double diffuse of scenario. 10:31:46 While we can envision it as a mix layer followed by staircases right and then in trouble and scenario where assume that the layers are smeared out by the background trouble mixing. 10:32:01 So, just looking at the observations. We already kind of know that double diffuse if scenario is present in the central Ross Shire under, she is. 10:32:14 So, these, we will we quantified a vertical heat flux is at two different scenarios double diffuser from trouble and, and we just used to standard prioritization for a vertical heat flux is for the feasibility. 10:32:34 We used the parameter ization developed by Flanagan. 10:32:40 That did their analysis and the Adele, and with LG fire. 10:32:46 And what we. Oh yeah, and then VCs our 1d model setup. 10:32:56 So to understand the fact of these two different mixing processes we followed the surf winter surface mix lamp buoyancy budget by partisan 1990. 10:33:12 So here, for mixed lay hit budget we've got atmospheric heat or heat flux out of the mix layer so we call it atmospheric cooling. 10:33:23 And then two terms for heat flux is coming from beneath of the mix layer in treatment heat flux and diffuse of heat flux that can be either double diffuser, or purely trouble and according Ciao, two different scenarios 10:33:45 for the salt budget. 10:33:48 We've got Brian rejection, and also to fluxes from the base of the mix layer in treatment flux and diffuser flux. 10:34:00 Now for the deeper portions. 10:34:05 We just have what we computed the evolution of temperature and salinity, looking at the convergence and divergence of heat flux and we also took into account Ekman pumping. 10:34:23 So I'll just show the wall some results from these simple one, D model modeling, together with observations. So here, a Showtime series from June 1 till the end of October. 10:34:42 And this is our winter months. 10:34:45 And the upper panel shows mix less salinity middle one shows thickness of the mix layer and then the bottom panel shows thickness of ice, we don't have. 10:35:03 Well Institute observations of the sea ice thickness, but from the previous studies we know that CIC thickness. 10:35:12 In winter time in the Rose giant is roughly between point five meters and point eight meters. So what I'll one the numerical experiment tells us is that makes less the linens and mixed layer thickness fall within the observations for two different scenarios 10:35:34 solid line is double diffusers case, when we used double diffuser parameter ization for GPU DVDs and dashed line shows the results for troubling case. 10:35:46 So, they both fall within this the the observations. But for ci thickness. 10:35:55 trouble in case significantly aren't estimates at. 10:35:59 So now the question is, what are the processes that might limit the background trivalent levels in winter, and allow a staircase to forum. 10:36:10 And we proposed a ice production and treatment feedback mechanism. So as you as the winter starts. 10:36:21 There is some sheer in there in the shallow regions and through Thermohaline sheer instability that was reasonably well, couple of years ago proposed by team or radical, 10:36:35 we form a double diffuse you have layers. 10:36:41 And so that is the big producers double diffuser staircases next as the winter progresses. 10:36:49 Double diffuser staircases may act as a filter for internal driven trouble and internal internal waves right so the presence of doubled issue staircases limit the internal wave field, and therefore limit background, internal wave driven trouble and levels. 10:37:16 This allows. 10:37:19 She is to grow right so we don't have vigorous trouble and mixing. We don't have heat flux is strong hip flexors that come from upper circumpolar deep water into the mix layer. 10:37:32 So, 10:37:34 double defuses straight case acts as a filtering. 10:37:38 As a filter, and that allows for CIOs to forum. 10:37:42 So now we have see eyes, and we propose that it may act as a buffer and further reduce internal wave activity. So there were some studies in the, in the Arctic Ocean showing that the presence of sea ice reduces troubling levels in the, in, in the interior. 10:38:06 And these further enhances. 10:38:12 Well formation and evolution have doubled issues of staircases and then it generates more size, and this loop closes. 10:38:23 So that's a proposed feedback mechanism to sort of explain the presence of double diffuser staircases underneath the mix layer in winter months. 10:38:40 But it would be very nice to test it, either with further observations, or, or model studies. 10:38:49 Of course I didn't go through the details of this study, you could take a look at the paper. It was published a few years ago, but I'll just go through main conclusions. 10:39:01 So, it was the first study that showed the presence of double diffuser staircases under the surface mixed layer in in the Ross Shire in winter months. 10:39:13 In the study we showed that the presence of double the field of straight cases enhances cis formation by suppressing vertical heat flux is from upper circumpolar deepwater, and perhaps there is this particular feedback mechanism. 10:39:28 And we call it ice production in treatment feedback mechanism. 10:39:33 And for the group for the discussion I suppose, 10:39:39 I would like to test this hypothesis. 10:39:46 Somehow, either through further observational studies or modeling studies. 10:39:53 Thank you. 10:39:55 Great, thanks for those really interesting. 10:39:57 And if anyone has any questions, please raise your hands Oh we got one from Louis 10:40:04 Right. 10:40:05 Yeah, this is a great study thank you yeah, I'm wondering because I'm really interesting. 10:40:11 With the CI subjects. Do you think that Brian rejection is kind of key to your to the different mechanisms you've proposed to get some layering or could we just have a nice shelf with no Brian and maybe the mechanisms will still apply. 10:40:29 Right. Um, so Brian rejection, in our setting, just x. 10:40:36 To deepen the surface mix layer, and it causes the entrainment we call entrainment heat flux on salt Fox from the layers below into the mix layer. 10:40:50 Now, the double diffuser hip flexors. We're assuming you know, wanting model that they operate, sort of, 10:41:01 on their own. 10:41:04 So in this way. 10:41:09 They just operate together. I don't think that's Brian rejection. 10:41:17 Yeah, which is included in, you know, feedback loop. 10:41:24 So it primarily controls the thickness of the surface next layer right yeah in this setting right, yes. 10:41:30 Yeah. 10:41:33 And can we hear from Tanya. 10:41:37 So I thought it was an interesting hypothesis that developing layers could suppress internal way of activity. And I was wondering if there is any, any evidence to back that up. 10:41:51 And because I was thinking that, you know, we do have modes of internal waves that exist in layered systems. So just the fact that you have layering doesn't automatically kill off internal waves and but it would depend on the vertical what mode number 10:42:07 and so on of the internal way so I was wondering if there's any, any observational lab or numerical experiment evidence that lead systems tend to suppress the internal wave activity. 10:42:21 Right. Um, well, I think maybe Bruce can comment on that. So we were just. 10:42:33 I think building on the Bruce's 2016 paper. 10:42:40 I personally don't know if there are any observations of that. 10:42:46 This is just more. 10:42:48 Suppose mathematical study right. 10:42:53 I guess I'll weigh in. 10:42:55 The work it was purely theoretical. 10:42:58 But it did pace a limit, based on the layer step size, and how many layers you have of what size of internal ways can actually make it through the staircase, or be reflected from the staircase. 10:43:13 but it's, it was that again a highly idealized study so far from real observations on it. 10:43:20 It was again a highly idealized study so far from real observations on it. I think it's a fascinating idea to have this coupled interaction between the two between double diffusion and a double diffuser staircases and then turn away so I hope, yeah. 10:43:30 People look at it more. 10:43:32 so that that that would be my major I think question for the group. 10:43:37 So just think about guy. Can I see something on that point. 10:43:41 So, not in the context of ci smelting, but we have been doing some simulations in our group about oil and oil blooms in stratified environments, and what we have found is, at least in one of the simulations, what I've been observing is that is formation 10:44:01 of these institution layers as the oil goes up. 10:44:07 And then, in the case when these intuition layers was not there safe, in case of a thermal. 10:44:11 We had this beautiful internal gravity waves, those structures, seems to vanish. Now, we don't know. I haven't analyzed the data enough to say that if the internal gravity waves are completely killed or not. 10:44:25 So these are DNS simulations we are talking I'm talking about. 10:44:29 But, I would. Now, this gives me an idea to additionally look into what you guys are saying that. 10:44:36 Is it too but in the velocity structure we are seeing in, and I can, I can try to pull up a finger and share it in the slack. 10:44:45 What we found is that it seems the, the internal grab, at least the structure of the internal gravity wave this 45 degree angle, we see that thing is, isn't there. 10:44:55 Now once the we have these, these intrusions coming up so I don't know if it exactly is what you guys are talking about but it seems similar. 10:45:07 We can bring in Alexis you got your hands interesting yeah, I'm just following up on on Bruce is sort of comment on Sundays question, Tim Arad coaches published a JFM paper in the last like three or four months that was doing some pretty rigorous and 10:45:22 topics talking about internally it's like large scale internalized interacting with staircases that also suggested for both for both sort of versions of double diffuse of unstable systems that that the wave decays because of the staircase. 10:45:37 And then I also think McHale had had his nice poster set up in the in the gathers setting, looking at some, some internal wave staircase interactions to he might have thoughts on this. 10:45:49 There was also some laboratory studies out of Tom peacocks lab, a few years ago from CSN games he would have been the lead author and I think they also saw decay for smaller scale internalized propagating through staircases sort of growing evidence I 10:46:03 I think overall that supports this idea that that interaction, at least in some regimes does decay internally of energy. 10:46:12 That's great. Thank you Alex, would you mind putting these references here in chat and I suppose the plan is to just copy this chat to slack. After this, I need to dig them up so I'll post some random slack. 10:46:33 All right, thank you so much access. 10:46:34 That's great. We've got a question from Mikhail or thanks I just wanted to mention that yeah so I do have a poster on my work I'm doing some simulations of internal waves, going through these layers and I just want to mention that I have also been looking 10:46:46 for papers of observations of this phenomena and I think the only things that I've really been able to see our observations of a storm that happened in August 2012 in the Arctic. 10:47:03 And there it was mainly just observations of general how active the internal wave fields are in that mix layer versus below it. And it seems like even if there is this storm which creates our higher internal wave energy above, then you don't really see 10:47:19 more internal energy wave activity below the picnic line. 10:47:40 But I will also say like something that I've seen a lot in my simulations, especially with few numbers of steps, is that if you have a certain match between the wave number of the wave, compared to the thickness of the layers you might actually have a 10:47:46 lot more transmission than you might expect. It's kind of a tunneling effect. 10:47:59 But I get to figure out whether or not that still holds when you have these staircases with dozens of steps. That's sweeping she actually Michael I have a quick question on the paper you referenced this Arctic paper, or they showed no difference. 10:48:22 So, I, my impression was that those staircases are kind of shield from 10:48:21 from basically the trouble and ocean that is way above it. Is that is that is that the one you're referencing here. Um, yeah so I'm looking at two different ones I think you referenced the rainbow and Woodgate people talked about it but then there's another 10:48:36 one which doesn't necessarily talk a lot about internal waves, but it's by Simmons and Richard diva called the Great Arctic Cyclone of August 2012. 10:48:50 And they mostly just talk about, like, activity on it is very, it's not specifically talking about this staircases. Okay. 10:49:00 Yeah. The reason I brought it up is that in dark secret case it's a little different, I suppose, because the steps are much further. 10:49:09 So, they might be just shield from all the atmospheric forcing that's happening at the isolation interface. 10:49:22 I social interface. Yeah, yeah, it is also different because they generally in the Arctic these are present all year round, as opposed to, I think, in the Southern Ocean their seasonal there Yeah, Yeah, yeah. 10:49:36 I mean, yeah. So I sort of wanted to link this back into maybe some of the the the chat we had at the beginning, and I was wondering, you talked about that, that upper layer and sort of think about these double diffuser boxes. 10:49:47 so it in your model is it that the mix layer is always a mix layer forced by by sharing and then, then the double diffusion happens, lower. 10:49:59 And then, to that. Is it possible to have, You know, that whole. 10:50:05 Basically I'm wondering if it would be possible to have that as a double the piece of collectively forced mix let, Paul, or whether that's going to be on the realms of. 10:50:15 Right. Um, well, of course, we considered just the most simplified model. And the next layer was maintained by, 10:50:32 by, basically. 10:50:34 Salt reduction in in winter months and grew. 10:50:39 Because of that, we just used constant temperature of the next layer throughout the year. Just freezing temperature. 10:50:50 And it's certainly a limitation. 10:50:52 And we use constant atmospheric heat flexes. 10:50:57 Those heat flux is out of the, of the mix layer. 10:51:00 So I think there are many ways to 10:51:07 sort of make this model more sophisticated and maybe the next layer can be represented by the way you suggested here earlier. Yes. 10:51:16 Yeah, 10:51:19 it's I think it's interesting in the it's, you kind of, you have these modeling papers like Maddie is where you kind of, you have this really quite strong diffuse of conviction. 10:51:29 And I'm kind of wondering why it's not been something we've seen so much especially if you know there's potential for these feedback mechanisms such as you've demonstrated like my sense with myself observations is there are so few but you know anything 10:51:44 could be true in terms of which one of them is representative is the same true of these sorts of observations or are they much 10:51:53 human, whether the next layer is whether you get these staircases for instance, you know like, how how frequently you see those in the data and is there, how much data is there. 10:52:14 Yes, that's that's a good point there is not much data with the proper resolutions to correct race straight cases so there are some floats, or go floats in, and I'm talking about the Ross gyrus specifically in the middle Jr. 10:52:23 I might not know some existing find scout resolution data. 10:52:27 And there are occasional I suppose microstructure measurements in these regions. 10:52:33 So, in that sense, I'll MRV floats played a big role in discovering these layers, because they're not microstructure data but they're also. 10:52:42 They also have much final resolution that then Argo floats. 10:52:48 So there is not much data. 10:52:51 Although the data that we have are going to show my screen and and just share my screen and show you the sequence of of profiles. 10:53:03 Just give me one moment. 10:53:06 And, and in winter, for winter profiles. 10:53:10 We see layers presents and pretty much all of them. 10:53:15 Yeah. Just give me one moment. 10:53:18 I'll show it to you. 10:53:21 Okay, here we go. 10:53:31 Okay so here I show a sequence of temperature profiles in the top, and then salinity profiles at the bottom. And, 10:53:43 yeah, well maybe the scale of the figure isn't good enough but you kind of can see that pretty much all profiles have these layers. 10:53:52 It's very noticeable that these layers are not coherent. 10:53:56 Unlike the Arctic case, when an individual layer can be tracked for hundreds, hundreds of kilometers here. 10:54:04 We do not observe that and that might be because of constant merging and splitting events associated perhaps with intimidated troubling mixing. 10:54:15 Great, thanks very much. and we've got a question from Martin. 10:54:20 Yeah, let me just unmute, Okay. 10:54:23 I think that there's two points I want to want to mention. So when concerns or questions that you were just discussing. 10:54:30 So, what is feasible in your area so far can you I mean how what extent of the next day are you able to to model to simulate, actually, like how many meters. 10:54:40 Awesome. 10:54:41 And so, the one that I was specifically using is, it would be unfeasible to to go that far, because we will resolving in a couple of scales and but with sort of a more sort of a proper le s with, you know, proper sub grid scale thing like. 10:55:01 It's currently some work that's been done by some collaborators, Catherine rodent Hill, who, who published this earlier. 10:55:10 And she's doing larger scale ones which which can resolve things like. 10:55:14 And the question is whether you can get any double piece of buses in those and I suspect we on some like. 10:55:19 I mean this was pretty interesting question because I was investigating Egmont layers and the ocean 2019 together with some colleagues among areas was really just the first centimeters in the ocean what we were looking at. 10:55:31 And now you're bringing up the question like how does it extrapolates, like when you have a different setting up to hundreds of meters. And of course, you cannot reach this kind of information with alias, so I mean this is kind of interesting that you 10:55:42 brought up also this one the modeling together, 3d equation modeling that we kind of may have to devise a strategy to to extrapolate the DNS right so what the earliest so you see me yes it is feasible and some upper part of the, of the ocean and we know 10:56:03 the staircases which are kind of developing further down walls. Now the question is like how can we bridge the gap. And I can, I guess this is pretty interesting to have this lower dimensional models for that. And I think if, if I can present next week's 10:56:12 able to learn more about this or maybe you've seen Alan Christine's talk about this is only a team model. So I think this is would be one, one of the possibly candidates that might do it or even ballsy may do it, but Francisco has shown his talk. 10:56:25 So I'm pretty excited about these options to use the word I'm others in order to reach these kind of gaps as we are actually facing and the atmosphere as well as in the ocean. 10:56:41 Yeah, and I think I guess that links back into your last point about and wanting to, you know, model the situation, you know if you can couple some parameters ation for the dumping of internal waves, you know, potentially that's something you then capture 10:56:47 that feedback in one day model, maybe, I don't 10:56:54 know, probably details have to be discussed them so that's why I'm pretty excited to see the series increase you know from the Antarctica staircases differ so that's pretty a pretty exciting. 10:57:06 Kind of just wrapping my head around and when I was wondering the second point for me, just go ahead and opt out so people kind of raising the essence of movement. 10:57:14 I just want to point out we're getting pretty close to the hour so we would like this time To sum up, maybe I'll just remind you that piano and Leo decided to proceed. 10:57:27 It's just, again, pretty quick question, I don't know, so uh, yeah. 10:57:36 Okay. Why don't you say it and then we'll discuss it later. Okay. 10:57:40 I was wondering if there's a difference in this year between the article ends and article so I don't know the data but you say the second so different. 10:57:53 And maybe I was wondering if so she might be the reason. That's an interesting point one, I'm sure we enjoy chatting about on slack has been a great discussion guide, and I've really enjoyed all the, all the different aspects that have come up a lot of 10:58:09 quite different from the ones that I might have initially thought which is perfect you know I guess that's exactly what you're looking for. And, yeah, I don't know if you have some good thoughts in summation Yella. 10:58:14 Oh here we we got we discussed a lot of different points. 10:58:19 To sum up, I suppose we identified, just major areas where the this this research can can go to. 10:58:34 Yeah, I didn't take notes to give a full summary. So, at this point, given how the discussion went, maybe the best option is just to ask Bruce who took notes to to to least the main, main questions and and summary points on the spot, but I think maybe 10:59:02 a little bit. But 10:59:01 yeah, this is I think what happens when we are flexible I don't follow the schedule but this is something to learn for the next two sessions, but I do believe that the discussion was was very, very amazing. 10:59:18 I think the way the two of you ran the discussion was absolutely perfect. It was really good discussion throughout. I have cramps my fingers from taking so many notes. 10:59:25 I couldn't possibly actually get my brain dump into words right now. However, all the notes I'm taking I am going to write into a succinct summary, which I will pass to Yana and Leo for approval. 10:59:40 And then that will be posted on slack and hopefully become a document that will also appear I did the same thing last week with a PD staircases and I'll be doing for the next two as well. 10:59:52 So, there will be a record of these discussions that hopefully will stimulate further discussion. 10:59:58 Are we all good. 11:00:05 Yeah. 11:00:03 I mean Sonja said going together down so I'll probably go there for anyone who wants to chat more about I stuff. Absolutely, please feel free to join the gather town I think we've given enough instructions on how to do that but you can find the information 11:00:13 on on slack Besides, I just want to remind everyone to that will be meeting next Monday again that session will be led by Alexis Kaminski and Nicole Shipley with the topic being on turbulence and his influence on staircases some of what you heard of today, 11:00:32 and also perhaps observations were turbulence is actually creating the staircases, so that'll be our discussion next Monday. 11:00:39 Thank you, especially to young and Leo Great job, and I look forward to seeing everyone next week.