 of the ASP summer tutorial. Today, the topic is air sea interactions, and I'm delighted to introduce the first speaker, Professor Chui Cheng, as a professor of atmospheric sciences at the University of Washington. Her research focuses on air sea interactions and precipitation in the tropics and coastal environment using airborne and satellite observations and coupled atmosphere wave ocean models. She has been the lead scientist for many field campaigns and also served on advisory boards, such as the vice chair of the National Academies Board on Atmospheric Science and Climate. It's a pleasure to have you here, Chui. Please go ahead. Thanks, Judith. Thank you for the introduction and good morning, everyone, if we're in this sort of a three time zones so far, but still all in the morning. Okay, so I'm going to share my screen now, just so can you see my screen? Yeah, I can see it. All right, so I'm going to turn the camera off, just making sure the bandwidth is okay. Okay, so, you know, after so many two days now, you'll continue to hear a bit more. So this particular lecture, I'm going to focus on a multi-scale air sea interaction process from a convective system known as the center storms, to MGO and ENSO. So in terms of the time spatial scale, we tend to think we know weather and we know the climate, which is ENSO side, it's a longer time scale. Then MGO is somewhere in between. So I want to knowledge this particular work that I will be showing you quite a bit of results coming from multi-students post-docs and contributions to my group, Yaaklin Gorkin. It's the one that contributed quite a bit of the material in here, Brandon Kerns, also Aida Saravan, all PhD students in my group. I will start with this image here showing a instantaneous satellite image with typhoons or tropical cyclones. You can easily identify that's cyclonic circulation northern hemisphere, anti-cyclonic southern hemisphere. So the question is, can you identify MGO on a weather map? So MGO is a phenomenon that do show up on everyday maps. So how do you identify that? What is MGO? As it turned out, these big systems here, that background is the water vapor, vertically integrated water vapor, and the precipitating part is the green-gray area. So MGO is this not a very well-defined, large-scale convective system that can contain lots of different pieces of convection with some central piece that like strong surface wind and then sometimes spin-off tropical cyclone as it forms in the ocean, propagates through maritime continent and move over to the western Pacific, along with the way it's interacting with lots of different phenomena as we all know. And also, as you heard, so I just kind of have time to turn the screen here. This is a map from the work from the people I've shown. You have seen this quite a bit about how to look at global impact of this multi-scale phenomena. MGO is the phenomena sitting here, usually formed over the Indian Ocean and propagates toward west Pacific and central Pacific and interacting with this phenomena downstream like the ITCZ and so on. So last few days, you also have heard a lot of the connection to the mid-latitude wave trend and so on and so on. However, the complicated piece of MGO is none of the MGO individually are like your composite exactly. So MGO can propagate toward the south that center of the heating is actually sitting in the southern hemisphere. That has a very different implication than the heating source is sitting on the equator. On the other hand, MGO can also propagate northward. So this is where I want you to check out all your pre-notion about, you know exactly what MGOs like because MGO individually is quite diverse if we are looking at the downstream influence like the tropical cyclones heat wave and so on. The heat source is sitting northern hemisphere is very different than the one sitting in the southern hemisphere and traditionally this has not been identified, we don't have a tool to identify this diversity of MGO. So again, what is MGO? How diverse is it? How can we actually identify individual MGO events on the weather maps? And what is the air-sea interaction and the role in the coupling cross-scales? So how does individual convective systems interacting with MGO as a large-scale dynamic system and how the MGO could be upscaling to interacting with and so that's the subject in the next 20-25 minutes. So I will be using quite a bit of data as Judith said. I'm a fan of the few campaigns. I use a lot of observation to describe the phenomena we observe, understand the physical process. In addition, we also use the coupled atmosphere ocean models to understand the physical process. So this is usually you would see in this half-molar diagram in Zhang's paper 2005 that you're looking at precipitation. This time goes up and then this is a global scale here and then you're basically looking at eastward propagating this enhanced precipitation associated with 150 millibar winds, westerly winds. So that just half-molar diagram tells you the zonal property of the MGO. And then the same time we decided we need to describe MGO in a little simplistic way. So we're making composite. So over many years we accumulate data, we're making composite, which means we treat MGO as anomaly. So this is a composite where it started to identify certain properties, eastward propagating, a connection that is positive anomaly and negative anomaly. Then we do EOF analysis and so on. So we end up with the RMM index, which is all very familiar to many of you. These are very, very useful and the convenient tool to use. So basically we're looking at phase one, two cross maritime continent. This is Indian Ocean going all the way around and then get to phase eight. So that's composite. Individually does MGO look like it's a composite? And the same time does this thing representing not only zonal, meridional variability, right? Obviously this measure doesn't actually do that. So with that in mind we decided to come up with another way to track MGO that as a physical quantity using precipitation maps. So this two paper by Kerns and Chen 2016 and 20 that we use trim data over 20 years to decided to track the largest scale precipitation. The way it works is that we'll filter the data and the 3-day accumulation. So we're making sure that what we're tracking is a large scale precipitation object. Then we track them in time. This tracking method has described in there. I wouldn't go through too much details here, but you do need a threshold in terms of a millimeter per day and how big spatially and so on, so on. So we want to track this at least greater than seven days. So that means we want these truly large scale beyond synoptic scale precipitation patterns. Then lastly the MGO have another concentrate that has to be greater than 10 days. We want to be sure that it's into the sub seasonal scale at least the 10 days. And we also concentrate that it needs to be propagated eastward. So with that in mind, this is an actual precipitation map top left corner and then we accumulate in precipitation and making the tracking part. So this black curve is the one that I talk about large scale precipitation object and that tracks the whole system eastward. So this tracking method leave us with an entity that is called MGO LPT systems. So the lower diagram shows you the something tracking time. So this color representing time. So now you can see the MGO has certain places to start in the ocean wobbling around the north and south and so on, so on. Make this big footprint of a phenomenon that we usually don't think about them. But we're thinking them phase one, phase two. But in fact many MGOs doesn't quite get to the entire phase of things. So the advantage of this is that now we're not only having MGO property eastward, we also can look at meridional structures. This happened to be extremely important for downstream influence of how MGO influence global patterns. So I include an event compared to this particular event was observing during the dynamo field campaign. Then you can track many of them and you can see a lot of complicated wobbling around and this particular event propagate quite farther into the south and southern hemisphere. So the implication or its dynamic properties of this MGO versus that is quite different. So to summarize all them up on this map, the top panel shows you the tracks of each tracked system over 20 years. Then the starting point is this blue dot and the red is ending point that these systems at least to be seven days tracked. We call it LPTs. The lower panel just show you the density of these tracks and tell you where they reside more so than the other stuff. So then you can see these system can be tracked across Pacific. So if I add a constraint to 10 days and beyond has to be propagating eastward and now we're left with MGO LPT. That's over 200 events. Now you can see the diversity of MGO. They tend to form in the equatorial region, these blue dots, but they propagate not only along the equator, they end up north and a lot of this northern system is in the summer and the burial summer and also winter season, January, February tend to be farther south. So there's quite a bit of seasonality, individuality and so on. With this system we can actually identify MGO individually. So now I want to leave you with an important thought about MGO. So given that we can track MGO individually, we can look at the percentage of MGO or contribute the total precipitation in the annual precipitation over the globe, entire globe. So if you're looking into this region in the Indian Ocean and the West Pacific, so in this paper we described MGO itself contributed close to about half of annual precipitation. Think about that. MGO has been viewed as anomaly, but in fact MGO is the base state here, contributed to half of that. So we have a problem in terms of looking at MGO as anomaly. So many of the theories and things build on that as anomaly, but not as a real event. So another diversity of the MGO is MGO have the property that formed over Indian Ocean and some cross-merit continent, some cannot. So non-propagating ones. And apparently this was referred to as a barrier effect of MGO. Some form in the over the West Pacific and continue propagate into Central Pacific. So it's quite diverse. And we've done a lot of work to try to understand this barrier layer effect. So so far I don't think I have time right now to go through all the detail except to say that this barrier layer effect is still a mystery that continue to really drive a lot of research on that. So not getting into the couple of the system. So if you look at that, the MGO properties which is observed. So the ones formed in the West Pacific, the bars here is observed, the first one is a trend. It's the blue bar is how many events over this 20 years. And the green and the brown are the ones formed over Indian Ocean. Then the brown ones is the one that cannot propagate cross-merit and continent. And then the green ones does cross, right? So if you look all the global models here, you don't need to pay attention to individuals. You can see model over all have difficulty to actually have MGO formed over the Indian Ocean, right? So really very little green bars and so on ended that. But the models way overdone over West Pacific. But if you look at these pointers of a several coupled model in general, coupled atmosphere model does much better in general to produce MGO with one particular property that coupled model does is that coupled model tend to produce better speed and zonal latitude, longitude range. So if we plot a dot, this is a trim. So basically you can see this propagation speed in the zonal direction. And this is the range of zonal in terms of degree of latitude, sorry. And then you can see here trim set here. And this is ECMWF reforecast. So in general, this are kind of a good corner. And if you see this line, majority of a coupled model actually does much better than majority of uncoupled model. So uncoupled model is way too slow, right? So now I'm going to try to get you a few things about physical process. What are the physical process the ocean now contributing to better representation of the MGO? So in the past, you probably heard about ocean have this slow time scale. If you are looking at a phenomena that in a few days time scale, you can forget it. You don't need the ocean. And then if you go on to longer time scale, then ocean started playing a role. I want you to drop that concept. Ocean contributed right way and interacting with the MGO in the much shorter time scale. And in fact, ocean contributed to the detail of the structures as well as the east war propagation in this very closely coupled system. So that's just what I said. Okay, how do we know that? We did have a very comprehensive view campaign that's called a dynamo took place 2011 to 12. And this is a map showing the lots of different instrument and platform being used aircraft ships and so on over this area. So I was an aircraft scientist. So I flew through a lot of these flights. These are the red ones the area will fly through. So in the aircraft measurements, we can measure much higher resolution in terms of a storm itself, SST, the cold pools generated by center storms. And then we also drop the ocean drop. We have both GPS drop sounds and ocean AXPTs measure temperature in the ocean in the upper ocean and the deeper ocean part. This is sort of a picture of that. So then you can tell that center storm can generate cold pools almost degree to two degree cooling in this very high frequency time scales. But then at the same time, after month and a half measurements, we also captured a well observed MGO event. So just give you an example of this kind of a measurement that can tell us a lot of things about air sea interaction properties. So just to focus on the left side. So we measure from the aircraft and the way I plotted it, we plotted here. By the way, all these results is in this BAMS paper, 2016, you describe atmosphere, the winds and so on, the atmosphere boundary layer, ocean, upper ocean mix layer, and so on and so on. So we can categorize it as a surprise phase, which is not very active MGO phases and then active MGO phases. You can see the upper ocean in the MGO surprise phase is warmer during the active rainy period of time MGO ocean upper oceans cooler. And then at the same time, you can see the cold pools from the collective system are much stronger in the surprise phase because it's dry. As you know that in treatment from the dry air can induce a very big cold pool. So if you put all together, as you can tell that this air sea fluxes, this is the sensible heat flux, latent heat flux. From surprise phase to active phase through transition period, you can see the air sea fluxes increase tenfold. These has all different roles in playing in how to getting MGO from surprise phase to the active phase. So without too much time, I'm here only example why this thing matters in the MGO time scale. You can tell from the synoptic weather scale they do have a lot of details. To put them all together, I want to come back with one question that I asked before I get here is what is oceans role in terms of making MGO propagate eastward or help? So on the top row, so now we're using both observation and the models we can do uncoupled atmosphere simulation, coupled simulation, and better air sea fluxes are corrected by the observations. So the top row shows you the MGO propagate eastward in terms of both precipitation winds and induce very large ocean cooling over the entire Indian basin. So if you run the uncoupled model, you don't have that property. So precipitation sits still because there's too much excessive energy that it doesn't exist in reality. So rain tend to stay stationary. And then if you use the one that I just showed, you can see the propagation observed and the very stationary MGO over in the ocean. So once you start coupling, you start seeing the eastward propagation even better when we correct air sea fluxes based on observations. So air sea coupling matters. So with the rest of the part, I'm going to tell you a little bit of current work that we're doing is that MGO propagate toward the West Pacific interacting with this phenomenon that you know very well. That's a large-scale insole patterns. So this is a diagram probably doesn't really need much explanation, except to say that this is a line in your state. There's a temperature and a very strong treat wind, eastward winds. During the normal phase MGO, the insole, so you can see the precipitation and the SST map. So the warm part of SST closer to the central Pacific. The important thing is the thermal client. So the thermal client is much tilted. And then by the time you go into line in your state, the thermal client tilted downward. So you have a much deeper warm pool and then bigger warm pool in the central Pacific and so on. So what this thing has to do with MGO, right? So as it turned out, MGO has a very important property that project onto insole spatial and time scale in terms of upscaling. So this is MGO that if you look at MGO precipitation and also the westerly winds, it's very large scale. So as MGO propagating into the west Pacific and central Pacific, that projecting onto the scale that insole, a matter to insole. So this happened to be some events that we observed and also we modeled. Quickly. So over the last 20 years that we find in this paper that the insole event over the last 20 years, the on-site of insole, this is a Minio 3 or 4 index. So you can see in general right before the on-site, you have enhanced MGO event, every single insole event. So there's definitely MGO contribute to this eastward warm pool extension. So to project it onto the normally phase that if you look at MGO, it gives the warming, I'll just say this last panel, that the warming really projected onto the Minio 3 and 4. So what happens when MGO coming over the west Pacific, they influence the ocean. So the way MGO influence oceans through both dynamic process and thermodynamic process, MGO rain and produce the strong winds and generated cabin waves, that's well described in this paper that deepened the thermal client. So repeat MGO event can really continue that process projecting into large scale. The other process is through the thermal dynamic process, again it's observed through fuel content during PogoCore. This paper describes another phenomena that is after the MGO rains and dump a lot of fresh water. So the fresh water on the upper ocean gave you this freshwater lenses and this particular lens that can generating this pool of warm water because after MGO is gone, you have enhanced the solar radiation. This warm pool can help generating this something called barrier layer. You can go into the detail in this paper to describe the salinity distribution, temperature distribution. You have a very deep mixed layer during the strong wind events, but then you have this very shallow warm layer and then the barrier layer is by its name is blocking the cold water through this barrier layer get to the upper ocean. So the upper ocean warm part gets protected through that process. So this particular event, then we did a modeling study over this region. So we find that MGO influence, I don't have time to go through the model detail, apologize, but this will be made available to you. So then we have made this couple simulation over eight months. We have a multiple MGO event right before this outside of the insole. Then we'll find each MGO event contribute to deepening of the thermal client. And also most importantly, MGO also produces fresh water pool. You see the blue ones that MGO after MGO, it's pushed toward the central Pacific and with a very interesting fresh layer. So I don't think I have time to go through the details. I'm just going to summarize in this schematic to show you this multi skill process. So this is the schematic that many of you are familiar with it published from based on a publication some years ago. So over the West Pacific, you have a warm pool and this is central Pacific. The couple system has zonal winds, it's a treat winds, Easterly and MGO sit over here. So when the MGO produce a large freshwater pool, then generate a barrier layer. So then this freshwater pool become active and move eastward. That's the MGO. So by the time that the warm pool, the pool started to become active in the ocean, now has life on its own. So this water now is actually propagate eastward against the wind. This is something that is brand new we just discovered from the couple model simulations. So the first was came as a surprise, but if you think about it from the Adam's third point of view, you know the dense water here and the less dense freshwater is going to have this sort of frontal system that the lighter water sliding above the dense water the subduction of this. At the same time, this part generate a pressure gradient. This pressure gradient in the upper ocean itself drives this very dynamic water in here. This water is the result of MGO even though after MGO precipitation is gone, this piece of water remain to be alive and that actually communicate from the West Pacific toward Central Pacific and East Pacific. So after multiple events, you can see the deepening of the thermal client by MGO through Kevin waves. So eventually after multiple MGO event, you can start seeing the coupled effect. So the water pushes Central Pacific the reduce the SSC gradient. Now that property is going to relax this treat when as soon as the treat when the relax, you're getting this fully coupled mold for insole that's the onset of insole. So this you can't miss everything. The one because the high frequency one that determines how good the freshwater pool is. And once the freshwater pool down in an ocean, then ocean takes its own dynamics and then that contribute to much larger basin scale properties. So I don't think I have a time to get to this one here. Do I have one minutes to say something? This is my last slide. I run over time a little bit. Yes, go ahead. Okay. So I want to touch on one thing that the last yesterday, both Eric has mentioned this heating sources that MGO in the Northern Hemisphere versus MGO in the Southern Hemisphere and the on how describe the global impact of MGO, right? So as it turned out, this is a good example of a detailed MGO event. This event versus that event have a very different downstream influence, right? So the way we did this MGO influence is that you can look at on the sub seasonal to seasonal scale. You can look at MGO influence on tropical cyclones, atoms, the rivers, heat waves and the floods and drought and so on, so on. This particular slide just show you one example. That's the tropical cyclones, right? So the way we plotted this thing is that we will look, if you have MGO versus you have no MGO, then you will compare how many events that are associated with MGO, how many events that tropical cyclone occurs without MGO. So that actually gave you a quite good indication of MGO downstream influence. So basically, this is the event of tropical cyclones associated with MGO versus MGO, the tropical cyclone without MGO. So pretty much the lentic basin, it's very interesting because almost twice as much of that tropical cyclone is associated with event in MGO. On the other hand, if we break down to MGO in the northern hemisphere versus MGO in the southern hemisphere, so here is this bar. You can ignore all the bar, just look at this last one pair, okay? In the north Pacific, north Atlantic, then you can see the ones when the MGO sitting northern hemisphere, you get majority of tropical cyclone event. This is a very large difference, right? So then for Atlantic, it's the same. So that tells me that MGO detailed MGO structure has a lot to do with downstream influence, but you need to distinguish where the heat source is. So with that, I'm just going to summarize this, which mostly already stayed. I want to make sure that we all sort of gave a detailed look to think about MGO that can be a source of predictability through this multi-skill air-sea interaction process that bridging both weather and climate. So with that, I can probably take any questions if I have some time. Thank you so much. Thank you, Shoei.