 This is definitely a male and then a single male. So this is, it's also men, but it goes under my shirt because men go the other way, you know, and it comes on the outside, so for women. I will put it on this one. Okay. Yeah, I already introduced myself. Okay. So welcome back. And I will try to walk you through a review of teleconnections on intracisional timescale. And this is actually the topic of a recent review paper that we just published in review of geophysics. Before I begin my presentation, I would like to explain the many of the acronyms that you see on my title slide. There is no need to explain the ICPP acronym, our host and main sponsor of this activity. Here on the left, these two acronyms represent two major programs of WMO, the WWRP, which is focused on weather and WCRP with a focus on climate. The third acronym here represents the sub-seasonal to seasonal prediction project, which is a project sponsored by WRP and WCRP designed to bridge the gap between the weather and climate. And finally, this acronym here, YTMIP stands for the Year of Tropics Mid-Latitude Interactions and Teleconnection, and this is an international project fostered by the sub-seasonal to seasonal project. And it started this summer and will go on until summer of 2019. The main objectives of this project is to try to answer some of the still remaining questions regarding the teleconnections on intracisional timescales. So the outline of my presentation includes a description of the influence of the tropics on the mid-latitudes, and I will present this from an observational perspective, divided between the northern and southern hemisphere, and then I will look at some modeling studies in the northern and southern hemisphere. And the modeling studies I'm talking here are those related to try to understand the mechanisms explaining these teleconnections, something along the lines that David talked about. I will talk next about the influence of the mid-latitudes on the tropics and then the two-way interactions and feedbacks. Then I will review some of the work that has been done to forecast the teleconnections on intracisional timescales. Finally, I will try to summarize some of the remaining challenges, hopefully provide you with some food for thoughts for your future research. So the idea of relationship between the tropics and exotropics dates back to 1950 and can actually be attributed to Herbert Riel, who was analyzing the so-called, he called the topography of the 300-millibar surface on a particular day, which was August 26, 1945, and this is not a climatology, it's a real-time map, and he noticed that the most outstanding features on this map are the breakdown of tropical atmosphere into a train of work pieces and the complete interlocking of flow between the tropics and high latitudes. Riel hypothesized that heat is injected into the polar zones in a few narrow strips of longitude that he marked with these lines here, and he thought that at least in part changes of the flow configurations and intensity in high latitudes must be dependent on the availability of disturbances in the tropics to extend to create these troughs. And Riel hypothesized that the source energy of these disturbances comes from the latent heat associated with convection. Of course, at that time, people were not talking about teleconnections. This is something that we later try to attribute. About 15 years later, Birkness was comparing the distribution of the mid-latitude circulations for various SST anomalies in the eastern equatorial Pacific, going from not so warm anomalies to warm SST anomalies, and he noticed that the strongest circulation that develops in the zonal winds of northern hemisphere winter tends to be associated with the warm SST. And one interesting aspect, an interesting hypothesis that Birkness advanced at that time was that this will have implication for the future seasonal forecasting of what he called the climatic anomalies over North America and over Europe. So a long time ago, people started to think about the potential of doing seasonal forecasting. So for more than 20 years, this subject was set aside and reemerging in the early 80s when people started to talk about this relationship between the tropics and the extra tropics. And I would like to point out that it took more than a decade after the discovery of the MGO that people started to try to make connections between the MGO or convective activity in the tropics and the circulation in the mid-latitudes. And probably the main impediment was the lack of observations with global coverage. So based on relatively short record, these pioneers work showed that the mid-latitudes circulation can be affected by the tropical convection during the periods when convective activity is moving from the Indian Ocean into the western Pacific. And they demonstrated that when the convection is located in this region, the associated circulation consists of this pair of downwind cyclones located near the region with suppressed convection. These cyclones extend over the whole Pacific region and their meridional structure allow them to extend into the mid-latitudes on both hemispheres like we see here and near the 354B north. One particular paper that probably you were expecting to see listed on the previous slide is this one by Lipman and Hartman. And I wanted to mention it separately because they also found the relationship between the circulation in the mid-latitudes and the tropical convection. But they claim that this relationship can, this exchange can take place only in the exit region of the South Asian jet. And they, in this schematic, they show that the convective activity located here at the jet entrance region will induce this meridional circulation. We'll create this meridional circulation which will be induced by the aegiostrophic winds that exist in the jet entrance region. And they found no relationship between the regions of tropical convection and the anomalies over North America and the Atlantic region. So this was a slightly different conclusion than the previous studies. And one of the main questions that drove the future research and it's still an open question today is what are the geographical regions where the tropical forcing is most effective in exciting the extra tropical circulation anomalies. So once the reanalysis became available, people started to look at the variability of the climate patterns in the mid-latitudes such as the NAO and PNA and try to link this teleconnection pattern to the activity in the tropics. Around that time it was found that in the northern hemisphere there are two modes of variability with the periods of about 48 and 23 days. And also the North Pacific circulation anomalies develop one to weeks after the appearance of the convection anomalies over the tropical Pacific. So later on when the more data became available, people started to think of the MGO as a forcing of the mid-latitudes. And it was found that some studies were suggesting a phase locking between the PNA and the MGO with the MGO explaining about 30% of the PNA variability. Whereas other studies like this one by Riddle at All published in 2013 were suggesting that MGO may only excite a PNA like pattern but not a pure PNA pattern. So here in this figure taken from the paper by Riddle at All, we are looking at the probability, the frequency of occurrence of regimes in which the PNA events take place. And we are looking at each panel corresponds to an MJO phase, which for those of you who are not familiar with what the MJO phase is, you will hear something later on. And then the left panel corresponds to the positive phase of the PNA and the right panel corresponds to the negative phase of the PNA. So these red cards here indicates the high frequency of occurrence of the PNA pattern whereas the blue lines indicates the low probability of the PNA pattern to take place. So when we look at the distribution or the frequency of the positive phase and the frequency of the negative phase, we see that we see an asymmetry and we also see that the MGO doesn't have such a large impact on the negative phase of the PNA. So this is the reason that they reached the conclusion that this is not a pure PNA pattern, but it's a pattern that looks like the PNA. There are other studies so that looking at the influence of the MGO on to the North Atlantic region and it was found that the NAO is influenced by a tropical convection when the MGO is in either phase 2-4, which means the convection is located over the Indian Ocean or phase 6-8. It was also found that the activity of the North Pacific storm track during the winter is modulated by the tropical convection associated with the MGO. The winter surface air temperature over North America is characterized by a 70-day oscillation which is forced by the MGO and also it was found that the MGO phase speed affects the mid-latitude interconnection patterns. So far these results that I mentioned were mostly focused on the boreal winter, but we do have a very small number of studies looking at the influence of the boreal summer convective activity on the mid-latitude interconnection. The moon at all looked at the influence of the active phase of the South Asian monsoon and North American summer monsoon on the extra tropical circulation. In particular the surface temperature anomalies and they do find significant teleconnections and these teleconnections show both quasi-stationary characteristic and also eastward propagating characteristics. Also, these studies and these results that I mentioned so far, as you probably noticed, are focused on the northern hemisphere. But what about the southern hemisphere? Because when I presented that understanding of the teleconnections, we said that teleconnections take place in both hemispheres. Well, for the southern hemisphere, we have a very limited number of studies and the result seems to be slightly different from the northern hemisphere. So initially a lot of studies were suggesting that there is no significant correlation between the tropical convective activity and the large-scale circulation over the southern hemisphere. It was until 1993 when Berberi and Hegel found that the impact of the tropical heating on the mid-latitudes has a seasonal dependence. So here, this plot shows the composites of the 200 hectopascal velocity potential for days when the large oil-large convective anomaly is located at 127 east, around here during the southern hemisphere summer. So in this composite, we see these circulation anomalies that extend forward like a Ross B-Wave train. Whereas during the winter, the same composites, while they show these anomalies, the circulation anomalies, they do not show these forward propagation, the Ross B-Wave train that we see during the southern hemisphere summer. So moving on in time, observations with higher temporal frequency became available and the research started to be focused on the relationship between the tropical convection and the mid-latitude weather and in particular extreme weather events. So here I have two examples. One showing that the winter extreme precipitation over the US west coast is increased after the MJO is in phase two and this figure here taken from a paper by Jones shows the percentage of occurrence of extreme events. During the active MJO phases, which is this black line and non-active MJO events, the boxes that are listed here on the horizontal axis corresponds to various regions that they used to analyze. So it's very obvious from this plot that the frequency of events associated with active MJO events during phase two, it's much larger than the frequency of events that take place when the MJO is not active or is not in phase two. Another example is related to the other side of the Pacific and Jong in 2005 looked at the extreme cold surges in the surface air temperature over East Asia and he plotted here the distribution of these extreme events and they are represented by these larger blue dots and the other and this diagram shows the phases of the MJO. We will learn more about this diagram later and you will actually get to use this diagram in some of your lab sessions. So when we look at the distribution of these cold events, we see that they tend to be favored by the MJO convective activity located over the Indian Ocean. Another type of extreme events is the so-called atmospheric rivers and they are called extreme events because they are associated with intense streams of precipile water and they can cause heavy rains that are usually resulting in flooding. So a number of studies found the coherent relationship between the MJO, the associated raspy waves and the atmospheric rivers, especially the atmospheric rivers that takes place on the west coast of the United States. And they have also found that the MJO in phase six, which means that the convection is located over the western Pacific tend to favor these atmospheric river activities. And not only that, the MJO tends to modulate the atmospheric river activity in various regions over the world, like in Korea, Japan, Alaska, Europe, in southern hemisphere. And here, I'm just showing you as an example of large influence of the MJO on the atmospheric rivers. I'm showing the frequency difference which for the atmospheric rivers that affect Hawaii as function of the MJO phases. So again, we see that there are some phases of the MJO in which the frequency of occurrence is negative, which means that this particular phase of the MJO doesn't favor the development of atmospheric rivers. But there are phases of the MJO which favor, which leads to the increase of the frequency of atmospheric rivers. So these are these examples that I show here and these results were based on observations. So the conclusion is that observations provide enough evidence that support a statistical significant relationship between the tropics and mid-latitudes. But as you probably noticed, I haven't said anything about the mechanisms driving this teleconnection. So all the results were based on statistical analysis, just looking at observations. So now I want to show you some results from, I call modeling studies because they are based on models, but mostly theoretical models that of various complexities starting with very simple barotropic and baroclinic models like the one that David described, that were designed, these studies were designed to try to understand the mechanisms driving the teleconnection. So again, some of the early, in early days when only very simple linear models were available, people have used these models trying to understand the relationship between the tropical heating and the distribution of the mid-latitude circulation. And here is an example taken from a paper by Hoskins and Crowley, where they, in this linear baroclinic model, a two-layer model, they put a source of heat in the tropical region, and then they analyze the response of the mid-latitude circulation by looking at this 300 hectopascal geopotential height. So in this map, we see that the response of the mid-latitude circulation to this tropical heating, are these ways that propagate poleward and eastward in the upper troposphere. So there were also other studies that, in particular, David mentioned. They studied by Simons and all, and they also, using a similar simple model, they were able to show that propagation, mid-latitude perturbation over the northeast Pacific, are excited by tropical forcing located over southeast Asia and tropical western Pacific, and also the Atlantic perturbations, mid-latitude perturbations in the mid-latitude circulation, are forced by the tropical convective activity located to the southwest of this region, mostly Indian Ocean and western Pacific. So as David described, the mechanism of driving this teleconnection was the Rossby wave train. In this paper, they promoted a different school of thought, and according to which the mid-latitude perturbation are associated with the fast-growing mode of barotropic instability of the mid-latitudes, which are these barotropic instabilities are forced by the Rossby wave trains. And by advancing these hypotheses, they were trying to justify that, why the response of the mid-latitudes is not independent if you have the source of your hitting south of the equator or north of the equator. So, like David said, another important result of that time was that the atmospheric anomalies in the extra-tropic have a barotropic structure. And they also found that this variability of the mid-latitude large-scale flow is dominated by an oscillation with a period between 28 and 72 days. So one question that was posed at that time, and it's still something that we don't completely understand, is how and why do the baroclinic atmospheric anomalies in the tropics transition to barotropic anomalies by the time they reach the extra-tropics. Because, like David showed us in his pictures, when we look at the distribution of the disturbances in the tropics, they have this baroclinic structure, whereas the disturbances in the mid-latitudes, they have a barotropic structure. So, moving on in time, people realize that all of these experiments that were done with this simple model use very idealized sources of tropical hitting. And in the real atmosphere, the distribution is slightly more complex or more complex than these idealized sources. So Feranti et al. designed an experiment still using a simple barotropic model in which they derive the structure of the tropical hitting from the observations. They took the first two EOFs of the distribution of the oiler anomaly in the tropics and they imposed this forcing into a barotropic model. And surprisingly, using this very simple model, they were able to reproduce with high fidelity most of the futures of the 500 millibar pattern that corresponds to the observations. And so the main conclusion here was that the characteristics of the tropical hitting really have a large impact on the circulation anomalies that develop in the mid-latitudes, regardless of the complexity of the model. So next, we started to enter into the so-called GCM era when models became more complex. And Troubert and Park first tried to look at the global initialized analysis produced by the ECMWF model. And they found that the, or they show that the PNA appears to have its main source of energy in the mid-latitudes and the link to the tropics manifest as a phase locking with anomalies forced by the tropical convection. So this is a result, a confirmation of the result obtained earlier by other studies just looking at the observations. Since the earlier experiments in which source of hitting was specified were done with very simple models like the linear barotropic and baro-cleaning models, once the GCMs became available, people started to repeat these experiments with the GCMs that are more complex. And they found that the propagation of the rusty wave is sensitive to the zonally varying basic state. So this is a, it was a new result because the basic state in the previous models were just linearized model about the basic state. So this was not possible to be explored. And they also found that the preferred path for the rusty wave chain propagation is in the regions with prevailing westerlies. They also found that the response of the northern hemisphere to tropical heating is much stronger than in the southern hemisphere. And the rusty wave response to a fixed tropical heating establishes in 10, 15 days. And this result, it is actually consistent with what the previous results based on the observations. An interesting result was obtained by Hailean who's here in 2007 when they use a primitive equation dry atmospheric model. And they show that this model can simulate tropical intracisional variability with the structure of the Kelvin wave. And the divergent flow associated with this Kelvin wave generates wave activity into the PNA region. I thought this result was intriguing because everything that I discussed so far was based on the source of tropical heating which was associated with atmospheric convection. And an earlier study showed that if in a model you turn off the convection then the MGO will disintegrate in dry Kelvin waves and Rossby waves. So convection it is necessary for maintaining the coupling between these two waves and to maintain the 30-day oscillations that they were investigating in that model which translated into the MGO. Of course when we started to analyze the GCMs people started to wonder how well the GCMs are able to simulate these observed teleconnections. The early simulations suggested that GCMs tend to produce the observed relationships between the tropical convection and mid-latitude circulation normally, especially during the boreal winter. But they also show differences from observations and these differences were attributed to the model's inability to simulate the right location of the tropical convection. Another type of more complex experiments designed to understand the mechanisms driving the teleconnections were the so-called intervention experiments in which in a very complex GCM you intervene by nudging the tropical heating to some observations or some idealized heating and these experiments confirm the earlier results from observations that the extra tropical response is sensitive to the phase speed of the forcing. They also confirm that the tropical convection outflow anomalies lead to Rossby waves which then interact with the mid-latitude flow in preferred locations and where these interactions take place they extract energy from the mean flow and they allow the growth of the barotropic modes. Other results show that the mid-latitude response to the MGO depends on the history of heating and cooling and it's not just the response of some heating and cooling at a particular longitude. And there is also some lag between the time of the forcing and the response. And another result which David used in his presentation was that the short pulses of tropical heating also affect the mid-latitude and these effects can persist more than two weeks. Another interesting paper that definitely the experiments here were not designed to look at the teleconnections but the results are so relevant that I decided to include this paper in this review. Are these experiments by slingo and slingo in which they look at the impact of the cloud radiating forcing on the mid-latitude circulation. So they do experiment in which they cut off the cloud radiating forcing and they found that the longwave cloud radiating forcing in the tropics accelerate the subtropical jets and generate perturbation in the mid-latitudes with the barotropic structure. And they also found that the longwave cloud forcing over South America induces a barotropic cyclonic circulation in the mid-latitudes and anti-cyclonic structures in the northern hemisphere. So now I want to move on and discuss the influence of the mid-latitudes on the tropics and I will start again with a review in time of the evolution of this subject. And I think the motivation of this kind of studies were based on the need to explain the energy source for the observed tropical waves and also that were predicted at the same time by Matsuno. There was also another interesting paper that was published a long time ago in 1965 which was trying to explain the formation of hurricanes in the eastern North Atlantic by the coal surges from the Antarctic region. This is the only paper making this suggestion and I don't think anyone else after that has tried to look at this. So the idea was that the coal there from the Antarctic region will once it propagates in certain conditions it propagates northward and will erode the static stability in this region and that will enhance the easterly waves activity that will lead to the formation of tropical cyclones in this region. So very early observational results suggested that only exotopical waves with westward phase speed larger than the zonal mean flow can actually propagate into the tropics. And the strongest influence manifest over the Pacific and Atlantic Ocean. So these are the two regions where the activity of the large scale circulation in the mid latitudes can influence the tropical convective activity. We also have learned that baroclinically unstable disturbances of the mid latitudes can modulate the tropical convection and some of the results suggested that the mid latitudes to arms maintain the intracisional variability of the tropics. You have already seen this figure in David presentation. But the reason I am showing here is that there is a number of studies showing that the mid latitudes circulation has an impact on the IPCC, SPCC and South Atlantic Convergence Zone. And this happens when the exotropical transient upper level crops, which David showed us here, can propagate eastward and affect the cloud bands into this region. And probably you have heard of this as referred to the so-called PV streamers, potential vorticity streamers. And how this affects these regions, it has been observed that during the boreal winter, a rosby wave train dominates the northern hemisphere, whereas during the boreal summer, the rosby wave train dominates in the southern hemisphere. Another impact of the mid-latitude circulation on the tropics has been observed in association with the so-called summer monsoon breaks. And we have a number of studies showing that the mid-latitude circulations that have large tropes penetrate into the westerlies over India and they can cause breaks into the monsoon. And these breaks have been observed to have, can last between, anyway, between three to five days up to 20 days. The West African monsoon also experiences some dry spells, which are also associated with the mid-latitude intrusion of coal and dry air from over the Europe and Mediterranean region. So, another type of event in which the tropical weather is affected by mid-latitude circulation are the so-called cold air surges. And the cold air surges take place in many regions in the Asian Pacific region, South Pacific, South America, Central America, Caribbean, Africa, Indian Ocean, maritime continent along the east coast of Australia, and also in North America. For example, we experience freezing in Florida, which affects the orange. So, now going into the theoretical part, trying to understand what are the mechanisms allowing this to take place. Using a simple model, people show that the propagation in the meridional direction in the presence of the mean easterly flow is only possible only if the phase velocity of the mid-latitude wave is more easterly than the mean flow. Then the large scale disturbances generated in the northern mid-latitudes may have a significant influence on equatorial region if a westerly duct is present. And some of the studies suggest that the equatorial response of the exotropical forcing does not rely on the presence of this westerly duct, but it's a direct projection of the forcing onto the equatorial trapped waves. So, one of the big questions that remains in trying to understand the influence of the mean latitude into the tropics is what are the systematic aspects and mechanisms of the exotropical initiation and maintenance of the organized tropical convection. So, moving on, because I think I'm getting close to the end of my time, I want to go very quickly over the two-way interactions and feedbacks between the two regions that David also mentioned. And the current understanding is that, for example, rusty waves excited by the MJO will propagate into the subtropics, will break here, and when they break, they can influence the convective activity in the tropics, and then the cycle repeats. And the Mansun Braves is also an example of the two-way interaction between the mid-latitudes and the tropics, because the changes in the convection induced by the mid-latitude circulation then will affect the teleconnections of the boreal, somewhat intracisional oscillations. So, a main question that remains is to what extent are these dominant tropical and extra-tropical intracisional oscillation connected, and David tried to address a little bit of that. And then more discussion was initiated by the audience. So, of course, it's the ultimate goal of understanding the mechanisms and these teleconnections is to use this knowledge for improving the forecast range, and the sub-seasonal to seasonal variability seems to be a time scale that can largely benefit from these teleconnections. And, for example, a number of studies have shown that the extended range of the mid-latitude large-scale circulation with small errors in the tropical simulation of tropical hitting, it is actually, they are actually skillful. So, this figure taken by a paper by Frederick Wittar is showing the NAO index for day 1925. The solid line corresponds to cases when the MJO is present in the initial conditions and the dashed line corresponds to cases when MJO is not present in the initial conditions. So, if we looked after year 2006 here, we see that the cases with the MJO show a larger correlation which translates into a better forecast scale than the cases with no MJO in the initial conditions. So, now you may say, but what about what happened before? Well, before the model did a very poor job in simulating the MJO. So, actually when the MJO was supposed to be present in the initial condition because of the large errors in the location of the tropical hitting, the model had very low skill. So, I will wrap up with listening some of the remaining challenges that the YT-MIT project is trying to address and also the sub-seasonal to seasonal project is working on too. So, for example, can we understand the mid-latitude teleconnections from the fluctuating tropical hitting as a time-leg stationary wave response to the hitting or does the time-dependent wave interface play a role in the response? So, how do intense mid-latitude storms and forward-propagating tropical storms interact with the polar vortex and alter the annual modes on sub-seasonal timescale? So, this is a little bit of an extension of the teleconnections into the much higher latitudes. To what extent are the dominant tropical and extratropical intracisional oscillation connected? What aspects of the intracisional hitting arising from tropical convection are most important for forcing the extratropical responses? We heard of various results where various studies have looked at different aspects and found different aspects being important, but we don't actually know the relative importance of all of these factors. Other questions that remind what is the sensitivity to vertical and horizontal structure and to temporal evolution of the hitting and why? Does tropical forcing amplify the intrinsic intracisional variability or excite new perturbations? So, I think this has not been established and, for example, most of the studies, for example, looking at the influence of the MJO on various weather events, they show an actual modulation of that event, but none of the studies are actually showing a triggering of that particular event. Something that we would like to understand is what explained the hemispheric asymmetry of the responses to tropical forcing. There is only the lens distribution where there are other factors that contribute to this asymmetry. And one very important aspect is the role of the basic state errors in the simulation of these teleconnections. And I think this is my last slide. So, thank you very much and I am happy to answer questions.