 Hey, good morning everyone. So in the in this period of trying to understand mechanisms and using a hierarchy of models I'll be presenting today a process study that we did trying to understand the role of the connection between the tropics and the extra tropics to possibly generate low frequency of our ability in the Equatorial Pacific and this is a work that I've been doing in collaboration with Franca, Moltenia and Fred Kujarski So we've been listening a lot already about the PDO and what the PDO might be so this is a little bit of a summary that the PDO might be a Connection of many different things for example stochastically driven It could be a passive ocean response to atmospheric fluxes But it could also be perhaps a couple mode of interaction between the ocean and atmosphere as somebody has postulated before About what the PDO is made of it could be associated with both tropical forcing It would be listened to that through an atmospheric bridge of low frequency answer signal but also local exotropical IRC interactions and We've been listening to this talk already about the PDO which is probably not physical mode But it's just a statistical mode made of different physical processes so I've been Looking at one of these possible physical processes and what is the role in generating? Answer the KDL variability so they're all of tropical exotropical interaction So answer the KDL variability. There's a few hypotheses out there and some of them are for example the connections in this case now would be a Atmospheric teleconnections from atmospheric bridge, but the other around from the exotropics to the tropics It could be just tropical noise and and mean state changes Generating this low frequency answer modulation, but it could also be a teleconnection from other basings So for example the Atlantic generating low frequency variability in the Pacific and If we manage to understand some of the processes that might give rise to this low frequency variability in In the tropical Pacific then perhaps we can also put forward yet another explanation for the warming hiatus So this is the observed PDO Which at the Kdom timescales as I said is probably a Third a third of it is remotely driven from the tropics, and that's the time series of the observed PDO from Trumburth et al So we know also that there is a very strong correlation between PDO phases for example positive and negative phases and And so low frequency variability. So basically when you have a War mode in the PDO you have mostly aninos and the opposite So the connection between the PDO and the global means sees a first temperature and trying to understand The possible reasons for the warming hiatus This figure has been shown a similar figure has been shown already in a few papers So basically at the top. I'm showing here the global means sea surface temperature up to a month ago and the PDO index from observations and so you see the recent rebound in the PDO and the Accelerated signal already seen in the global means to surface temperature And again, you can go back to your previous hiatus periods and see and see that when you have a negative PDO You have hiatus periods and the opposite on on a PDO in a world phase So explanation for the recent warming hiatus. There's been a few already and for example The one from Kosaka and Shia in 2013 which says that a quarter Pacific liquid or a Pacific cooling has probably The most Influence on this warming hiatus and England et al has shown that if you account for the trends in the Trade winds in models, then you can reproduce the warming hiatus in model And this is a schematic that we've seen that we've seen before So all of these studies are pointing to a major role for the local Pacific in explaining the warming hiatus But there's also as I said teleconnections from the Atlantic which can cause low frugacy viability in the Pacific This is the sea surface temperature difference from observations between Two decades straddling the year 2000 You can see this negative phase of the PDO and in this case you can clearly see the AMO signal Even though there is a lag between the two indices so correlation teleconnection, sorry between the two basins can give rise to a Equatorial Pacific sea surface temperature anomalies and this is a pacemaker experiment by Kuchasky et al In which you can see that if you force the Atlantic with prescribed Well, if you prescribe sea surface temperature in the Atlantic in this case with a negative AMO signal Then you can force a positive sea surface temperature anomaly in the Pacific So part of those sea surface temperature anomalies in the Pacific might also be driven by teleconnections from the Atlantic So here I will try to put together two hypotheses that have been Going around for a few years, which are the atmospheric bridges that we've been listening to already but also the oceanic tunnels so also the ocean dynamics plays a major role in in tropical and isotropical interactions or at least we think they do and The upper circulation so the thermocline circulation gives roughly the time scale of these Seculations which is on the order of a decade so the the main hypothesis is that If there is a tropical forcing pattern that can force Exotropical flow response like in generating part of the PDO index then perhaps those atmospheric Signals in the isotropics can also foreenshune response that can feed back to the tropical Pacific And so the how the ocean does this is mainly through the isotropical cells these are observations from Zhang and McFadden in 2006 in which They are showing a quarter work convergence of the isotropical cells So the effect of both isotropical cells converging Mass flux at the equator and here the time series is reversed to show the highly anti-correlation with Equatorial C-surface temperature anomalies. Okay, so this is Data and here you can clearly see a low frequency signal in the isotropical cells and the high correlation with the C-surface temperature anomalies in the equator So we're going to be using a model and it's not going to be a couple model It's going to be an atmospheric model firstly Forced by C-surface temperature and then we're going to use an ocean model driven by those atmospheric anomalies So the model is pretty coarse is but it's the global region a global ocean model to the great resolution We have a control run which is forced with an atmospheric state Which is called the core coordinated ocean ice reference experiment Which includes all sort of data sets and the atmospheric model is the speedy model Which is developed here by Franco Moltenia and Frank Kowalski So what we're going to do is first we're going to run the atmospheric model Forced by SSTs observed SSTs and we're going to run a 10-member assemble of that and the Interanel variant SST are only present in the Pacific but in the Other regions Atlantic and Indian we're going to be using a climatological so We can be sure that the signal that we see is just produced by the interanel of our in SSTs in the Pacific Then we're going to take that ensemble member and we're going to take all those Atmospheric fluxes and we're going to compute anomalies the K-del anomalies straddling the year 2000 And we're going to feed those anomalies to our ocean model, which is forced by climatological fluxes Climatological Anomalphic state plus the anomalies derived from the atmospheric simulation So this is the anomalous forcing that we get By forcing the atmospheric model with SSTs observed SSTs so you can see the temperature The temperature up there the sea level pressure signal and the wind and wind curl Okay, so you can see Familia pattern here in the Pacific a very strong asymmetric response in terms of sea level pressure with your Newton law and The wind stress and we stress curl in this case that we are forcing with a negative SST anomaly at the equator The wind stress and we stress curl are responding in an in an opposite sense as the climatological State and that's going to be very important So now you might wonder where is the anomalies that we see in the exotropics coming from is it? Coming from the tropics, or is it local air sea interactions? So what we did is to force the those 10 member ensemble with Only anomalies in the equatorial Pacific so that we are sure that all the forcing comes from the equatorial Pacific and now from The local air sea interactions and this is the pattern that we get which is very similar to the previous one So most of the response in terms of wind and winters curl is really coming from the equatorial Pacific Forcing okay, so we are going to be running the ocean model now with all those anomalous fluxes and So as I said, we have a control run. It's very long control run Quite spun up and then we have a few perturbation experiments in which I will add the anomalous forcing to my Ocean model. Okay, there's a few of them Speedy all I put in all the anomalous forcing into the ocean model and down to the most simplified And the one where I get the largest response, which is pdw where I'm only using Anomalous forcing in the wind stress and winters curl So this is the atmospheric response, which is what we expect if we're forcing with a negative anomaly at the equator Then we get a weakening of the Hadley cell. This is The first top plot is the decade 2009 and this is the atmospheric stream function for the decade before If there's a difference between the two you see weakening of the Hadley cell and also a shift towards the equator of the Hadley cell Which is constant with the thermally driven Hadley cell if we look at the heat transport Here you see the climatological heat transfer in the model, which is quite realistic And here you see the anomalous heat transport again taken as the difference between the two thousand decades and the 190 decade and you see a negative Anomaly in terms of heat transfer in the northern hemisphere consistent with a weakening of the Hadley cell and the response from the SST Forcing so that's all very consistent with what we expect the atmosphere to respond to that force Now we look at the ocean model and we force the ocean model with those Anomalous fluxes and this is and this is the first response that we get we do an EOF of the SST in the North Pacific and we see something that might be similar To a PDO like button. This is a switch on experiment So what we see is a is an adjustment to the Forcing the anomalous forcing that we are putting into the model, which has a roughly a Decadal timescale adjusting to this anomalous force So now we look deeper into the ocean to see what is happening into the ocean yesterday. We've seen a Plot of the meridional overturning circulation in death space now. I'm showing the meridional overturning circulation in density space so this is Potential density the y-axis and this is latitude in the x-axis So here you're supposed to see two subtropical cells Around the equator and this is the anomalies again When you force the ocean model With those anomalous fluxes as compared to the control simulation You see a Weakening of the subtropical cell in the another hemisphere and a reduction in the heat transport associated with that in the non hemisphere as well and the response that we get at the equator is Because the subtropical cells are weakening. There's less heat transport away from the equator less upwelling at the equator And so we get a positive SST anomaly At the equator which is actually damping the original cooling that originated the whole thing So again, you might think What is the ocean really responding to is a responding to the anomalous winds at the equator Or is there responding to the anomalous winds that are north of the equator subtropical and exotropical regions So we did another set of experiments where we are forcing our ocean model only with anomalies north of 18 degrees pole worth of 18 degrees or Only within a band of 18 degrees at the equator and see what is the response of the ocean in that case So those two experiments are called speedy no-trop. So wind anomalies only pole worth of 18 degrees or Anomalies only within the equatorial bond And what we see if we force the ocean model only with the wind anomalies Within the equatorial bond is that we don't get that weakening of the subtropical cells No weakening in heat transport and that not much of a signal in terms of SST at the equator If instead we force only with the wind anomalies pole worth of 18 degrees and so the exotropical wind anomalies Coming from the atmosphere then we recover most of the signal in terms of weakening in the subtropical cells weakening in oceanic heat transport and that warming at the equator which dumps the original signal and You know to see the timescale of this you can look at the heat content anomaly within a box in the equator Which is given there in the top 500 meter. So if you do the Trop experiments or only force in the ocean model with wind anomalies within the equator Then you don't have much of a signal. It's it's a fast response and is actually Trying to reinforce the original anomaly but if you force your ocean model with exotropical wind anomalies, then you get this Increasing heat content because you get that warm anomalies at the equator and actually it spins up On a decadal timescale, okay, and it's quite it's quite stable as a solution. Yeah So now we try to model this already a simple Framework with a with an even simpler Model kind of a toy model and this is starting from the answer delay the oscillator of swaths and shawth so we start with three equations one defining the evolution of the Temperature at equator one is going to be the jar index and another one is going to be the cell Sotropical cell index. So the the hypothesis here is that the C-cephus temperature are forcing the jar index circulation, which Is then connected to the Sotropical cell index. Okay, there's a few parameters. This this is the the well-known part from Swaths and Shawth and there's the The jar index as I said is connected to the temperature at the equator and the Cell is going on the same sense as the jar by definition in this case so with a proper Tuning of parameters depending on on the trying to get a Correct response in terms also of the observations then we get this this kind of behavior in terms of Temperature at the equator the jar index and the cell index. Okay, so this is quite a A Good representation of what answer is in reality and you can see the blue line is the low-frequency Evolution of this T anomaly which shows some the K-doll anomalies and there are by construction anti-correlation Sotropical cell index anomalies in this and this plot of the Sotropical cell index, okay To test whether the connection between the temperature and the jar is really essential in order in this model to generate those Sotropical cells Low-frequency anomalies and C-cephus temperature anomalies then we can play with this model and for example eliminate the connection between the C-cephus temperature at the equator and the jar and so Is the Sotropical connection tropical exotropic connection really necessary to generate that low-frequency variability and in this case It is if you take off that connection Then you basically don't generate any low-frequency variability in the tropical cells and the C-cephus temperature at the equator Okay, so I'll listen this model the connection between the jar and the equatorial temperature is essential in order to Spin up the Sotropical cells so You know that you are put together this two hypotheses the atmospheric bridge and the oceanic tunnel Then we kind of look at the simplest schematic in which you start with anomaly at the equator Which through an atmospheric bridge? Generates some atmospheric response in the exotropics that atmospheric response is able to force the ocean Which responds by spinning upward down the Sotropical cells and then you get a reversal of the original SST anomaly in your in your ocean Okay, and this is this is all based on previous theories in in both the atmospheric bridge and the oceanic tunnel So we're just putting two pieces together here So is this does this make any sense in our in a more real model, and I think it does So this here I get yeah In gray is observations. So gray is the C-cephus temperature evolution Over the past hundred years and the data lines the gray data line is the observations of the surface The Sotropical cell index the one that I showed at the beginning If we force a global the same old global ocean model with time varying Core to force things then we are able to reproduce the Sotropical cells variability in this model's okay So the the model is actually able to reproduce the isotropical cell variability during this period so if we Remove the wind forcing within the aquatoria bond if what I said up until now make any sense Then we are supposed to maintain that low-fringacy variability in the Sotropical cell and we actually do This is the same kind of experiments by just forcing the ocean model with time varying winds north of 1015 degrees north and south of 10 10 15 degrees And so we retain most of the Sotropical cells variability and most of the C-cephus temperature variability in the ocean model so Do couple models semi-fi models are able to reproduce any of this low frequency variability and actually they don't So at the top left you see the reanalysis data the black line is the Sotropical cell index and the Gray line is the C-cephus temperature and all the other models basically Not reproducing any kind of low frequency variability during the historical records So for some reason Sotropical cells in semi-five historical models are not able to reproduce any low frequency variability as observed So conclusions the atmosphere response to tropical forcing does seems to have some feedback on the Sotropical ocean which in turn is forcing an equatorial time delay response at least in this simple framework So we are basically putting together do two things the atmosphere beach and the oceanic tunnel here So also soft tropical force soft tropical cells variability I think is a key player in generating this low-fringacy equatorial Pacific variability and the Sotropical cells are forced by soft tropical winds not equatorial winds and This natural mode, which is just one of the possible low-frequency Genen one of the possible mechanism that generates low frequency variability Can perhaps be another one that explains the warming hiatus together with the local Pacific decadal variability and also the teleconnections Thank you very much