 So, I will continue today with discussion of El Nino Southern Oscillation that is ENSO. Now just to remind you, we did look at the boreal winter that is to say December, January, February in the last lecture. And what we showed was that there is a tropical Pacific wide circuit of air proceeding westward at the surface, rising over the warm region of organized convection and persistent precipitation over the west Pacific, returning eastward at higher levels and descending over the cool eastern Pacific. And this is what is called the Walker circulation and associated with the Walker circulation is the low surface pressure in the western Pacific and the high surface pressure in the east. So, this is what we saw that the boreal winter circulation, boreal winter circulation that is December, January, February circulation. In fact involves rising of air over the west Pacific and sinking of air over the east Pacific and this can be viewed as an east west circulation which is called the Walker circulation. Now, so Walker circulation is an important facet of the atmosphere in the boreal winter. Now let us consider the mean state of the tropical atmosphere over the Pacific in the other seasons. So, consider boreal summer which is June, July, August and if we look at the OLR pattern for June, July, August then what you see is that there is a TCG stretching across the Pacific which has a large latitudinal extent over west and central Pacific up to about 170, a narrower TCG around 10 north over east Pacific connected by a relatively weak connection a convection between 150 west and 120 east. So, if we look at mean June, July, August precipitation there is deep convection here and you can see that it is continuous the belt is continuous from the Indian region and then there is also a zone of low OLR of considerable lateral extent latitudinal extent just off the coast of the Central America and in between here from about 150 west to 120 west it is a much weaker and a thinner tropical convergence zone, but it is very much there we do have OLR below 240 there. So, there is a band of low OLR stretching across the Pacific all the way here, but here the latitudinal extent is very large and here also it is substantive. So, consistent with the OLR pattern during June, July, August a belt with ascent at 500 HPA and rainfall occurs around 10 degrees north over east Pacific. Thus the TCG in this case stretches across the Pacific at 10 north. So, now this is the precipitation rate for June, July, August and again you see heavy precipitation this is a monsoon season, heavy precipitation over Indian region, but now we are concerned more with the Pacific. So, there is very very heavy precipitation over west Pacific there is reasonably heavy precipitation over this part of east Pacific, but in between it is somewhat less this is over the region where the OLR was also slightly higher, but still below 240. Now, in fact the precipitation rate and the ascent velocity of ascent at 500 millibar is very comparable as we expect and we see that there is very weak ascent over this intermediate region reasonably strong ascent over this region and very strong ascent over this part. So, however along the equator and to its south the pressure is lower over the west Pacific then over the east and there is no convection or rainfall over the central and east Pacific suggesting sinking of air above this region. So, consistent with the OLR pattern during JJA we have this belt around 10 degrees north and the tropical convergence zone in this case is stretching right across the Pacific. So, when we look at the OLR or the precipitation zone we would think that this is much more like the Hadley cell. So, you have ascent at this latitude and descent everywhere else. So, this is much more like the Hadley cell circulation we saw where within a small latitudinal belt you have strong ascent and convection and rain and over the rest of the region the air is sinking and so there is no ascent and this is exactly what we see in the precipitation also and you see it very clearly here that you have an ascending zone here and it is surrounded by descending zones here and here. So, this could be thought of as a north south cell or what we called a Hadley cell. So, the circulation over the Pacific in JJA therefore, appears to be more akin to the Hadley cell than the walker circulation seen during the boreal winter. Because of course, you have ascent here and descent here, but if we look at this belt we have ascent here and descent in both these places. So, it does look like more like a Hadley cell. However, along the equator and its south see if we look at along the equator and its south then indeed the west is convecting west has rain and east does not have west has ascent east does not have. So, one can think of having a walker circulation also in addition to the Hadley cell. So, along the equator and to its south the pressure is lower over the west Pacific than over the east and there is no convection or rainfall over central and east Pacific suggesting sinking of air above this region which we have seen see there is definitely sinking of air all around here this whole region has sinking of air. Hence walker circulation could be thought of as persisting in the boreal summer in addition to the Hadley cell. So, we think of the Hadley cell in which ascent is occurring here and descent is occurring in the surrounding region and walker cell in which we have strong ascent here and descent here. So, it is a combination of both of them, but mostly in literature it is only the walker circulation which is emphasized as far as Pacific is concerned. So, what we have here is a picture like this. This is the ITCZ and on the one hand there is this east west circulation which is the walker circulation and on the other hand you have a Hadley circulation which is a north south kind of cell and what it is is a combination of both of them over the east Pacific. Now, let us consider the seasonal cycle because we have so far looked at only what happens in the winter and summer of the northern hemisphere or the boreal winter and boreal summer. So, let us see now how the seasonal evolution takes place and that we will do by looking at monthly OLR patterns. So, this is a snapshot of six mean monthly OLR patterns goes from January here, February, March and then April, May and June. Now, one thing important to see is that around the west Pacific the picture looks sort of similar in all the months. There is variation of a certain kind which I will point out, but basically you have convection over the west Pacific. Now, as far as January, February, March is concerned that is all there is to it. In the tropics the convection is by and large restricted to west of the date line. There is no convection in the equatorial or tropical regions at all on the eastern side. So, this is true for January, February and March. The low OLR region is restricted to this part of the tropical Pacific over this part of the tropical Pacific, what we could call west and central Pacific and there is absolutely no region of low OLR over this part here. So, this is a very important feature and this is what was described the circulation associated with this kind of convection was described as walker circulation with rising here and sinking here, but see what happens from April onwards. From April onwards you begin to see a nice zonal band and east west band of low OLR right here. It is there in April and note that in between here it is slightly weaker than at the other two extremes. Of course, in April it kind of ends here, but by May it has flared up of the coast of the Americas. So, you have two relatively strong convection parts joined if you wish by a relatively weak convecting part, but this is a single band on this monthly scale stretching right across the Pacific. It is a TCZ or an ITCZ in this case, because the convergence which is associated with this is indeed inter tropical. So, we have an ITCZ here, notice also in addition to that we have us what we call SPCZ South Pacific Convergence Zone which is an additional Convergence Zone which we see right from January. In fact, it is very strong in January, February, March it begins to become weaker from April and actually then May and June it actually becomes much less weak, much more weak compared to. So, now we have we have seen here what happens up to May and in May already you see onset has occurred over part of the bay and this region is getting the cloud band here is moving to the north here and here we see a well developed ITCZ. Now, what happens in June? We have this is June, this is July, August and September and July, August are the mean monsoon months and what you see here is actually the tropical Convergence Zone very, very prominent and where specific is also very intense this part is also reasonably intense SPCZ is weaker and you see of course, a lot of convection this is our monsoon. So, the belt has moved all the way to go over the Indian region here by July, August, but this one remains pretty much in the same place and September also. Then October it still is persisting, but now it is weaker and over Indian region now things have come to the south. November now this is a bit like April where the band does not quite reach Americas and by December this is beginning to evaporate this is beginning to go away by January this whole convection will disappear. So, in the nature of the patterns during January to March is rather different from that in the rest of the months. In January, February, March while there is intense convection over the west Pacific there is no low L R region east of 150 west over the tropical Pacific this is what we saw here that there is absolutely no convection east of 150 east of this region. There is no convection in the tropical belt at all in January, February, March and also east of the dateline north of the equator. So, the convection is not there over the eastern part thus for J F N the prominent feature is the east west asymmetry in convection and the walker circulation is a large part of the story. So, you can see here that for January, February, March large part of the story is that air is rising here and air is sinking here and we could conceive of this as a cell east west cell in which there is a scent here and there is descent here. So, for the boreal winter which is January, February, March the east west asymmetry in convection and walker circulation is a large part of the story. However, in the other months the prominent feature of the O L R over the region east of the dateline is the zonal band of low L R stretching across the Pacific around 5 degrees north associated with the ITCC. Thus in these months the east west symmetry is restricted to the equator and south tropical Pacific and this is what we have seen here that in these months see if you look at this one then we can say that the air is rising here, but really it is sinking only over this part particularly when we go to the peak months July and August you see that the air is rising over this entire band and this looks like it is a north south circulation. This is an important point which is often not emphasized in the literature. So, in other months the prominent feature of the O L R over the region east of the dateline is the zonal band of low O L R stretching across the Pacific around 5 degrees north associated with the ITCC. Thus in these months the east west asymmetry is restricted to the equator and southern tropical Pacific is the way you could put it. Now of course, it is important when we say we want to understand the seasonal variation of tropical convection over the Pacific what we would like to know is why is there no convection over this part in January, February, March how does it appear there in April what decides on this kind of seasonal variation across the from month to month. So, this is what we would like to understand now how do we understand this see we will begin to understand it by looking at the sea surface temperature patterns why is that because the SST and its spatial variation play a critical role in determining the location of the TCCs over the tropical oceans. See the key factor which determines where the convection will be where the low O L R region will be the key factor is SST and its spatial variation. Now it is therefore, important to understand the nature of the relationship of organized convection or rainfall over the tropical oceans with the local SST and we have touched on this problem before, but I would like to mention here that in fact, this the nature of the relationship of rainfall over the central Pacific to the local SST was first elucidated by Berkness in 169 by analysis of data for Canton Island. This paper of Berkness is one of two papers he wrote which have turned out to have a enormous impact on our understanding of ENSO. It is interesting that in this paper he addressed the question of the relationship of rainfall over Canton Island to the sea surface temperature. Now the central Pacific region to which Canton Island belongs it is also considered part of the line islands there are about 6 or 7 line islands and line islands precipitation data is also available and has been analyzed. So, Canton Island is one of those islands in the central Pacific you can see it is at 172 east and 3 degrees south very close to the equator and these places these stations island stations have enormous amount of inter annual variation that is in most of the years there is very little rainfall and then suddenly in a few years there is copious rainfall. You can see that by looking at what is the mean see this is the median and this is the mode this is January, February, March and this is all rainfall in centimetres. Now, so the mean mean January rainfall is less than 5 centimetres you can see that here and or median is less than 5 centimetres mode is also this, but there is this extreme highest recorded in 25 years is 51 centimetres. So, it is many many times the mean and the same is true for all these months you have huge amount of rain occurring there towards the end of the year December, January and even in other months of the year the rainfall can be substantive, but most of it is in the in the months of at least the highest rainfall seems to be from November to about January. So, this rainfall distribution is very skewed most of the time you have hardly any rain and then some years you get a lot of rain. So, variation of the monthly rainfall at Canton Island the SST and the air temperature from 50 to 67 from Berksness study is given here and what you see here is rainfall SST is solid and air temperature is that we will not worry too much about air temperature right now. SST is solid and I have marked here what corresponds to about 28 degrees here. So, this is about 28 and this is the rainfall recorded at Canton Island. So, by and large you see the rainfall is very very small you see these little little bars here, this is the 10 centimeter line. So, the rainfall is rather small over most months, but suddenly it rains a lot and you see what has happened this is where SST is less than 28, but suddenly when it is well above 28 you get this enormous amount of rain. Similarly, again here SST is lower and there is no rain at all, but if you go here again SST is high and you get enormous amount of rain and more over it is sustained for several months you see that it is sustained for several months during this and again now you see SST is very high. So, this is what happens when you see it rains a lot and you see what has happened this is where SST is less than 28, but suddenly when it is well above 28 you get this enormous amount of rain. Similarly, again here SST is lower and there is no rain at all, but if you go here again SST is high and you get enormous amount of rain and now you see SST becoming cold and rain has been suppressed generally below 10 centimeters. So, generally the monthly rainfall is less than 10 centimeters, however there were periods in which monthly rainfall was sustained at a high level well over 10 centimeters for several months and this occurred when SST was above 28. So, this I think although he does he did not call it a threshold and he did not actually discuss this in too much of a detail in the paper. He was only concerned with the impact of SST on rainfall which is very clearly brought out here, but actually the concept which came to be known as presence of an SST threshold was first discovered by Birkenness in this study. So, presence of the SST threshold for organized convection of about 28 degrees centigrade about which I have discussed at length in another earlier lecture was first demonstrated by Birkenness with the data from Canton Island in the Central Pacific. And the way he put it was he said most of the time water of equatorial upwelling origin now you remember from the last lecture that there is a tongue of cold water on the equator. And towards that kind of the limit of that is around the on the Central Pacific. So, most of the time this water of cold equatorial upwelling origin reaches Canton Island and it is in the cold SST regime that is well below 28. But the major maxima of SST in late 57, early 58, late 63 and late 65 are proofs of the occasional elimination of the upwelling process. This is what he says and it is during these that rainfall increases very much when SST becomes high. So, Birkenness study of the relationship of the rainfall over the ocean to the SST was very illuminating. But it was after all relationship at one point relationship of the rainfall over one island to the SST of the region surrounding the island. Now, before the satellite era we did not have any objective measurements or synoptic measurements of the clouds over the ocean. They became available only over in the satellite era and later on even direct measurements of tropical rainfall became possible with satellites. So, systematic investigation of the variation of convection or rainfall over the tropical ocean and the relationship of this convection or rainfall with the local SST or sea surface temperature of the ocean above which this convection is occurring became possible only after the availability of satellite data. I am not going to go into details here about the different studies, but just summarize what was learnt. So, analysis of the variation of different measures of satellite derived convection it all began with subjectively derived satellite measure of cloudiness intensity as Sadler called it. Then came the OLR which we have used very often outgoing long wave radiation data. There was also data on frequency of highly reflective clouds because Albedo of a cloud depends on how high the cloud top is and how deep the cloud is. So, highly reflective clouds is a data set that was compiled also and that also was analyzed by Walliser and others. What is very nice is that the result that was originally derived from cloudiness intensity turned out to be so robust that similar kind of nature was relieved for all these different measures of convection which became more and more fancy. In fact, in the discussion there we have even shown the results of the latest measurements with cloud set which is a high resolution cloud measuring satellite. So, the relationship of convection with SST of the ocean beneath all led to the same conclusions about the nature of SST convection or SST rainfall relationship. And what are the important facets of this relationship? We have to remember this because we are going to apply what we have learnt to understanding the seasonal variation. Now, such scatter plots were generated by for various indices. What I show here is a scatter plot for the Indian Ocean region of the OLR outgoing long wave radiation and co-located SST for July only of 82 to 98. So, several years are here. Now, notice that the OLR is actually plotted in a reverse direction. So, it increases downward which means that as you go up convection increases, OLR decreases. The radiating surface is going higher and higher in the troposphere. Now, what has the major by the way the size of this star indicates how many points are in this particular bin. So, for example, this corresponds to a bin with 29 degree centigrade SST and this is 180 OLR. So, this many the points here are much larger for example, than the points here and here there are 0 points, here there is few points and here there are more points. So, size of the star tells you gives you an idea of the number of points in these bins. And it is these kind of scatter plots that give you an insight into the relationship between the two. So, what do we see then? This is 240 which is a reasonable threshold for identifying organized convection, deep convection in the tropics. This is 240 watts per meter square. So, above this line we can say there is considerable convection. Now, and this is the threshold of 28 degree centigrade for SST and what do we find? See there is a whole lot of points here. So, below 28 the points tend to have high OLR higher than 240. So, there is high propensity for high OLR for SST below about 27 or so. So, this is this part here below about 27 see there is there are very few points with SST low with OLR lower than 240. So, all of this is below 240 below the line of 240 which means OLR is higher than 240. So, these are all relatively non-convective systems and so there if one is below 27 there is a very high likelihood high propensity for high OLR whereas, if you go above 28 then there is a very high chance that OLR will be less than 240. So, very high propensity for occurrence of deep convection or occurrence of low OLR for SST above 28 and you see a market change in the distribution across the SST threshold of 28 here. It is also important to remember that once this threshold is crossed once the ocean is warmer than 28 then it does not mean that you will always have a lot of convection you still have points here. There is a whole range for any given SST which is higher than the threshold OLR varies over a large range from close to 240 to very very low values where which implies very deep convection over large regions. So, there is a very large variation for a given SST once the threshold is crossed and this is why we said SST being above the threshold is a necessary condition for deep convection, but it is not sufficient. Now, this is the mean OLR versus SST again for the same data and what you see is that the mean is increasing somewhat gently here and when I say mean is increasing I mean convection is increasing OLR is actually decreasing very sharp increase in convection or decrease in OLR occurs across the threshold of 28 from below 27 below 28 to above 28 this part has very sharp increase in the mean and then you can see that above about 28.5 it flattens out this is where you know what the mean OLR or the mean convection has no relationship with SST. So, there is a sharp decrease of mean OLR with SST in the range about 27 to 28.5 around the threshold. Now, if we look at the frequency distribution for specified SST ranges that is to say for pixels these are all 5 degree by 5 degree regions and remember that there are lot of pixels for every July and there are several Julys. So, all these are added together in this analysis and what we find is for example, when the ocean is for pixels which correspond to cold SSTs this is 24.5 to 25.5 vast majority of the of the pixels have very very large OLR. Now, actually the OLR starts to decrease as the SST increases, but what is remarkable is the change in the distribution across here because the mode has always been less than 240 rather larger than 240 for OLR for SST below 27.5 suddenly it shifts to values which are which indicate deep convection or OLR of the 40 or so. So, here the mode is at 250, here the mode is at 260, here the mode is at 270 watts per meter square. So, 270, 260, 250. So, still the mode implies that most of those pixels are not convecting, but across the threshold when we suddenly go to this 27.5 to 28.5 then we find the mode has shifted to this this is 220 and this is 230. So, 230 and 220 and this is 210. So, it has suddenly shifted to convective things and it in fact shifts further towards lower OLR a little bit and then remains more or same. Of course, when you go way beyond 29.5 there are not that many pixels itself. So, it is one has to take this with a pinch of salt. So, what are the major results that there is a high propensity for organized convection over warm oceans with SST above about 27.5 or 28 called the SST threshold. Now, while SST being higher than the threshold appears to be necessary for organized convection over the tropical ocean, it is not a sufficient condition. We saw this and that is why there is such a large spread of the OLR for such SST with several points having high OLR as well as several points having low OLR. So, there is a high propensity for organized convection once the threshold is crossed, but if the threshold is crossed there is no guarantee that you will have organized convection you may have no convection at all. So, you have a variation between no convection to intense deep convection and Graham and Barnett showed that over such warm oceans convection is determined by whether there is convergence in the atmosphere or not. In other words SST has to be above the threshold, but once it is above the threshold also unless there is convergence in the atmosphere which leads to ascent of air from near the surface of the ocean there will not be any clouds. So, dynamics favorable dynamics is very important and therefore, if this is absent then a part of the ocean warm ocean where SST is above the threshold will be without organized atmospheric convection if the dynamics is not favorable. Unless there is convection you will not get unless there is convergence there is no convection. Now, so far we have been talking of satellite derived measures based on clouds, but now satellite derived rainfall is also available and the SST rainfall relationship can also be studied. Now, I am now going to focus a little bit on the Pacific ocean which we are discussing right now because that is where ENSO occurs and so let us consider the relationship of the rainfall over two parts of the Pacific with local SST. Now, first is a key region for El Nino this is the so called Nino 3.4 region which is 170 west to 120 west and 5 south to 5 north this is the region of the central Pacific you remember Canton Island was very much in this region it was 170 west and 3 south I think and warm part of the tropical west Pacific which is 120 to 140 and 10 to 20 for June to August. So, this is the Nino 3.4 region central Pacific and again what you see here is a scatter plot only thing is the number of points is now indicated by colors the black and the gray correspond to many many points and the blues somewhat less number of points. So, this is done for June July August for several years and it is all pooled together and what you see is what you saw before there is a threshold of about 28 before that by and large most of the rainfall is less than 3 millimeters per day it is only above that that things change a bit and above 28 you get very long tails here with very heavy rainfall amounting to 12 millimeters per day or even more. So, now let us look at this Nino 3.4 region this is the frequency distribution of the SSTs and what we find is that a large number of points here are just around the threshold see this is 28 and a large number the mode is just around the threshold here on the other hand if we look at the west Pacific all the points are above the threshold of 28 and we see no relationship between rainfall and SST it is just a huge blob like this indicating there is no relationship between SST and rainfall whereas, here if we plotted the mean and we would get a relationship because once the threshold is crossed you see this minimum also lifts up. So, once the threshold is crossed then you always get some rain you always get some rain once the threshold is crossed that is very clearly seen here. So, the differences in the nature of variation of the SST of these two regions are clearly brought out in the observed frequency distributions of SST which are also depicted there. Note that for the Nino 3.4 region the highest frequency of occurrence is for SST just below 28 degree centigrade. So, this is what we saw here in fact this is 28 here. So, the mode is for SST just below 28 and this is just above 28. So, between 27 and 28 we have a large number of points here of the order of 30 percent of the points are between 27 and 28 and you have substantive number just above 28 also about 15 percent of the points. So, highest frequency of occurrence for this Nino 3.4 region is just below SST of 28 on the other hand for the part of the tropical way specific considered SST is always well above 28 and that we have seen this is 28 and there is not a single point where this region is below 28. So, SST is well above 28 for the entire region. Now, the mean SST for Nino 3.4 is 27.14 for June, July, August. Hence, we expect positive SST anomalies to have a large impact when the threshold is crossed and typically we have seen the SST anomalies of the El Nino. We have seen those before in the introductory lecture to El Nino and there of the order of 1 degree or even more. So, with such anomalies you will get this threshold to be crossed. In fact, we saw that the threshold was actually threshold of 28 was actually crossed for the El Nino that Birkenes looked at in his figure for Canton Island. So, for if we have positive anomalies, we can have a large impact and the rainfall will increase. If the anomalies are negative on the other hand, then the rainfall remains suppressed. So, this is the asymmetry and I will come to this later, but this is the problem if one looks at only anomalies and not at the actual SST. One can be misled because the response is not symmetric. We are saying that a warm anomaly or a positive anomaly will enhance rainfall, but negative anomaly does not depress rainfall. Rainfall is anyway very small once it is below the threshold. So, there is no depression with negative anomaly. So, the response is not in any way antisymmetric in terms of actual rainfall because what matters is the actual SST and not the SST anomaly which is a creation of our own thinking. We subtract the mean from the SST and then go on looking at anomalies to see how things are different from average. Now, that can lead some give some insight, but it can be misleading if we try and look at SST rainfall relationship. Now, since a large part of the variation of the SST of the Nino 3.4 region is in the critical range around the threshold, the relationship of the rainfall with SST is strong and very high rainfall say greater than 7 millimeters per day occurs only when SST is above 28 degree centigrade. So, as I mentioned before when we look at Nino 3.4 then very very high rainfall say 7 or so which is here will occur only above the threshold. See this cap, this peak of the distribution is entirely above the threshold also above the threshold you will always have non-zero rainfall that is another thing which is important. This occurs only when it is above the threshold. On the other hand over part of the way specific for which SST is always above the threshold the variation of rainfall base no relationship with local SST. This is also important to remember. So, SST variations can have a very large impact of on convection or rainfall of the atmosphere if the SST variation is around the threshold. If it is always above the threshold then you cannot say very much on the basis of SST because dynamics will then determine whether you have convection or rainfall. On the other hand if it is very much below the threshold then whether you have positive or negative anomalies of SST you will get no rain at all and therefore there is no relationship between SST and rainfall. It is only in the intermediate range when SST varies across the threshold that we see very strong relationship of the rainfall in the atmosphere with the SST over the ocean, SST of the ocean over which it is raining. Now, this is necessary to revise this background on the relationship of convection or precipitation with SST in order to understand and interpret the seasonal patterns as well as the monthly patterns. So, let us begin with the simpler ones the seasonal pattern. So, during the boreal winter the low oil region is prominent over the west part of the and part of the central Pacific and over the region 120 is to 170 west with the SPCZ more intense than the ITCZ to the north. Note that this OLR region now sorry. So, this is the boreal winter. So, you have convection going here this is the ITCZ and this is the SPCZ. So, the low OLR region is prominent over west part west and part of the central Pacific and there is also SPCZ. So, within the low OLR region see this is part of west Pacific it is this is the date line. So, it comes right up to central Pacific and in addition there is this SPCZ which is sort of going this way and just notice how close the pattern is to the warm SSTs remember the yellows are 27.5 and here we see very close correspondence that the low OLR region is always within the warm oceans. It is always over the warm oceans you never see low OLR over cold oceans that is very clear, but you can have warm oceans there are small bits here over which there is no convection and that you can see partly here as well. So, in the Indian Ocean as well as Pacific what you see is that the low OLR region is over the warm parts of the Pacific with the SST well above the threshold for organized convection. See this is where these dark regions here are where SSTs well above the threshold and that is where the low OLR region of the Pacific is. So, this threshold does in fact determine the special patterns of OLR on the other hand the SST is below the threshold over East Pacific and in the southern hemisphere and OLR is high over the region. So, you can see here OLR is well above 240 and you can see that SST is below the threshold in this equatorial tongue as well as the East Pacific south of the equator. So, here it is well below the threshold and there is no convection at all. So, during DJF the SST is above the threshold only for very small regions near 5 and 15 and the OLR is not low over the small patches of warm water this is what we saw here these are the little patches of warm water and low OLR does not occur there they are perhaps too small for organized convection. Now, we come to the this was the boreal winter now we come to the boreal or northern hemispheric summer and here we see that for June, July, August again now you see sea surface temperature is very warm over this whole patch and a large part of that is covered with low OLR region or convection. Now, you also see that sea surface temperature is warm also over this part here this is east of about 120 west and there again you see fairly you see a low OLR region with reasonable latitudinal extent here. Now, in between you see this is a relatively thin strip of warm SSTs and on that part of that is actually a thin strip of low OLR here. So, again here also we see a rather good correspondence that organized convection over the Pacific and also over the Indian Ocean in the boreal summer is just like in the boreal winter is entirely confined to warm oceans with SST above the threshold the low OLR region is prominent over the west and part of the Central Pacific. I mentioned also Indian Ocean you can see here here also the Indian Ocean is warm in this patch and you can see this is the low OLR region almost over that patch identical. So, the shapes are determined to a large extent by the SST itself. So, in addition to the west Pacific low OLR region we already saw there is a relatively narrow band zonal band stretching eastward around 10 north from the low flow OLR region over the western Central Pacific up to 120 west and then a low OLR region with a larger latitudinal extent over the coastline of America. So, we have this huge region huge blob of low OLR with intense convection then a thin TCG and somewhat larger latitudinal extent characterizing the convection over this region. This zonal band is located where the SST is maximum. So, low level convergence is expected to occur and hence conditions will be dynamically favorable also. Now, that is clearly seen here you can see here that the SST is maximum in this zonal band. If you go north and south then SST is colder. So, if you have warm SST here and cold SST here because of the gradient right warm air being lighter you will have lower pressure here relative to here and this will lead to convergence here from the north and south and that should lead to convection since SST is already above the threshold. So, the dynamics is determined by the location of the maximum SST and dynamics is also favorable here and the threshold is crossed. So, we have in fact this zonal band here this is important to remember the low OLR region is prominent over this and. So, the zonal band is located where the SST is maximum and so there is low level convergence is expected to occur there and conditions will be dynamically favorable also for convection. Now, note that the convection is restricted to a narrow band and is relatively weak where the SST is not much higher than the threshold this is the region 150 west to 120 west. In fact later on we will see see this region here this is 180 this is the date line and this is 120. So, in between see this region in between where OLR is just below 240 and where you can see there is only a thin line above 28. It is this in fact it turns out that the major difference in the OLR patterns associated with El Nino and La Nina are presence or absence of convection in this region here 150 to 120. We will see that later so far we are looking only at the mean patterns. So, later on we shall see that a major variability in convection occurs over this region between El Nino and La Nina. Now, we note that as in DJF SST is below the threshold over the east specific in the southern hemisphere and there is no convection over the region. So, this is something we notice see how low the SST is here over this region and there is no question of convection it is all right there. So, this is something that we saw in the other season as well. Now, we will look at so far what we have done now is shown that the pattern of low OLR regions which are the regions where there is organized convection can be understood in terms of the SSTs. In other words SST patterns of where SST is above the threshold appear to determine to a large extent where the convection is going to occur. So, this is a very strong relationship and the fact that SST has to be above the threshold seems to be a very rigid constraint on where organized convection can occur. So, this is very clear from the seasonal patterns we have seen. Now, we will go to monthly mean patterns and see how the evolution occurs as far as the mean climate is concerned from the boreal winter through boreal spring to boreal summer and then austral spring and austral summer or our autumn and our winter. Now, it is interesting that we see that, but it is important to now we I did mention that one thing to notice in the seasonal patterns is this is the JJA. You see this low OLR region as a coherent zone. Now, here it is of a enormously large latitudinal extent, but you see there is no gap between the two. It is coherent between the Indian longitudes and then stretching across the entire Pacific and in DJF of course, the story is somewhat different, but in June, July, August we see it very much as that way and then we will see what happens on the monthly scale, but as far as June, July, August is concerned then we see that the low OLR region is certainly coherent between Indian longitudes and the Pacific and it would be surprising if the variability of OLR or rainfall over the Indian region is not linked to variation of convection or rainfall over the Pacific. So, it seems to me that we are looking at a part of a gigantic planetary scale system. We are looking at one part and obviously, other parts of the system are going to have an impact in terms of the variability of the monsoon. So, this is something we expect. Now, next we will look at monthly mean OLR patterns, how they evolve and then go on to see the characteristics of El Nino and La Nina. Thank you.