 Good afternoon. We have in the last lecture looked at what are the mean rainfall patterns like over the Indian region as well as the all India rainfall. In this lecture, we look at the problem of nature of variability of the Indian monsoon by looking at what conventional as well as satellite data have told us have shown us. So, today's lecture the outline is first we will consider the inter annual variation of the Indian summer monsoon rainfall that is to say the year to year variation of the all India average of the summer monsoon that is to say June to September rainfall. Then we will talk of variation of the various facets of the monsoon we considered last time. We consider the mean date of onset over Kerala. So, now we look at variation of the onset date over Kerala and variation in the advance of the monsoon. We will also look at active spells and weak spells or breaks. Now in any year the rainfall over the Indian region differs from the seasonal mean which we have seen in the last lecture and the dates of the onset retreat also vary from year to year. In addition within the peak monsoon months of July and August when the monsoon zone has the monsoon established over it. In fact, wet and dry spells occur wet spells with continuous rain for several days interspersed with dry spells relatively dry spells with varying duration. Now understanding and predicting the variability of such facets is the central problem of monsoon metrology that is to say understanding and predicting the year to year variation of the Indian rainfall average as a whole over the Indian region as well as average over different spatial regions as well as intra seasonal or variation within the season between wet spells in which you have continuous rain for several days and dry spells which may be relatively dry spell in which you have relatively less rain or it could be totally dry spells in which you have no rain at all. So, let us consider first the inter annual variation of all India summer monsoon rainfall. Remember this is what our finance minister Pranab Mukherjee was concerned with when he talked about praying to Indra for a bountiful monsoon. What he meant was that for India as a whole the rainfall should be good during the summer monsoon that is June to September. Now long term mean of the Indian summer monsoon rainfall is about 85 centimetres. Now consider first the variation from year to year of this and what we always look at in metrology is so called departure from mean or the difference between the actual value and the average value this is called also anomalies. So, an anomaly is defined as the actual value in any year of say in ISMR that is Indian summer monsoon rainfall minus the mean which is about 85 centimetres. Now what you will see in this slide is this anomaly of rainfall for each year is plotted as a percentage of the mean rainfall. So, when the anomaly is positive that means in that year say for example, this year here 1878 1878 in fact and the anomaly is about 15 percent or 14 percent or so and it is positive which means then 1878 1878 rather the rainfall received ISMR was above normal and above normal by about 14 percent of the mean or 14 percent of 85. So, when the sticks are above this line that is positive values of anomaly that means rainfall was in excess of the average when the sticks are below this this means rainfall was below the average. Now it turns out one measure of course of variation of any quantity is the standard deviation right. You have a set of values in this case going all the way from 1876 to 2010. So, 100 and odd years of values and these values will have a mean which I said is about 85 or so and they would have a standard deviation. So, a natural measure of variability is the standard deviation. Now it turns out that for Indian summer monsoon rainfall the standard deviation is about 8.5 centimetres or it is about 10 percent of the mean. So, whenever we have deficit rainfall that is to say rainfall is below the mean below normal or below the average value and the deficit is very much larger than the standard deviation that is to say larger than 10 percent of the mean that is when we call it a drought. So, a drought year in which ISMR anomaly is large and negative and larger than 10 percent of the mean or larger in magnitude than one standard deviation. Opposite end you will have excess monsoon years and for excess monsoon years like 1961 which is when the highest rainfall occurred. So, in 1961 ISMR anomaly was positive and actually the rainfall was 20 percent in excess of the mean. So, this is the largest excess that has occurred. So, now what do we see here that there are fluctuations in ISMR from year to year and in some years you have droughts these are marked by big red lines and in some years you have excess rain which are marked by green line. So, red line is drought and green line is excess. Now so what you see here is how the ISMR or the all India summer monsoon rainfall has varied from year to year from 1876 onwards it has fluctuated between above normal and below normal and between deficit below normal is deficit and between droughts and excess seasons. Now what is very interesting in this graph is if you see in fact you see that these droughts are sort of clustering around in some patches that is to say if you consider this epoch here between 1899 to 1920 there were many many droughts and 1889 to 1920 is a 22 year epoch 21 year. So, in 7 out of these 21 years there were droughts. Now similarly you had very frequent droughts here from 65 onwards we have been having very frequent droughts 65 to 87 we had 10 droughts in 28 years. Again from 2002 we seem to have hit a bad patch we had a drought in 2002 another in 2004 another in 2009 and 2012 has been close to a drought though actually not a drought in this years monsoon 2012 actually the deficit is about 8 percent. So, it is rather close to a drought but not actually a drought. So, what is interesting is that there seem to be what we call these decadal scale variations or epochs that is to say several decades in which we have a high frequency of drought interspersed with decades in which the frequency of droughts is very very low. So, in some sense the inter annual variation of anomaly of ISMR can be thought of as variation of frequency of what we call droughts or excess rainfall season or extremes of ISMR. So, this is the story of the inter annual variation from the largest longest time series. As I said before the standard deviation of the ISMR all India monsoon rainfall is 10 percent of the mean and so summer monsoon seasons with ISMR smaller than 90 percent of the mean that is deficit more than 10 percent in magnitude are called droughts and ISMR with anomalies greater than 10 percent of the mean that is excess greater than 10 percent or rainfall greater than 110 percent are called excess rainfall seasons these are the extremes. Now, in fact the Indian monsoon is rather reliable and we will go back to this slide and see how reliable it is. See the worst drought is less than 30 percent deficit and the best rain is about 20 percent. So, actually it is a fairly reliable thing in many languages in India there is a saying that every year we do have a rainy season the monsoon does come every year and that is very true. Sometimes the strength is high and you get you know ISMR positive anomalies or excess rain above normal up to 10 percent sometimes it is very large in deficit, but it is still only 30 percent of the mean it is not as if there are years in which the monsoon never comes it is a very very reliable feature, but it is variable this has to be borne in mind that it is one of the most reliable facets of tropical circulation. Now, although the amplitude of this inter annual variation is not very large because the standard deviation is only about 10 percent and the maximum range of variation we have seen is about 40 percent of the mean. So, the variation cannot be considered very large, but even then it has a very large impact on the food grain production of the country and on the gross domestic product GDP. Now, I will give a separate lecture in this series on the impact of monsoon on agriculture and GDP and there we will see how large the impact is. So, the even today the monsoon continues to govern the very pelts of life in the country if it turns out to be a normal monsoon like it did in 2010 and 2011 the nation heaves a sigh of relief and carries on with business as usual, but if it turns out a drought like the last severe drought was 2009 and it took a long time for us to recover from the effects of that drought major drought relief programs are launched. Thus the importance of prediction of ISMR cannot be over emphasized this is why people are very keen to find out what the future holds for us what is the next monsoon going to be like. So, we never take the monsoon for granted although it is a reliable facet and every year as the heat scorches the countryside in May we eagerly await the onset of the monsoon. Now, as I mentioned earlier the first important event of the drama which is repeated year after year of the monsoon is the onset of monsoon over Kerala which is associated with a dramatic increase in rainfall of South Kerala. Now, in common parlance the term monsoon onset in fact refers to the onset over Kerala. Now, just like we never have any year in which monsoon simply did not give any rain obviously, there is no year in which the onset did not occur. So, onset of the monsoon is also very reliable phenomena and it occurs year after year without fail there has never been a year in which the onset has not occurred. Yet the announcement by the India meteorological department that the onset has occurred over Kerala is invariably greeted with great joy in the country and even the stock markets respond at once this is a really amusing that the monsoon seems to have so much impact on the pulse of the country. So, you remember this is what we saw last time this is the monsoon onset over Kerala and this is what we are going to look for how what is the variability of this onset over Kerala like how does it vary from year to year. In fact, there is considerable variation as in this facet in every facet of the monsoon there is considerable variation from year to year. Now, if we look at the 100 year data then 100 year mean of the onset date is first of June and its standard deviation is about a week 7.4 days. Now, this is the inter annual variation of the date of the monsoon onset over Kerala and you can see that it has varied all the way from mid June to very early around mid May. So, the range is quite large, but the fluctuations are not very high amplitude they are close to the normal and you will see more about that. So, this as you see is the date of the onset and this is the year and this is plotted only up to 2000. So, you can see for example, in from 1900 to 90 to 2000 it did not vary too much it varied between end of May 20 something of May till about 6th of June about 2 standard deviations also less than 2 standard deviation. So, this is now the frequency distribution for the date of monsoon onset and this is from 30th May to 2nd June this is 3rd June to 5th June. So, how many years what percentage of years had the onset date between this period 31st May to 2nd June this is where the mean is right first June is the mean date. So, this is the frequency distribution of monsoon onset over Kerala and you see that the mode is actually close to the mean in other words most likely it is going to occur between 31st May and 2nd June that is most likely, but there is quite a spread and you have one here between 16th and 20 1st May and another between 18th and 20th June. So, this is the variation in the onset date of Kerala and this is what we are trying to predict for a specific year where will the onset when will the onset be where will the point lie in this frequency. So, as I said in only about 25 percent of the years onset date is close to the mean that is during 31st May to 2nd June it is still the mode. So, the most probable date of monsoon onset is still within this 31st May to 2nd June, but actually it occurs only on 25 percent of the occasion in about 50 percent of the years it is between 28th May and 5th June. So, that is a bit of a better window if we see for about a week that is few days before and few days after 1st of June then almost half the years have onset date within this period. As I said the earliest onset was on 11th May in 1918 and the most delayed onset was 18 June in 1972. Now, in India we consider the onset of the monsoon over Kerala which is the commencement of the rainy season as a very important event because it is the first act of this play as I called it. So, many people are under the impression that somehow when the monsoon onset occurs over Kerala holds the key to the future course of the monsoon for that season. So, it is pertinent to ask the question to what extent does the performance of the monsoon in the season or what we called all India monsoon rainfall ISMR. To what extent does ISMR depend on when the onset of the monsoon occur over Kerala. So, what you see here is plotted all India rainfall which is the ISMR as we called it all India summer monsoon rainfall and this is the onset days starting from 31st May. So, this is for all the years in the last century 1901 to 2000. So, you have so many days in which the onset occurred after 31st May so many years in which it occurred before 31st May and what you see here is that there is no relationship between what will happen to all India rainfall for the season and the onset date that is very clear because given any onset date say here 5th of June you know the rainfall can be anywhere between 70 to 95. So, it can be varied this is the mean rainfall. So, it can be deficit and with equal probability it can be excess and this is true for any onset date that you take. There is a large spread of all India monsoon rainfall values for every date of onset and not surprisingly if you still want to compute a correlation of something like this it comes out to be not at all significant it is minus 0.097. So, it is not significantly different from 0. In other words what data have shown us is that even if the onset is delayed one does not have to really worry too much about what will happen to the monsoon as a whole. So, ISMR is not related to the onset date over Kerala. In fact, in the two cases in which onset close to 20 days before and after the mean date ISMR was same let us go back and see that in the two cases in which here you are this is the spectacular thing see this is a case in which ISMR was almost close to 20 days before the 31st May and this is a case when it was almost 20 days later than 31st May and yet you see the all India monsoon rainfall is the same for these two cases. Another manifestation of the fact that there is no relationship between all India monsoon rainfall and the onset date. So, while the event is of great meteorological interest when it occurs is not important for the Indian summer monsoon rainfall. Then you remember we talked of after the onset over Kerala occurs the advance of the monsoon northward and westward propagation of the monsoon northward advance and westward advance of the monsoon and by 15 July it is set. Now you may ask the question these are mean dates after all mean date here is 1st June and mean date here is 15 July and therefore in the mean that is if you look at average over all years it takes about 45 days for the advance of the monsoon from here to here ok. But from year to year the advance varies the when it onset occurs on Kerala varies how fast it progresses northward varies how fast it progresses westward varies. So, all these dates vary from year to year ok. So, because of that the total time taken for advance which in the mean case was just 45 days from 1st June to 15 July actually varies from year to year. Now we have already seen that the all India summer monsoon rainfall is not at all related to when the onset occurs over Kerala there is no relationship. Next question is is it at all related to how long the monsoon took to advance from Kerala to the north western corner here. So, this is the question we ask and when we look at the data. So, the normal duration of the advance phase is about 1 month if you look at this part, but if you look at the final destination of the monsoon it is about 45 days. However, there is considerable variation in the duration of the advance phase from year to year. The relationship between isomer and the duration of the onset phase onset phase begins with the onset over Kerala and ends with the onset over north western parts of India. So, the duration of the onset phase is the time it takes for the advance of the monsoon and now we want to ask the question does the total rainfall over the Indian region during the Indian summer monsoon depend on the on how fast this advance took place or the days between onset over Kerala and onset over the north western part. Now actually it turns out that there is some relationship between the number of days it took to advance and the all India rainfall. By and large if it is a very fast propagation you tend to get above normal rainfall if the advance is very rapid you get above normal rainfall. If the advance is slow actually the situation is not very clear you seem to get above or below, but if it is extremely slow you are more likely to get below rainfall than above rainfall that is very clear. And so, the correlation is 0.345 it is a negative correlation of course. So, Indian summer monsoon does depend on how long the monsoon took to reach its destination in the north west of India after starting from Kerala. Note that when the advance is faster than the average the probability of above normal ISMR is very high we saw that. In fact, so if the rate is faster than normal then above normal rainfall probability is higher than below normal rainfall probability, but if it is slower than normal then the picture is not. So, clear when it is slower than average the chance of ISMR being below the average is only somewhat more than that of ISMR being above the average. So, there is not much signal there, but if the advance is fast then you can guess that ISMR is likely to be above average. So, far we have seen the variability of two important events of the monsoon. One is the date of onset over Kerala and we have also seen all India monsoon rainfall is not related to the date of onset over Kerala. The second facet we saw of the monsoon was how long it takes the monsoon to advance from Kerala to the north western region or how long is the onset phase the duration of the onset phase. So, it turns out that the monsoon does depend on the duration of the onset phase, but more so when there is a rapid advance of the monsoon. So, that the duration is less than average then you are more likely to get above normal rainfall. If it is a sluggered case if the monsoon takes too long to advance it is difficult to say which way the monsoon rainfall will go it can be positive ISMR anomaly can be positive monsoon can be above normal or it could also be below normal. So, these are two important facets of part of the monsoon season that we call the transition. This is the transition to development of the full-fledged monsoon and this is the onset phase. Now, at the end of the onset phase what happens at the end of the onset phase the rain monsoon is established over what we call the monsoon zone. And let me see if it is here it is not here it does not matter we have seen it in the last lecture the monsoon zone. So, it gets established in the monsoon zone which is the seat of the major rain belt by beginning of July around the beginning of July and even after it is established there it is not as if it rains there every day. If we looked at the monthly mean picture you would think oh there is rainfall there in June, July, August and September. Remember that is what all the stations in the monsoon zone showed that they get rain in June, July, August, September that is very true if you total the rainfall in the month. But if you look at daily rainfall nowhere is it true that it rains day after day during the rainy season or during a monsoon. In fact even in the monsoon zone during the peak monsoon months after the monsoon is established in the monsoon zone it fluctuates in intensity between active spells and weak spells active spells in which you have rainfall continuously for several days and weak spells in which the rainfall is subdued and dry spells in which it is absolutely absent. So, there are fluctuations between active spells with continuous rainfall for several days interspersed with weak spells or what we call breaks of the monsoon which I will come to. Now let us as we did for the case of interannual variation let us first look at the all India average picture and by the way I should mention that diagrams like this for every year are available in the website of the Indian Institute of Tropical Meteorology Pune to which I have already given a reference to that website. So, now what you see here is the daily variation of rainfall and you know these there are the sticks when the sticks are high you have wet spells when the sticks are low it is relatively weak remember this is all India average. So, relatively weak means rainfall is occurring somewhere, but not everywhere and not sufficiently intense. So, these look very much like the Manhattan towers that people who make cricket commentary talk about. So, you have these many runs code in some of these and hardly any action in some of these other spells. So, these are these are the active spells and you can see even within the active large scale active spell here there are four short scale active spells and there are weak spells here and then again active spells resume this red line here is the climatology that is to say it is the mean all India rainfall. So, when the sticks are above the red line all India rainfall is higher than the mean when they are below all India rainfall is lower than the mean. Now, 2002 was a drought year and what you see is that for a many many days the rainfall was below the normal 2009 was another drought year and both with actually two recent severe droughts with ISMR deficit greater than 20 percent. So, these are really severe droughts 2002 and 2009, but you can see that even they are similar in terms of how much the deficit was in an all India average you still see that there are differences in how the monsoon evolved. See in this case or in the when the dry spells or weak spells occurred in the case of 2002 June was quite this is month of June, but in July you got many many weak spells whereas, in 2009 June itself was a huge deficit and then it recovered somewhat in July and then we got another major deficit slot in August. So, there are differences on the daily scale also between droughts now you look at normal monsoon years. So, called normal monsoon years which means at the end of the season the all India summer monsoon rainfall was very close to the average in this case ISMR was 2 percent above normal for these two years and these are 2010 and 2011 by the way the two years that followed the severe drought of 2009. And what you see is that even in normal years the rainfall fluctuates it is not you know it is unlike climatology which is so dull and flat if you plot the rainfall daily rainfall of any season it fluctuates a great deal whether it is a normal monsoon season or whether it is a drought. It fluctuates a great deal and typically you have wet spells of few days duration interspersed with weaker spells and this is what you see here as well. So, every year the all India rainfall fluctuates between active and weak spells active or wet spells with rainfall well below the average and relatively weak spells. Now, as I pointed out in the years of 2002 and 2009 the weak spells are longer and more intense and we can see that if you look at this you have a fairly long and a sufficiently intense weak spell here also one long weak spell another long weak spell and if you were to compare it with the normal years normal years also have weak spells, but they are much shorter. So, droughts do seem to have longer weak spells and while in both the years there is a this is something I pointed out while seeing that you know where the weak spell occurs seems to differ from one drought to another. Now, there is an interesting facet of these sub seasonal variations we call this sub seasonal variations because we are looking at variation within the monsoon season we have derived them from high frequency data that is daily data. So, these are variations within the monsoon season and what we find is one of the major characteristics of this is what I what is called break in the monsoon. Now, this is a word that has become extremely popular in fact to the extent that people have started defining breaks the way they like them which has created considerable confusion and we will come to that when we discuss intra seasonal variation in great detail. But let us now see what are the distinguishing attributes of breaks. See breaks are intense dry spells in the monsoon zone after the monsoon is established in the zone that is to say during the peak monsoon months of July and August if we have intense dry spells in the monsoon zone these events are called breaks. So, breaks are special cases of weak spells in the monsoon zone. Now, again we will get into the detail of this when we look at rainfall for active and weak spells and the data for several years and so on, but let me show you just the average or the composite picture. So, if you just take the average rainfall anomaly over all the breaks that have been identified in the 100 year time span then what do you find you find that there is a very severe rainfall anomaly over the monsoon zone here which is to be expected because breaks are dry spells of the monsoon zone. So, you have a very huge rainfall anomaly large deficit over the monsoon zone interestingly although in the definition of breaks and derivation of this data what we used was only the rainfall over monsoon zone it turns out that rainfall over west coast also shows very large deficit during the breaks. So, rainfall is has very large deficit over the monsoon zone which is the critical region for the monsoon it also has very large negative anomalies or deficit rainfall over the west coast. However, you see opposite sign of ISMR anomalies you notice here blues means rainfall is more yellow and red means negative anomaly of rainfall. So, you have large negative anomalies of rainfall here, but positive anomalies of rainfall over southeastern peninsula and over the this northeastern region and Himalayan foothills. So, this is a very interesting pattern that the rainfall anomalies are not coherent all across the region rather they are very coherent over the monsoon zone and the rainfall anomaly over west coast tend to go with monsoon zone that is to say if monsoon zone has positive negative anomalies as here it will have negative anomalies too if it has positive anomalies as for active spells west coast also tends to have positive anomalies. Then compared to in relation to what is happening over the monsoon zone in fact the southeastern part of the peninsula always has opposite sign anomalies. So, in breaks it gets more rainfall than normal and in active spells it gets less rainfall than normal and Himalayan foothills again have the opposite sign. So, that in active spells you have negative anomalies here of rainfall and in weak spells you have lot more rain over the Himalayan foothills and the northeastern region. So, these are the typical patterns of breaks and active spells. Now, this I just wanted to remind you is the mean rainfall and it is the rainfall over this monsoon zone that was used for defining breaks and active spells and you can see that the monsoon zone is around here and that is where the anomalies are maximum. Now, suppose we consider we have already seen that if we look at the daily picture and you look at the mean rainfall over monsoon zone vis-a-vis mean rainfall over the entire Indian region then the two curves tend to go together this is what we saw in the first lecture that the variation is very very similar particularly during July-August. Now, I ask the question how important is monsoon zone rainfall in determining all India monsoon rainfall that is ISMR. Now, to answer that what we plot here is ISMR which is the all India monsoon rainfall for each year versus how much it rained in the monsoon zone. This is all based on I am sorry the period can be read I think it is 54 to 2005. So, in fact, this is based on the grid data grid rainfall data which was compiled first by Rajiv and which are now available on the website of India Med Department. So, from that grid data we can actually calculate how much it rained in the monsoon zone and what you see here is the all India monsoon rainfall for the summer monsoon versus monsoon zone rainfall for the summer monsoon and what you see is that they are highly correlated. In other words if we wanted to predict all India summer monsoon rainfall if we could predict rainfall of the major rain belt which is the monsoon zone then you have a prediction for all India monsoon rainfall as well. So, since ISMR is highly correlated with the rainfall over the monsoon zone we expected to be also related to the number of break or active days. Now, you have to remember that we have defined breaks or active days depending on the rainfall over the monsoon zone active days being those in which rainfall is in excess of normal over monsoon zone and break days being those in which it is highly deficit over the monsoon zone. So, if that is the case if the rainfall is very much depressed over the monsoon zone for many days then we expect that all India monsoon rainfall for the season as a whole will be deficit and. So, we expect some relationship between the number of break days and ISMR we have to also ask the question is the all India monsoon rainfall related to how many active days were there in the monsoon zone and we look at both those. So, this is the ISMR percentage departure. So, this is the ISMR anomaly which we had seen earlier expressed as a percentage of the mean and this is the number of break days. So, first and foremost if we looked at the years in which actually there was no ISMR anomaly that is to say the all India monsoon rainfall was actually normal mean then we would get a large variation I am sorry I go back. If we look at this 0 line this 0 line represents years in which there were 0 days of breaks never did the monsoon rainfall zone rainfall over monsoon zone become very large deficit not on a single day that is what 0 counts for and even then there is a large variation in ISMR it can be as large as 21 percent this must have been 1961, but it can be also as small as minus 7 percent. So, there is a huge spread given the number of break days that occur there is a huge spread then we start with typically we have counted breaks only if they are more than 3 days in duration. So, this is what matters this is the relationship that is relevant for us and in this relationship then we see that there is definitely a relationship the chance of getting below normal rainfall increases as number of break days increase and in fact if the number of break days are beyond 17 or so you are not only guaranteed to have below normal ISMR, but you are guaranteed to have a drought. So, droughts intense droughts are all which is see here more than 15 percent are all associated with actually large number of break days. So, there is this relationship between the fluctuations within the season how many break days there are and how the monsoon performs for the season as a whole the all India summer monsoon or what we call ISMR. So, this is in fact negatively correlated however please notice that while for large number of break days we are getting some signal in the relationship for break days within about 10 days actually you cannot say very much about what will happen to ISMR that is to say the all India monsoon rainfall really where is a great deal if the break days are less than 10 it varies between a drought to an excess season anywhere between the two and so it is very difficult to say anything about it in other words the relationship is not very strong when the number of break days is less than 10 when it is more than 15 or 17 it becomes very very strong. Now, this is the correlation with number of active days let me just point out that here the correlation is highly significant it is 0.61 which means that the number of break days explain about 36 percent of the variance of ISMR. So, this is a very significant correlation and this is the first glimpse we have of a relation between sub seasonal variation or intra seasonal variation variation within the season and year to year variation or inter annual variation. Now, when we look at number of active days actually the story is not that great if we look at number of active days the correlation has dropped substantively. So, that now the number of active days explain only less than 10 percent of the variance. So, correlation has dropped substantively and you can see for each number say here you have 7 active days and the ISMR varies all the way from minus 15 percent which is a drought to plus 15 percent which is excess. So, there is very little relationship between number of active days and ISMR except when ISMR except when the number of active days is reasonably large, but even there when you have number of active days say exceeding 13 or so in 4 then you always get above average ISMR in 4 out of 5 cases. In fact, you are getting significant anomaly it is more than 5 percent excess of ISMR over the mean is more than 5 percent, but then you have a year like 2006 in which you had so many active days and yet the rainfall was very close to 0. So, you can say a lot more about what the ISMR corresponding to a situation with a given number of break days is provided of course, you have had a large number of break days relative to what you can say about ISMR from the information you have about number of active days. Now, so this says that even if your final aim was to understand only the inter annual variation of the all India summer monsoon. Even then it is important to understand breaks because it is very clear that you have a one to one correspondence between very long breaks and droughts and so it is very important to understand the morphology of breaks. You know what is the structure of the breaks what are the rainfall patterns elsewhere in the world and so on and so forth. Now, how does the transition to from active to break phase occur? We saw several of these transitions earlier in this all India case you saw several transitions this is an active spell and there is a transition to a weak spell. This is a weak spell and there is a transition to an active spell again active spell and transition to a weak spell. So, there are these transitions, but we ought to be able to understand how these occur and do we understand the mechanisms that lead to these transitions? It is only if we can understand the mechanisms we will be able to model them and we will be able to predict them. So, now to we have seen all this now thus it is important to understand the morphology of breaks the transition from active to break phases and revival of the monsoon from breaks. So, these are all facets of the monsoon which we will look at in great detail when we look at inter-seasonal variation. In this lecture I am only identifying the important facets. Now, rainfall is the most important attribute of the atmosphere in the tropics. So, I have discussed the mean rainfall patterns and variability of some of the facets based on rain gauge data over the Indian region. So, far what we have talked about is something that was actually known from before the satellite era itself from before we got satellite observations. These were data rainfall data that were analyzed and all these facets of fluctuations between active weak spells morphology of breaks and so on. People have studied and worked on with the rainfall data, but what happened is with the advent of meteorological satellites it became possible to literally see the cloud systems that rainfall over that result in rainfall over land or ocean. See after all we get rain from clouds and for the first time we had an eye in the sky which actually let us see every day where the cloud systems are over the entire tropical belt including the Indian region. So, in fact it turns out that the eye in the sky has taught us a great deal. In fact, you will see as we proceed that the kind of things we learnt from the satellite, the kind of information that we could derive from satellite data has actually contributed not only to our understanding of variability of the monsoon, but also it has helped in elucidation of physics of the system which is responsible for the monsoon physics of the mean monsoon that is the basic system responsible for the monsoon. So, this is another case in which new observations have given new insight into the physics of the system. Now, so we are going to talk about what satellite data have told us. Now, initially satellite data used to comprise images that satellite. So, basically there were cameras and satellites and we got pictures taken by satellites and in those pictures bright bands or bright regions represent clouds. Why is that? Because sunlight falls on the top of the surface and when there are clouds it gets reflected and it looks white. So, what the satellite camera notes is a white region where there are clouds and darker regions where there are no clouds. Now, satellite image of an active monsoon day is shown in this one and what you see this is a sector and I hope you people can recognize what it is. This is the Indian region you can see that and perhaps even a yes you can even see Sri Lanka here. So, this is the Indian region and what you see is a whole band of cloud going across India. In fact, it is going across the monsoon zone as we had identified earlier. It is coming all the way from the western parts and extending here right across eastwards. See this is now 70 degrees east and this is 90 degrees east. So, this is close to the longitude of Calcutta and 70 degrees east is close to the west coast and you can see that this is 70 degrees east here. That is 70 degrees east and this is 90 degrees east here. It is going right through the head way of Bengal. So, typical pictures that we saw from satellite were pictures like this and in fact the very first study of daily variation of these pictures itself revealed several new features which are particularly important in the seasonal transitions as well as intracesional variation of the Indian summer monsoon. Now, this first study was carried out by us in 19 and published in 1980 and I will talk about what satellite data has taught us in the next lecture. At this point this is a good point to stop this lecture. Let me just revise what we have learnt. What we have learnt is that there is considerable variation in most if we look at rainfall data alone. We see that from year to year there is considerable variation in many facets of the monsoon. We have seen that if you look at all India monsoon rainfall it varies from year to year but it is a reliable facet of the monsoon and the total range of variation that we have observed from 1876 is only 40 percent of the mean. So, it is a reliable facet but it does vary from year to year and although it is only 10 percent is the typical variation 10 percent is the standard deviation even then prediction of Indian monsoon rainfall is very important and particularly as we will see later prediction of extremes whether you will have a drought or not whether it will be an excess rainfall season or not is extremely important. So, variability of all India monsoon rainfall is very important then the facet of the monsoon we looked at was the onset of the monsoon and onset of the monsoon over Kerala is an event which is very dramatic and which is looked forward to during the summer season but which does not seem to be too much related to how the monsoon will perform in the rest of the season. So, if we were only interested in the all India summer monsoon rainfall then we would not worry too much about onset when the onset occurred over Kerala. Then we also looked at advance of the monsoon from Kerala to the northwestern region and then we see that in fact the performance of the monsoon for the season as a whole that is ISMR is indeed related to how fast it progressed across the country or how slow it was and in particular when it progressed faster than normal then the chance of above normal ISMR was higher. So, this is a signal we got then we looked at fluctuations during the peak monsoon months of July and August in rainfall over the monsoon zone which is the seat of the rain belt and we consider the fluctuations between active spells with continuous rain for several days and weak spell with subdued rain or intense dry spells called breaks. And we showed that in fact number of break days is well correlated with the performance in the monsoon of the monsoon in the season as a whole or ISMR and the number of break days are highly correlated the correlation was 0.6 or so and in fact if you had a season in which the number of break days exceeded about 14 or so then you were more or less guaranteed a drought. So, and also all the severe droughts were associated with long breaks or large number of break days because it need not be one break. So, that pointed to us that it is very important to study also the sub season and variation even if you are interested only in the all India summer monsoon rainfall you know the performance for the whole season as a whole because of this relationship of the active and weak spells rather the breaks with the ISMR. So, we have to study the morphology of the breaks what is the structure like can we predict the breaks in other words can we predict the transition from active to break can we predict the transition from break to active. Now, these are all open problems and we will see how far we have gotten in terms of understanding the mechanism that lead to the transition and prediction of this inter seasonal variation. So, so far I talked about what rainfall has taught us, but as I mentioned with the advent of satellite it opened a new dimension totally to us and this is the dimension that I will discuss and I will discuss that in the next class I will begin with talking about what the satellites have shown us in terms of monsoon variability in the next lecture. Thank you.