Normal Monsoon Forecast: Indian Ocean Dipole (IOD), the Wild Card

The Indian South West Monsoon or Indian Summer Monsoon Rainfall (ISMR) is planet Earth’s most productive wet season. The monsoon generates over 80% of the annual rainfall across the country and is vital for the economy, being the main source of water for agriculture, which contributes about 17% to India’s gross domestic product. Other than the 60% of the country’s workforce that depends on agriculture, the rains are also important for traders dealing in food and cash crops as any shortfall can inject price volatility in the markets.  

“When the monsoon fails, GDP falls by almost 2-3 percent. It has a huge cost on the agricultural sector. Even in 2009, agricultural production fell by about 7 percent from 234 million tonne in 2008 to 218 million”. 

This was Dev Raj Sikka, an experienced meteorologist being quoted by Forbes. It is said when India sneezes; the world’s agricultural markets catch a cold. Meaning if India imports food commodities, it sends global food prices spiraling upwards and if we do not require any imports and instead exports food, it has a depressing effect on global food prices.

 The correlation should be a no brainer - a good monsoon results in record harvests and poor monsoons results in agricultural production shortfalls and food commodity inflation.  This is because 55 per cent of India’s agriculture is still rain-fed. Production of crops like paddy, coarse cereals, pulses, oil seeds and cotton is mostly dependent on the four-month monsoon season starting June.

The rains also provide critical soil moisture that helps sustain production during the ensuing rabi sowing season. Last year though we had a pretty good South West Monsoon, the North East monsoon rainfall fell short by a whopping 40%. This means a significant portion of the farmland is currently devoid of any tangible moisture content.

Added to this, water levels of all major rivers are running low.  As on April 4, the water level in the 80-odd major reservoirs across the country was 48.19 billion cubic centimetres, 83 per cent of last year’s storage during the same period and 31 per cent of the full-reservoir capacity. 

It is obvious that there is not much of a buffer to a tide over a bad monsoon season. But that does not in any way suggest that the country should be facing an exceptional crisis this year. All it suggests is that our dependence on the South-West Monsoon is total, year after year, and if it fails in any year, its impact has far reaching consequence. 

A worried Indian agriculture ministry however is leaving nothing to chance. It has already issued advisories to all states and Union territories to prepare a contingency plan in the event of low rains. It is especially focusing on drought-prone districts of the country. Explains an official of the Ministry of Agriculture to Business Standard:  

“After two consecutive years of good monsoon, the law of averages says 2012 could be a year of uneven rains. Hence, we are leaving nothing to chance.”

Yet the Indian monsoon phenomenon is only partially understood and notoriously difficult to predict for weather agencies the world over. Last year, the country's official weather agency got it completely wrong. Even its updated outlook of August 1 suggested the total rainfall for the season at only 92.5 per cent of the long period average (LPA), whereas the actual figure turned out to be 101 per cent. The IMD then simply ignored the crucial La Niña event developing over the equatorial Pacific.

Given its record of not getting it right last time, predicting deficient rainfall when it turned out normal  – or, more seriously, predicting normal monsoons in the drought years of 2009, 2004 and 2002, or even when they get the broad trend right, they find it's outside the inherent error margin of plus or minus 4 per cent  – one couldn't fault the increasing numbers of those skeptical of the reliability of their forecasts.

In fact the success rate of the Indian Meteorological Department (IMD)’s statistical model is just 22% while those for US based NCEP-CFC dynamic model, marginally better at 24%. We just do not know what the success rate of models of European and Japanese models are as they do not reveal these figures. But we can assume that they are of similar range or worse - if otherwise, India could just go by their forecasts to hit the bulls-eye, but that’s not happening either.

A model is a set of equations in which you enter observations or data of several variables, and get a forecast. These variables are conditions in the atmosphere, oceans and clouds, among other things. The IMD’s model took into account 16 such factors while those of Indian Institute of Tropical Meteorology (IITM), Pune incorporate over 100 of these factors. And if they still go wrong, then we can’t actually blame them. A Forbes article gives us a deeper insight in their interview with  Rupa Kumar Kolli, Chief, World Climate Application and Services division, WMO and M Rajeevan, one of India’s best known meteorologists: 

RKK: “In tropical conditions there is a limit to predictability because things develop quite rapidly, which do not have prior signals...

MR: The chaotic component is more in the Asian monsoon region"

To adapt, scientists need to factor fast these changes into their models. This isn’t easy as it sounds. Goswami, the head of IITM and the Indian Monsoon Mission explains to Forbes: 

“If we try and do that, the number of calculations goes up many times. There are many factors like that. So, what I am saying is that these are all physical processes which can affect our monsoons. Now, in order to predict, actually we have to model them. This is what you call a physical climate model. This model, in principle, is a model of the atmosphere which is nothing but the equations that will solve the motion of the air. Atmosphere is air, right! So ultimately, you have to calculate where there is more heat and less heat in the atmosphere”

Small variations in the monsoon onset, in the spatio-temporal variability during the season and in the seasonal mean rainfall have a potential for significant economic and social impacts. Dr. L. S. Rathore, Director, IMD’s Agro-Meteorology Division admitted during a media interview that forecasting the monsoon is quite complex, especially at a 1-to-2-month lead time or sub-national scales. The year to year variation in the Indian Summer Monsoon Rainfall (ISMR) is primarily attributed to its association with the slowly varying boundary forcing such as sea surface temperature, snow cover, soil moisture etc. This is the predictable part of the inter-annual variability. The unpredictable part of the variability is due to the natural variability (internal dynamics) of the monsoon system such as oceanic phenomena like El Niño Southern Oscillation (ENSO); Indian Ocean Dipole (IOD) etc 

In a scenario where meteorologists virtually throw up their hands in forecasting the monsoon, we find foreign funded NGOs like Oxfam, ActionAid, ChristianAid, World Vision and other NGOs peddling their climate smart agriculture (CSA) agenda within this country. If the agricultural practices they peddle were to truly reflect climate smartness then these NGOs really need to shame meteorologists by coming out with more timely and accurate monsoon forecasts as this is the single most significant variable that affects agricultural productivity in India. Yet, even this basic expectation of them of forecasting the seasonal mean rainfall is not met by them.

 

As they further work in the grassroots in various parts of the country they should also be in a position to address the problem of the high seasonal spatio-temporal variability of the monsoon. For example rainfall variability is as high as 40 per cent in northwestern India and as low as 7 per cent in northeastern India.  

Besides rainfall has been within the normal window of plus or minus 10 per cent (which is quite substantial) roughly 70 per cent of the time. This is so because, given the large geographical area of the country, there are enormous spatial variations, and drought over any region tends to get offset by an excess in another region to bring the overall rainfall within the large window. Given this, regional and sub-regional monsoon predictions need to be carried out if agriculture practices needed to be climate smartened. Do our climate smarty NGOs even think of carrying out such forecasts? You bet not.

In fact these NGOs really need to go much further in forecasting  to justify their climate smartness claims. A more significant monsoon variation within a season is colloquially called as active-break cycles. These intra-seasonal variations occur with a typical period between active phases of between 20 and 50 days. During the active phase, copious rainfall occurs, while during the break phase, little or no rainfall occurs. Prediction of the intra-seasonal rainfall variations is of prime importance as these variations can have dramatic impacts, affecting the timing of crop planting and crop selection, and the management of water and nutrient resources in the affected regions. So where again are these predictions of the so called climate smartys of the NGO sectors? We find none again. Why? It is obvious that they deliberately claim to an expertise wherein they have no clue about! Providing a forecast that goes wrong repeatedly and consistently will puncture their bubble image of climate smartness! 

ANALYSIS OF FORECASTS

 

The table above lists the early monsoon predictions by several agencies, both within and outside the country as published so far. With meteorologists proving unreliable and with so called NGO “climate smartys” deliberately shunning forecasts, this is the reason why we have also included predictions of the fast growing community of independent weather blogs within the country. 

Vagaries of the Weather is administered by my friend Rajesh Kapadia.  His experimental weekly weather forecasts have demonstrated a reasonable degree of accuracy and closely followed by many both within and outside the country. My own prediction is included though I admit I neither have the vast experience nor knowledge that Rajesh has to be taken seriously. However, there is a striking convergence in our model projections.

Backgrounder to the Model Frameworks

 

To be noted is that the IMD would give its forecast only by last week of April.  The IMD model is based on a statistical "power regression" platform that uses 16 parameters in an empirical relationship with the total quantum of monsoon rainfall over the entire country. Historical data over a sample period (1951-1987) have been used to identify the 16 atmosphere-ocean-land variables that have significant influence on the monsoon behaviour. 

These atmospheric factors, the "predictors", are both regional (such as the Himalayan snow cover) and global (such as the El Niño, the anomalous warming of the ocean off the Peruvian coast in the eastern Pacific), the latter's influences being termed "teleconnections". Measurements of their stabilised values, which occur from December to May, are used in establishing a quantitative relationship with the total monsoon rainfall, the "predictand". Based on the data, weights or coefficients are assigned to several parameters that go into the model and that is used to predict the next season’s forecast. The model is stated to have an inherent error window of 4% on either side because of its statistical nature. Over the years the skill of these models in predicting the monsoon has gone down, especially after they replaced 4 out of the 16 original parameters with new ones, a step taken ironically taken to improve their model performance better! 

The Indian Institute of Tropical Meteorology (IITM) heads the (Rs 400 crore) National Monsoon Mission (NMM) attempting to evolve a more accurate monsoon forecasting model. The basic architecture of the IITM model is borrowed from the US agency - National Centre for Environmental Prediction - Centre for Forecasting Climate (NCEP-CFC), which perhaps explains their convergence in predicting an above average monsoon this year. The model by IITM is an experimental one based on a new dynamical model - a coupled ocean-atmosphere model that takes into account observations on land as well as the ocean which eventually designed to replace the current statistical methods used by Indian Meteorological Department (IMD) for its seasonal forecasts. All the foreign agencies forecasts found tabulated are based on dynamic modelling. 

The model incorporates more than 100 parameters. Why this is called dynamic is that the model factors in the current year by taking into account any change in the relationship between the 100 odd factors. So, say, historically the relationship between temperature between the Bay of Bengal and land is one of the determinants of monsoon and if there is for example a cyclonic storm or tsunami,  the new model will factor in the changes of such an event. This implies that different mathematical formula need to be applied to capture the interaction of different physical phenomena. The formulas will divide the world into tens of thousands of grids of 100 sq km each. Each year, fresh values are entered for each of these parameters for each grid. Then the Beast (the IBM supercomputer), will crunch through millions of calculations to solve the mathematical formula and project the amount of rainfall India will get in that year.  The process is  of course much more complex than stated but deliberately caricatured to facilitate easier visualization. 

As compared to those used by meteorological institutions, models used by the community of independent weather blogs are simple and crude. We use only a handful of parameters, in my case only 3. It is tempting to equate increasing sophistication with accuracy. But this is an illusion in practice. The UK Met Office for example uses the most sophisticated and expensive super-computer and yet Piers Corbyn, a climate sceptic meteorologist and physicist, running Weather Action, beats hollow the UK Met Office time after time, year after year, just using his pocket calculator!

Gilbert Walker was one of the all time great meteorologists who among other things discovered the Southern Oscillation and its influence over the Indian monsoon. Walker propounded the theory that an increase in the number of parameters does not enhance the accuracy of a forecast. This was more quantitatively proved by E.N. Lorenz of the Massachusetts Institute of Technology (MIT), the man who discovered the phenomenon of chaos in physical systems, in a classic paper way back in 1956. What the Lorenz study pointed out is that increasing parameters can only end-up increasing errors in recasting. This is probably why current monsoon model success rates range within the measly 20% band.  

The forecasts of the community of independent weather blogs as seen so far usually assume that though several parameters condition the monsoon, by identifying its 3 topmost drivers and the interactions among them; this could give a fairly good idea of the monsoons likely performance in any given season. If last year it was the Niño-Southern Oscillation (ENSO), then it is the Indian Ocean Dipole (IOD) that's the prime driver of the monsoon this season.

ANALYSIS OF FORECASTS

Despite the wide variations seen in these forecasts, what’s clear is that no agency has yet predicted a deficient monsoon less than 90% of Long Period Average (LPA). Records of past years, however, show that a monsoon has an 18 per cent chance of falling in this category [deficient monsoon].

Accordingly, all forecasts published so far fall within the broad category of a normal monsoon that immediately rule out a repeat of the 2009-2010 drought year. Atmospheric scientists define a normal monsoon as one where the rainfall the country receives is between 90 per cent and 110 per cent of the long-period average (LPA). Rainfall data for over a century show that there is a 70 per cent chance that a monsoon will fall in this category.

However, the normal in the scientists' parlance is again sub-divided into three classification:

- below normal (from 90 per cent to 96 per cent of LPA);

- normal (from 96 per cent to 104 per cent of LPA); and

- above normal (from 104 per cent to 110 per cent of LPA).  

The major differences in all these forecasts are tabulated within these sub-categorizations: 

 Note: 

[IMD uses 88 cm as the mean all-India rainfall or the LTA based on its analysis of the rainfall time series for the period 1901-70. However, there is no one figure for the LTA on which there is a consensus among monsoon researchers. The problem arises because of the inconsistency in the number of rain gauges used for collecting rainfall data, and the corresponding area averaging that is done, from year to year. The entire country has a rain gauge network of about 5,000 stations. However, because of the non-availability of rainfall data over a consistent set of rain gauges, different rainfall series have been constructed based on different rain gauge networks giving rise to different LTAs ranging from 84.6 cm to 90.3 cm.]

[The persistent monsoon mode has moderate lead/lag correlation with the Indian Ocean Dipole (IOD) SST as well with the ENSO SST in the Pacific. The strong relation of the persistent modes, which mainly determine the seasonal mean monsoon, when the SST leads, provides hope for long-term prediction of seasonal mean monsoon. The strong relation between the monsoon and the SST, when the monsoon leads, points toward strong influence of the monsoon on the variability of ENSO and IOD.]

ENSO

El Niño is associated with drought over the Indian subcontinent. The last such event took place in 2009-10, when the country received just 79% rainfall during the monsoon.

The El Niño  of 2009-2010 was immediately followed by a multi-year La Niña of 2010-12 which was responsible for the good rainfalls the last two seasons. This event has also now completely dissipated.

Key Pacific Ocean indicators are now at neutral levels (neither La Niña nor El Niño). The SST anomaly map (NOAA) for the week ending 26th March shows mostly near-normal sea-surface temperatures in the equatorial Pacific, with some residual cool anomalies across the central equatorial Pacific. Likewise, atmospheric indicators such as cloudiness, trade winds and the Southern Oscillation Index (SOI) are beginning to return back to normal values. Perhaps the most important atmospheric indicator is cloudiness which remains relatively La Niña like (high) in contrast to the SOI that have already returned to neutral value.

 

There is considerable uncertainty in all forecasts, which favours ENSO-neutral or developing El Niño conditions over a return to La Niña conditions during the remainder of 2012. Forecasters however remain split whether an El Niño would develop after the boreal summer. The El Niño is characterized by a large scale weakening of the trade winds and warming of the surface layers in the Equatorial eastern and central Pacific Ocean.  El Niño events occur irregularly at intervals of 2-7 years, although the average has been, until recently, about once every 3-4 years and lasting 12-18 months.

In our forecast of a normal-normal monsoon, we have assumed a high probability for the ENSO remaining neutral till end of the year. But even if an El Niño starts developing from June, we do not consider this significant because the atmospheric impacts often lag the demise of an ENSO episode viz. aspects of La Niña would still be reflected in the coming monsoon season. In fact the latter rationale appears what exactly the IITM, Pune models also seemed to factored in:

“Although a weak El Niño may develop in the later part of the season, the experimental dynamical model still shows monsoon is more likely to be on the positive side of normal”.

Moreover, while most models are forecasting that neutral conditions will prevail over the Pacific for the rest of the year, some predict onset of El Niño only towards the end of the monsoon season (Sept). Only one or two actually predict the onset of a new El Niño by June. 

  

IOD/EQUINOO

The IOD is marked by see-sawing of sea-surface temperatures (SSTs) from the west to the east of the Indian Ocean. When the west gets warmer (positive IOD) it helps the Indian monsoon and vice versa. When IOD turns negative, it adversely affects rainfall within the Indian sub-continent by sucking up the thermal convection of sea water away from the mainland. 

Over the equatorial Indian Ocean, enhancement of deep convection in the atmosphere over the western part is generally associated with suppression over the eastern part and vice versa. The oscillation between these two states, which is reflected in the pressure gradients and the wind along the equator, is called the Equatorial Indian Ocean Oscillation (EQUINOO). EQUINOO is the atmospheric component of the coupled Indian Ocean Dipole mode.

Unlike in the case of El Niño-La Niña (teleconnection), the impact from an IOD is more direct and immediate to a concurrent Indian monsoon. However, the ENSO can sometimes swamp the IOD effect. An example was 1997-98 wherein the Indian sub-continent faced a crippling drought inspite of a positive IOD due to a super El Niño. It happened again in 2009-10 where IOD was in a neutral state meaning normal sea temperatures in the western and eastern half of the Indian Ocean but a strong El-Niño episode caused instead drought. 

Overall the following thumb rule applies: when an El Niño event occurs in the absence of a positive IOD, the Indian monsoon tends to break down. Conversely, when a positive IOD occurs in the absence of an El Niño, monsoon rainfall is significantly higher than average. When an El Niño event and a positive IOD coincide, however, normal levels of monsoonal rainfall tend to occur. A positive IOD therefore disengages the relationship between an El Niño pattern in the Pacific and monsoonal rainfall over India.

The Regional Centre for Global Change (RIGC), Tokyo has been forecasting from two months ago about the prospect of a weak negative IOD evolving (warmer in the east part of the Indian Ocean and cooler in the western part). This is fundamentally what would decide the fate of the monsoon this year, according to RIGC assessment. Its prediction of a below-normal monsoon in India in its seasonal forecasts relies heavily on its modeling of the IOD. 

As seen from above chart of the Australian Bureau of Meteorology (BOM), the IOD ensemble mean indicates IOD neutral values may prevail during the entire ISMR season (June-September). Though neutral, it is important to note the ensemble mean projects the IOD with  weakly negative bias during June-July, neutral in August and weakly positive in September. What this suggests is a likelihood of June-July experiencing relatively lesser rainfall; August with normal rainfall and the tail end season, September with relatively above normal rainfall. 

The mean distribution of the rainfall within these 4 monsoon months can be found in the table above. We can assume the expected shortfall in June and the excess in September to cancel each other out. Since the August rainfall is expected to be a normal, this means it is really the shortfall in July that needs to be made up. So how much could be this expected shortfall?  Since the IOD at worse would be only weakly negative, we can expect the shortfall within the 0-4% range.  Factoring this outcome, this would place the monsoon overall within the normal-normal category (from 96 per cent to 104 per cent of LPA). Annually, a 1-4% rainfall shortage in the ISMR could be realistically made up, either in part or whole,  by the North-East Monsoon whose prospects looks very promising as the IOD ensemble mean suggests it would be weakly positive territory during that period.

We can also confidently rule out a repeat of 2009-10, a deficient monsoon-drought year. [Krishna Kumar et al. 2006 study concluded that droughts in India are shown to be associated more with the warmest SST anomalies in the central equatorial Pacific (ENSO) than those in the eastern Pacific (IOD)]. On an average, one of every 4 El Niño events is extremely strong. Since a strong El Niño took place in 2009-2010, our model outlook considered it highly unlikely that we see another such strong episode until after 2020.

Further, much depends on distribution pattern of the monsoon in terms of both space and time. If the latter is good, even if there is any overall deficiency of the monsoon, its impact will hardly be felt. If this is not the case, this would have more far-reaching implications for agriculture. In the latter scenario, to successfully adapt we need to adopt strategies such as late sowing, make greater use of shorter duration crops, early maturing varieties and conserve irrigation reserves till at least end of June. Monsoon withdrawal starts usually from around September 15 though a 100 year trend analysis show a slight shift of monsoon activity towards late onset and late withdrawal. If this trend holds, Kharif (summer) agricultural production may not be too affected and whatever affected could be made up by the rabi (winter) crop. 

Land - sea surface temperature difference

Halley (1686) suggested that the primary cause of the monsoon was the differential heating between ocean and land. According to this theory, the fundamental driver of all the monsoon systems is solar heating of the land during the spring season that helps to establish a land-sea temperature difference. This contrast, with the land being warmer than the surrounding ocean, triggers a low-level flow of moisture from nearby oceans, and this moisture is rained out during convection over monsoonal regions. In the Indian context, this is particularly relevant as the Himalaya/Tibetan Plateau forces moist air upwards, which then condenses out as rainfall.

Differential heating of land and sea is still considered as the basic mechanism for the monsoon by several scientists (e.g., Webster 1987). But according to S Gadgil: 

“There is an alternative hypothesis in which the monsoon is considered as a manifestation of the seasonal migration of the intertropical convergence zone (ITCZ; Charney 1969) or the equatorial trough (Riehl 1954, 1979) in response to the seasonal variation of the latitude of maximum insolation.”

However RIGC model appear to place a high confidence on the differential heating of land and sea theory. It forecasted that due to the 2-year La Niña effect, surface temperatures over the Indian sub-continent would have colder than normal climes in summer, throwing a spanner into differential heating of land and sea process, upsetting the monsoon intensity. RIGC argues that the La Niña’s effect – apart from delivering an above-normal 2011 monsoon and a harsh winter that followed – would result in less-than-optimal heating of the land surface and cloud-building necessary for widespread rainfall activity in the season ahead. 

The RIGC outlook is consistent with the theory that increased snow over the Himalayas can be linked with weaker summer monsoon rains over India. Last year, the Himalayas received record snowfalls. A research, published some years ago in Climate Dynamics, studied the mechanisms using the Met Office/Hadley Centre climate model. It shows that greater snowfall reflects more sunlight and produces a cooling over the Himalayas. This in turn creates a weakening of the monsoon winds that bring rain to India. The relationship is strongest in the absence of warm (El Nino) or cold (La Nina) conditions in the tropical Pacific, because these are normally the dominant control over Indian rains.

But this theory appears in the danger of being falsified.  For one thing cloudiness  remains high. Land surface temperatures in India during Dec-February have indeed run much below normal just as RIGC predicted due to the harsh winter. But our model reasoned that so was the sea surface temperatures around the sub-continent which is running a tad below normal too, somewhat offsetting each other in overall effect.

Further by March, temperatures in the Indian sub-continent have rebounded very smartly. In the early days of Pakistani summer the mercury has already reached 43 degrees centigrade in Mithi, Jacobabad and some other parts of Sindh. Meanwhile, the IMD says this summer is likely to be the most severe in the last 10 years, thanks to dry northerly winds blowing through South India. The dry winds have already ushered in an early summer with Telangana and Rayalaseema regions suffering very high temperatures. The temperatures are expected to increase gradually till the first week of June. IMD further forecasted the first week of April will see temperatures hovering between 37 and 41 degrees Celsius. Last year, it was few degrees lower and hovered between 35 and 39 degrees Celsius and in 2010, it was between 36 and 40 degree Celsius. Meanwhile, heat wave conditions now cover many parts of Rajasthan, Gujarat and Madhya Pradesh.

How then to explain the torrid summer? The only explanation is that these are only signals emanating from some abnormal heating taking place of the Tibetan plateau. The Himalayas play more than the role of just the orographic barriers for Monsoon. They help in its confinement onto the subcontinent. Without it, the SW Monsoon winds would blow right over the Indian subcontinent into China, Afghanistan and Russia without causing any rain. Himalaya/Tibetan Plateau forces moist air upward, which then condenses out as rainfall. When a plateau heats up, winds over it move horizontally unlike upwards in the plains. This creates a high pressure over the plateau, which then shoves away winds in different directions. The Tibetan Plateau is in fact acknowledged as a major driver of the monsoons of India and China and SE Asia. While the plateau heats up in the spring and summer, air pressure attracts moist air from the Indian Ocean in the south and creates the monsoon rains that sustain an estimated over a one billion people.

The US National Centre for Environmental Prediction (NCEP) recently warned that summer showers could become even more widespread from the middle of this month. The RIGC model also forecasted above-normal pre-monsoon (March-April-May) showers which can lead to incidental cooling of the landmasses. Our model discounted the RIGC outlook. Last year too pre-monsoon rains were fairly vigorous but did not succeed in lowering the monsoon intensity. 

Conclusion

 

“It does seem like El Nino conditions may not emerge this year during monsoon. However, in such years, local meteorological factors may play an important role. It just takes one month of bad monsoon, like in July 2009, to destroy the whole year.” 

This was M Rajeevan, a former forecaster with IMD in a LiveMint interview. 

Forecasting the monsoon this year could accordingly prove to be extremely tricky affair because of the climatic transitions now taking place in the Pacific Ocean (ENSO) and the Indian Ocean (IOD). And all the forecasts discussed in this post are to be treated as just early indicators. They could all change in the weeks ahead as more clarity on trends of these climatic factors is obtained. 

However, from our analysis, there appears no significantly increased risk of this season’s monsoon turning deficient or below average. This is contrary to what most global models are currently indicating. Least of all, we can be fairly confident that it would not fare as badly as in 2009-10 (drought). Although the La Niña has subsided, forecasters put a low probability on a warming that would lead to a new El Niño event, a possible monsoon killer, before the end of this year.

The wild card to the monsoons would be definitely the IOD. Should it swing positive, India can look forward to a bountiful monsoon - a third in the row. On the other hand, should it shift to its negative mode, India may be placed in more than a spot of bother. If the IOD retains its neutral status coinciding with a neutral ENSO, the chances are we are heading for a normal-normal monsoon.

Of all forecasters considered by this post, RIGC possess the best track record. RIGC was the first and among a handful of forecasters who correctly predicted the development of the (2010-12) La Niña.  In fact, they correctly predicted a multi-year La Niña well in advance when almost everyone was treating it as a single year episode. More significantly RIGC’s track record in predicting the Indian monsoon is a cut above all others. It hit the bulls-eye with its forecast of the monsoons for the last two seasons. RIGC is an agency that this blog places alot of respect and their forecasts have pervasively influenced many of our previous posts, related to La Niña and the Indian monsoons. 

RIGC is also part of Japan Agency for Marine-Earth Science and Technology (JAMSTEC) who developed the advanced ocean-atmosphere coupled general circulation model called SINTEX-F1 to successfully predict IOD event of 2006 from the November of 2005. And their outlook? 

“Sea surface temperature in the northern Indian Ocean will be colder-than-normal up to fall. A weak negative IOD will evolve in early summer and then peak in fall.” 

So if RIGC forecast a below average (90 per cent to 96 per cent of LPA) monsoon, they should logically be taken most seriously.

IITM on the other hand is the baby of the Central Government’s Monsoon Mission project. India hopes this model may eventually replace the current statistical methods used by IMD for its seasonal forecasts. The IITM model besides is found as one of the best models in retrospective forecasting. In the past 18 months, the institute has extensively tested the model using a method called ‘hindcast’ to check if it can successfully ‘predict’ the amount of rainfall in past years. For instance, IITM fed in data from past years and compared the model’s results with actual rainfall from the period 1982-2009 and found it to be 60 percent accurate - considered one of best for all models. The performance was mixed. For example it did not predict the drought of 2009-10 and a trial forecast last year predicted 14% excess rains! Their monsoon forecast this year however will be their first real-time test. Their forecast of an above average (from 104 per cent to 110 per cent of LPA) monsoon, if proved right would catapult their model into the headlines.

RIGC vs. NCEP-IITM is therefore quite a tossup. It should be tempting to play safe by aligning our forecasts to those of one of these big players in the global weather forecasting industry. But then we won’t be doing justice to the voice of independent weather forecasters in the country. In contrast to their prediction of either above average (from 104 per cent to 110 per cent of LPA) or below average (90 per cent to 96 per cent of LPA) our forecast remain normal-normal (96 per cent to 104 per cent of LPA).

So how confident we are of our forecast. In fact, we are quite confident:

·         There is a chance that we may be proved wrong in our assumption that a new El Niño would not emerge by the year end.  Even so, this hardly changes our outlook of the ENSO being not be a key driver of this season’s monsoon since at the most it is most likely to be a weak event.

·         The spectacular rebound of surface temperatures within the Indian sub-continent linked to the heating of the Tibetan Plateau seen last month (March) is at the root of our optimism of a normal -normal monsoon.

·         Further India's pre-monsoon showers are "normal" this year, indicating that the monsoon season, which is crucial to farm output, will arrive on time, two senior weather officials said Tuesday. Northeastern India, eastern India and southeast peninsular India are currently experiencing thundershowers, a sign that pre-monsoon season "is all normal," said a senior official with the state-run India Meteorological Department.

·         We are totally in agreement with RIGC outlook suggesting that the IOD holds the key to this season’s monsoon. But the IOD is an unpredictable phenomenon and can change modes when we least expect it. If it switches to a strongly negative mode, it can prove our prediction wrong.

·         We give our forecast (normal-normal) monsoon a 70% probability to succeed. We give a 0% probability to those forecasts suggesting a deficient monsoon; 15% probability respectively for for below average and above average monsoon forecasts.

This is our preliminary forecast which will be next updated in May this year with a second and final update in July.