Critical Failures of Indian Weather Prediction in a Climate Challenged World

Critical Failures of Indian Weather Prediction in a Climate Challenged World

In terms of a farming tragedy, the drought in many parts of the country during the 2009 monsoon season, was yet fresh in the lives of our peasants, and it also took a significant toll on the Indian economy. Here we must make a difference between large scale, inputs intensive industrial agriculture – practiced in large scale in our grain belts of Punjab, Haryana and to some extent in Andhra Pradesh & Tamilnadu also – which were affected to a lesser extent (but still adversely affected) due to their capacity to access powered-irrigation, from the roughly 60% of Indian peasant farmers who still practice rain-fed cultivation, not necessarily by choice but largely out of lack of resources.   NORMALLY, one would not expect another drought like condition so soon – in 2012, especially with no adverse ENSO (El Nino Southern Oscillation) condition in the beginning of the monsoon season, but these are not normal times in terms of the increasingly visible impacts of global climate change.  The beginning of the year 2012 saw a severe winter in Europe, by some accounts – the worst in the last 20 years.  Unusual snow-storms raged in southern Europe and also in north Africa.  Rome and Tripoli – of all places, experienced heavy snow falls. There was a large no of deaths in Europe from exposure to extreme cold, in late January and early February 2012.  This was followed by an unusually hot summer in southern and middle United States.  Its grain belt was scorched, leading to a significant fall in the production of corn, which combined with the US agro-fuel policy (of partly replacing mineral origin gasoline by biological origin ethanol and producing this from its abundant corn harvests) to create a global food crisis and food prices shooting up, leading to heavy hits being taken by the poor & food insecure in several countries.    And in the far northern summer, there are frightening decreases in the Arctic ice cover, and very worrying melting of the Greenland ice sheet, which are indications of things going way beyond normal.

The cold wave in Europe in the early parts of the year 2012 was caused largely by significant ‘incursions’ of the Jet Stream to the south over Europe, which allowed the colder air from northern latitudes to invade far south.  Northern hemispheric Jet stream is a large & strong band of upper atmospheric wind pattern between the temperate & arctic regions, and normally flow fairly straight from west to east – dividing the colder higher latitudes from the warmer southern ones, but when and where they protrude either south or north for long periods, these causes severe weather over these ‘invaded’ regions.  The jet streams (both northern & southern hemispheric ones), as is the ENSO phenomenon, are very much dependent on the global patterns of atmospheric & oceanic heat flows, both of which are undergoing fairly rapid (by geological time standards) and ‘erratic’ changes.  And these are causing drastic changes or ‘aberrations’ in the familiar weather patterns.  This is also evident from the apparently increasing frequency of the ENSO phenomenon, which were earlier occurring much less frequently – as shown by evidence from south American lake bed deposits (ENSO is an above normal warming of the equatorial waters of the western Pacific Ocean, from near the coast of south America, and the changes in precipitation & other resultant effects are recorded in the lake bed soil there). 

This global warming driven climate change is an unfortunate outcome of the energy intensive, consumption driven capitalist economic model that the world seems to have grown accustomed to, but we will not deal with that question here.  In this piece, let’s focus on the small farmers and other artisanal workers (including fish-workers and as such) of India, who have to depend on the blessings of ‘normal’ or familiar weather, and how the scientific community of meteorologists in India have failed to serve even their basic needs of reasonably correct information.   This is not to cut the usual jokes on our meteorological predictions (which are not entirely without justification) – as the challenges they face are significant, but to highlight how it is much more critical to generate and spread correct weather information to people dependent on such information for their livelihoods and also for the nation as a whole, not least in terms of disaster preparedness in the face of increasing number of devastating floods.  This is particularly important when the normal patterns of weather cycles people are accustomed to – are no longer holding true. It is also to bring to focus, that faced with this new challenge, new capacities need to be developed and deployed for the common good of the people, more so for the vulnerable, and here failures are and will continue to extract a heavier toll as the weather cycles become more & more ‘erratic’ or un-familiar from those observed over the last century.  We should also keep in mind – which is otherwise common knowledge – that for farming and also for disaster preparedness, the temporal & spatial distribution (time and geographic space wise spread) of rainfall is as important as the total rainfall, and this is one of the aberrations that is increasingly becoming a reality, with the number of rainfall days coming down along with an increase in rainfall intensity on a lesser no of days.  And it is here the failures of the India Meteorological Department – IMD – stands out more starkly.

Repeating its 2009 callousness, the IMD kept predicting a normal monsoon for the year 2012, though as a reflection of a lesson learnt from 2009 – they also added that they will closely watch the developing ENSO condition. This “normal monsoon” prediction continued till early July – again with an unfortunate similarity with 2009, in spite of atmospheric conditions over Pacific changing, and NASA satellites monitoring the developing El Nino and putting the information in public domains.  The other global signals – as highlighted in the first paragraph – were also clear.  There was hardly any good or positive signal from the Indian ocean dipole condition either. IMD also tried to assure the increasingly restless farming community and the other ‘stake holders’, by telling that the till-then deficient rainfall would very soon pick up and normal farming operations can resume.  This in spite of the rule of thumb that for Kharif wheat & paddy, July third-fourth week is normally the approximate cutoff period for enough rains for sowing, with 2-3 weeks of minimum lead time needed for field preparations. This is also dictated by the temperature regime and the time needed for these crops to mature.   Over the years, more and more farmers in mainland India have developed some (though not total – fortunately) dependence on the weather predictions by the IMD, in addition to their own indigenous methods.  The problem is, with climate change, these indigenous or traditional methods – which are based on long decades of observations of the connections between visible natural signs and the ensuing weather – are increasingly going off target.   The lamentable failures of IMD’s predictions  adds on to these woes, as farmers develop dependence on its predictions (see Economic Times Bureau article dated July 25, 2012), and pay a heavier price – while the IMD gets into yet another round of finding good excuses.  And it is not only in 2009 or in 2012 that we have seen this farmer-hitting costly prediction failures, as the figure below illustrates, the IMD’s reasonably correct predictions have come in less than 50% of the years, almost like the throw of a dice – painting its “scientific claims” in a very poor light. This is not a constraint of the science of meteorology, or the present state of the art, as evidenced by the ability of other agencies to ‘see’ well in advance – as evidenced by the USDA’s GAIN report, by the FRA-Japan’s reports etc.  This is indicative of the lack of expertise and effort on IMD’s part (starker in 2009 example), and unfortunate as we need a far better weather prediction success rate while facing the challenges of a world with rapid climate change.  The generous annual budgetary support to the dept has not improved the situation, will the new Rs.400 crore “Monsoon Mission” bring a better success rate ?

Since we are close to the end of the monsoon season (in northern India at least, by Sept. 20), let’s look at the total rainfall in different regions of India by end of July and also by mid-Sept 2012.

As a result of this deficient rainfall, the major Kharif  crop planting has also suffered (table below), putting many millions of farmers into distress.  Which the table below do not show is the amount of labour, money and other inputs that these farmers have put in, depending the “Normal Monsoon” prediction of the IMD, and the resulting financial distress.

Table 1. India: Progressive Planting of Major Kharif Food Grains

(Area in Million Hectares) Crop	Planting as of July 27, 2012	Planting as of July 29, 2011	Planting as of July 29, 2010
Rice	19.11	20.99	19.98
Corn	5.72	5.93	6.39
Sorghum	1.72	2.23	2.41
Millets	4.31	7.03	7.90
Pulses	6.30	7.39	8.18

Source – Ministry of Agriculture, GoI.

==================================================

Let’s take a look back into what happened in 2009.  The picture we saw of a weak South West

monsoon, unfolding as time went by.  About 177 districts out of a total of 600 odd had been declared drought-affected. Some more were facing drought-like conditions.  The ‘monsoon season

(June-Sept) was a little more than half way on, and reports of farmers’ suicides were already appearing from Andhra Pradesh, Uttar Pradesh etc.  Any travel in the dry districts of UP,  Bihar, AP, Jharkhand … showed vast tracts of land lying fallow or with dried up paddy seedlings.  And despite Govt. assurances of adequate buffer stocks, retail food prices shot up so much, as to force the non-rich to cut down on this essential consumption.

What went wrong ?  First let’s have a look at the rain-fall deficit for various regions of the country, and also for the country as a whole – from monsoon-start period of June 1 to July 29, 2009 (this data is  taken from the IMD release of July 31 2009) –

Seasonal Rainfall Scenario (1 June to 29 July 2009) – courtesy IMD

Period Ending	Country as a whole	Northwest India	Central India	South Peninsula	North East India
03.06.09	-32	-40	-50	-14	-32
10.06.09	-39	-31	-56	-15	-44
17.06.09	-45	-26	-72	-21	-46
24.06.09	-54	-49	-73	-38	-55
01.07.09	-46	-45	-59	-31	-41
08.07.09	-36	-50	-40	-18	-34
15.07.09	-27	-43	-15	-12	-40
22.07.09	-19	-38	3	-6	-43

Period ending	Country as a whole	Northwest India	Central India	South Peninsula	North East India
03.06.09	-32	-40	-50	-14	-32
10.06.09	-39	-31	-56	-15	-44
17.06.09	-45	-26	-72	-21	-46
24.06.09	-54	-49	-73	-38	-55
01.07.09	-46	-45	-59	-31	-41
08.07.09	-36	-50	-40	-18	-34
15.07.09	-27	-43	-15	-12	-40
22.07.09	-19	-38	3	-6	-43
29.07.09	-19	-33	1	-15	-39

So, this turned out to be a ‘bad’ monsoon, as most regions had large rainfall-deficits, except central India (which recovered late in July).  The grain belts of N-W & AP-TN were both badly hit.

What did the Indian Meteorological Dept. predict, in its early or “first stage” forecast ?

First Stage Forecast issued (by IMD) on 17^th April, 2009

“IMD’s long range forecast for the 2009 southwest monsoon season (June to September) is that the rainfall for the country as a whole is likely to be Near Normal. Quantitatively, monsoon season rainfall is likely to be 96% of the long period average with a model error of ± 5%. The Long period average rainfall over the country as a whole for the period 1941-1990 is 89 cm.”

That is way off the mark with the actual scenario, even during the first two months of the S-W monsoon season.  But that can possibly be explained by saying that this was the early prediction, Kharif sowing season was yet to begin, and the complexities of the monsoon system before many indicators became clear, added more errors.

The late or “second stage forecast” was issued on June 24, which is well into the kharif sowing season for several states.  It is far more crucial, as crores of farmers depend on this to decide what to sow and when.  The ‘climatic window’ of sowing kharif paddy (which needs large amounts of rain water in 58% of India’s agricultural fields which are un-irrigated)  --goes up to about July 15 or thereabout.  If rainfall is expected or ‘predicted’ to be significantly lower, farmers without access to good irrigation resources, use their back-up strategy of sowing crops that require much less water, like Maize, Millets, Sorghum, Pulses etc.  Thus it was critical for the IMD to make as accurate a prediction as possible, and they are supposed to be equipped for that with latest computers, sophisticated computerized models, access to many satellite imaging systems, thousands of meteorologists …. This writer kept wondering throughout July (and voiced this in meetings), what is the basis of IMDs optimism ?  

And what was their late prediction ?  See below  (taken from IMD release of 24^th June.)

“South-West Monsoon Season Rainfall

IMD’s long range forecast update for the 2009 south-west monsoon season (June to September) is that the rainfall is likely to be below normal. Quantitatively, monsoon season rainfall for the country as a whole is likely to be 93% of the long period average with a model error of ±4%. The Long period average (LPA) rainfall over the country as a whole for the period 1941-1990 is 89 cm.

  

ii)         Monthly (July & August) Rainfall

Rainfall over the country as a whole in the month of July 2009 is likely to be 93% of its LPA and that in the month of August is likely to be 101% of LPA both with a model error of ± 9 %.   The long period average and coefficient of variation of rainfall based on the 1941-1990 data for all India and 4 broad geographical regions are given here along ßwith the forecasts:”

Area	Long period Average (mm)	Coefficient of variation (%)	Forecast (% of LPA)
All India (June to September)	890	10	93
All India (July)	293	13	93
All India (August)	262	14	101
NW India	612	19	81
Central India	994	14	99
NE India	1429	8	92
South Peninsula	725	15	93

Please note that the IMD said that the most crucial July rainfall will be just below normal, whereas the important Aug. rainfall will be above that. Not really alarming or bad for paddy sowing. Only N-W India was ‘predicted’ to be having a significant shortfall.

As late as July 31, IMD was still assuring – wrongly it now seems clear – that the rainfall in August will pick-up fully and there’s no cause for worry.  The same ‘assurance’ was given by even the Minister for Agriculture! Even in their very late ‘prediction’ released on Aug.07/2009, the total rainfall was pegged down to 87% of the Long Period Average !!

So, again, WHAT WENT SO WRONG ?  Was there any clear meteorological indication(s) available of the possibility of a weak S-W monsoon ?  Was IMD aware of such a possibility ?  The answer to both seems to be “YES”. How do we say that ? 

Let us first look at IMDs own report, where it is clear that they were aware that El Nino Southern Oscillation (ENSO) conditions were developing in equatorial Pacific, starting in May 2009 itself.  El Nino conditions, or warming of sea surface in the equatorial Pacific, is known to weaken the south Asian summer monsoon, subject to a few other conditions, as they noted in their same June 24 release--  

“Since the middle of April, 2009, ENSO neutral conditions are prevailing with positive SST anomalies** observed over the equatorial Pacific from the beginning of May. The latest observations and forecasts from both dynamical and statistical models suggest high probability (about 60%) for El Nino conditions to appear during the monsoon season. The probability for ENSO neutral conditions is about 40% and that for La Nina is negligible. 

It is important to note that other factors such as the Indian Ocean Sea surface temperatures also influence the monsoon rainfall over India in addition to El Niño and La Niña events. Forecasts from few climate models suggest possibility of the development of a weak positive Indian Ocean Dipole event during the 2009 monsoon season, which may not have much impact on the Indian monsoon. However, IMD is carefully monitoring the possible evolution of El Nino conditions over Pacific and the Indian Ocean Dipole. “

(** “Positive SST anomalies” means higher than normal Sea Surface Temperatures).

The developing El Nino conditions, along with the absence of any countering sea surface temperature condition in the Indian Ocean (as happened in 1997, thus preventing a drought in India despite El Nino conditions in that year) , should have alerted any professional agency - and IMD is supposed to be a “Big One” – about strong possibility of  a weak S-W monsoon. 

The US Dept of Agriculture, in its Global Agricultural Information Network (GAIN) Report (no. IN9086), released on 29^th June, did exactly that. They also repeatedly warned that the “window” of kharif  paddy sowing, as well as those for corn, sorghum, millet etc will soon end, with even an warning of a possible drought.  A quote from that report  --

“The window of opportunity for planting of most kharif crops (rice, coarse grains, soybeans, peanut, cotton, and pulses) will be over by mid-July. If rains come in the next one week, planting operations will pick up. Otherwise the country will be heading for a drought, which could be more serious than the 2002 drought, which resulted in significant crop losses.

In some major rice growing states such as West Bengal, Orissa, Chhattisgarh, Bihar, and Uttar Pradesh, the crop is mostly rainfed and dependent on monsoon rains. Although rice is mostly irrigated in the major surplus states of Punjab, Haryana, Andhra Pradesh and Tamil Nadu, the crop is still dependent on monsoon rains for replenishing ground water reserves and reservoirs required for irrigation and generating electricity to run tube wells. What is hurting the crop more this year is high surface temperatures (4 to 5 degree Celsius above normal) which is causing high evapo-transpiration. The lack of rains will also result in low fertilizer application, which also will have a negative impact on yields. “

Yet, throughout the month of July, the powers that be kept on repeating phrases like – “no cause for concern”, “rain fall will pick up”, … and the media flashed these all around the country, misleading all to the extent of little or no preparation for the coming drought !!  Only towards the 2^nd week of August (damage already done), did the Govt. admit and start preparing !!

The NASA MODIS satellite images, showing El Nino conditions developing in equatorial Pacific were also available to IMD (as they were to this writer, a non-specialist science activist), from June2009 or earlier, and we attach these here. The first pair compares the more developed El Nino condition on 25^th June 1997, with developing El Nino conditions on 16^th June 2009.  Notice the already warming equatorial Pacific water (yellow & red colours) in the middle of the 2^nd image, extending from the coasts of Peru  & Equador  –

A second NASA satellite image taken in late July 2009 (attached below) more clearly shows the El Nino conditions over the same area, indicating strong possibilities that the rainfall during the remaining part of the S-W monsoon is likely to be weaker than normal.  That is exactly what happened, with August rainfall till date (15^th August, ironically our Independence day, without any sign of the Indian people getting independence from these ‘disasters by design’ !) being far from the 101 % of LPA that IMD predicted !!

And farmers in large swathes of the country, along with the poor & lower-middle classes, are paying the heavy price of this careless (to use a soft word) folly of our “professional” weather-men. At last the Indian Govt. and the IMD has woken up from their callous slumber, and are sending messages like – “Weathermen warns : Less rain” on Aug.10, or  “Centre advises states to adopt water saving technologies for Rabi crop” on Aug.13,  thus writing off the most important cropping season in India in one stroke !! And we are told that the livelihood of about 65 crore Indians depend on agriculture !! The other 50 crores are also dependent on that food production.

A timely warning from IMD and the Indian Govt. could have alerted crores of farmers, and prevented them from sowing paddy where irrigation is scarce.  Bigger and more resourceful farmers have managed to save part of their crops by spending large amounts of money in pumping out groundwater for irrigation, leaving the smaller farmers even more parched. Those farmers in the 6 million plus hectares, who have not sown paddy due to lack of water and left their fields barren, have at least saved the sowing costs (and the debt from moneylenders) and labour.  A timely warning would also have guided them to sow – when the climatic ‘window’ was ‘open’, crops requiring much less water, as they know well. Now they are saddled with lost paddy in their dry fields, lost investments, lost opportunity to get some return from other crops, and an empty future to look ahead to.  Is it any wonder that desperate farmers have started committing suicide (despite Govt. obfuscations)?  It is not only the kharif crops which have taken a big hit, rabi crops will also suffer, as there will be very little ground water available, made  worse by desperate & indiscriminate pumping out to save whatever part of planted crops.  To make matters worse, reservoir levels in most of our river-basins are running very low!!

We have not even taken up the issue of possible changes in south-Asian summer monsoon in future, due to global warming.  Sophisticated modeling studies (by Purdue University, by renowned climatologist James Hansen at GISS etc) are indicating both a backward shift in time, and a decrease of total rainfall during S-W monsoon, along with change in regional & temporal distribution. To cope with these changes – when they become prominent, we will need a dependable, competent and sensitive weather agency / meteorological dept.  Will we get that ?

And it is not only the always-abused food producers who are paying the price.  Consumer Price Index for rural & manual workers have gone up by over 11%, that for urban non-manual workers even more (whatever the Govt. figure – one visit to the ‘weekly bazaars’ bears this out) – all on the back of a weak monsoon and the anticipated food shortages. Economic ‘pundits’ are predicting at least a one percent hit to the GDP growth from this and the consequent drop in rural demand.  The non-farming poor, lower middle class & even the middle class are cutting down on food intakes, due to much higher prices, which will have obvious health impacts. The appallingly shameful 45% malnutrition rate amongst our below-five children, is sure to jump to an even more shameful figure, as possible alternatives to water-guzzling rice – pulses/soybeans (protein sources now gone beyond the reach of most of Indians), were not sown due to lack of timely rainfall information and the non-availability of “kits’.

The IMD can say that they were not “one hundred percent sure” about a weak monsoon, and they will be right. But were they “one hundred percent sure” about rainfalls to the tune of 96% or even 93% of the Long Period Average ?  Obviously not.  So, why did they not come out openly with the apprehension – based on clear meteorological signs – of a weak S-W monsoon ? This would have prepared farmers across the nation to face a water shortage. Which caused far more damage to the country & its non-rich people ?   Was the prediction-failure caused by a fear that the newly elected Govt. will loose the “feel-good” factor if a bad monsoon possibility is broadcast?

The Indian people are paying a very heavy price, and they must have clear answers.  If not for any advantage this time, at least in the long term interest of the nation and its productive people. And shouldn’t the people’s representatives – our MPs & MLAs, raise these pertinent questions ?

Soumya Dutta -  soumyadutta_delhi@rediffmail.com 

Bharat Jan Vigyan Jatha / Beyond Copenhagen collective