Was the 2002/03 Maha's Bumper Paddy Harvest due to El Nino?

: During the last six decades, Sri Lankan paddy production has systematically increased due to improved technology and enhanced cultivation. Along with this trend, there is also a year-to-year variation in harvest primarily due to rainfall variations. An analysis of these variations shows that during the global El Nino climatic episodes, the odds are that Yala harvest decreases and Maha harvest increases. During El Nino episodes, there is usually an increase in the rainfall from October to December which is the planting phase of the Maha season leading usually to an increase in paddy production. The 2002/ 2003 Maha was during an El Nino season and heavy rains fell during this period and it is likely that 2002/2003 bumper Maha crop was influenced by the El Nino event. Since the rise in paddy production was likely due to a climatic anomaly, it would be unwise to change agricultural policy to discourage paddy cultivation.


Introduction
During the last six decades, national paddy production has systematically increased due to factors such as high-yielding seed varieties, increased fertilizer application, increases in land under cultivation and use of tractors ([16], [5], [13]). Along with this trend, there is also a yearto-year variation in harvest primarily due to rainfall variations [15]. An analysis of the yearto-year rainfall variations shows that during the global El Nino climatic episodes, the odds are that Yala harvest decreases and Maha harvest increases ( [7], [17]). There are exceptions to this association. This may indeed have been the cause of the rise in paddy production in last Maha as an El Nino persisted through it.
The term El Nino was coined by fisher folk a century ago to describe the unusually warm waters that would occasionally form along the coast of Ecuador and Peru. This phenomenon typically occurred late in the calendar year near Christmas, hence the name El Nino (in Spanish for "the boy child", referring to the Christ child). Today the term El Nino is used to refer to a broader global scale phenomenon associated with changed oceanic and atmospheric states in the tropics [6].
The Pacific is a large ocean and changes in its surface temperatures affect other tropical regions through an atmospheric "bridge". Just as much as there is a land-sea-breeze day and night because of contrasts in heating rates of the land and oceans, there are east-west patterns of circulation caused by the successive layout of oceans and continents as one follows the equator. It is this circulation pattern that is affected by the warm Pacific and which in turns affects Sri Lanka on the other side of the world.
La Nina is the opposite phase of the El Nino and is characterized by cooler than normal sea surface temperatures across much of the tropical eastern Pacific. A La Nina event often, but not always, follows an El Nino and vice versa. Once developed, both El Nino and La Nina events tend to last for roughly a year although occasionally they may persist for 18 months or more. The time between successive El Nino events is irregular but they typically tend to recur every 3 to 7 years. El Nino and La Nina are both a normal part of the earths' climate and there is recorded evidence of their having occurred for hundreds of years.
The impact of El Nino on Sri Lanka is to increase the rainfall in the last quarter of the year and to reduce the rainfall thereafter ( [8], [12], [11], [7]). The El Nino is the major large scale influence, but its influence may be counteracted in some seasons by impacts such as variations in the Indian Ocean, Asian land mass changes and volcanoes. Thus in considering predictions based on El Nino, one needs to make allowance from one El Nino event to the next.

The bumper paddy harvest during the 2002/2003
Maha season led to some celebration and left farmers with poor prices and inadequate storage. Sri Lankan policy makers wondered whether this bumper harvest shall be sustained. Indeed in 1982/83, after three decades of sustained increases in paddy production, there was a bumper harvest in both Yala and Maha seasons and subsidies were reduced. Here, it is argued that it would be unwise to change agricultural policy to discourage paddy cultivation on the basis of the past bumper yield as the bumper harvest was mostly likely due to the climatic phenomenon known as El Nino.

ENSO Identification
The El Nino and La Nina years were identified following Trenberth [14]:

Rainfall Index for Sri Lanka
Sri Lanka has been divided into three principal agro-climatic zones [10]. The Wet, Intermediate and Dry zones have been demarcated based on hydrology, meteorology, soils and vegetation [10]. The Wet zone receives an annual rainfall of more than 2500 mm; the Intermediate zone receives between 1750 and 2500 mm; and the Dry zone receives less than 1750 mm in a year. A districtwise analysis of paddy production (Department of Census and Statistics, Sri Lanka, [1-3]) shows that the contribution to the island-wide paddy production from the Dry, Intermediate and Wet zone districts are in the ratio of 3:1:1.
The rainfall records from five stations located in paddy growing areas were used to compute a rainfall index for paddy production in Sri Lanka. Three rainfall-measuring stations were selected in the Dry zone at Anurhadhapura, Batticaloa and Amparai. The fourth rainfall measuring station at Kurunegala is located in the Intermediate zone and the fifth station was chosen from the Wet Zone at Ratnapura. The average of the rainfall in these five stations is used in this paper as a representative index. The annual climatology for  from 1952 to 1997 ([1-2]) was estimated from a five-year running mean (Figure 2 and 3). The seasonal production anomaly is computed as the difference between the paddy production in a given year and the running mean computed with a 5-year window centered on that year. The fractional production anomaly is the production anomaly normalized by the running mean production for a given year. The yield too shows an increasing trend and the yield anomaly is computed in the same manner as the production anomaly.  The trend for both Maha and Yala paddy production ( figure 2 and 3) shows a steady increase from 1952 to 1983. A reason for the decline in yield from 1989 to 1992 was the withdrawal of the subsidy given for fertilizer purchases to farmers [9]. Cultivation was affected in the northern regions of the island after 1983 due to civil disturbances. A declining trend in Maha rainfall contributed to the declining production trends. The mean Malta rainfall from 1983-1997 was 1110 mm compared with 1233 mm from 1952-1982. However, the declining rainfall trends was not the only cause of declining production; a similar analysis for Yala shows that while rainfall did not drop, there was a decline in production from 1983 to 1997.

During El Nino periods, the Yala rainfall drops from the seasonal mean usually. Since Yala is a water-constrained season, only half the land is cultivated and the impact of rainfall anomalies is brought out more clearly and there is a slightly stronger relationship between El Nino and production. If the La Nina lasts through both, a
Maha and a Yala then the combined paddy production tends to drop in Sri Lanka as the Maha harvest is about twice that of Yala.
If the El Nino lasts longer than two seasons, then the drop in Yala cultivation may be offset by the increase in Maha harvest. Indeed, since the cultivation in Yala is only half as that in Maha, the combined harvest shows only a weak increase during El Nino episodes. However, this tendency for one season to be more bountiful than the other has important consequence for the livelihood of the farmers, marketing, storage, national policy and historical outcomes.

Conclusions
During El Nino episodes there is usually an increase in rainfall over Sri Lanka from October to December at the start of the Maha season, and an increase in paddy production. In certain years, the rainfall was so heavy that there were floods so that the production got disrupted. Factors such as war, sudden changes in fertilizer policy and a hike in the previous seasons harvest can influence the paddy production along with the climate. Not every El Nino results in an increase in rainfall at the planting season and it is only the most important of several global climatic factors that influences Sri Lanka's climate. After the subsidies were reduced subsequent to the bumper harvest in 1982/83 (which was again an El Nino year), there was a dramatic drop in paddy production in the following years. Indeed, the rate at which the paddy production increased decade after decade dropped in 1982. There could be several reasons for that such as the reluctance of the government to provide subsidies, a reduced emphasis on self-sufficiency, the exhaustion of the so-called green revolution, urbanization, labour shortages, low prices, irrigation shortfalls and the civil wars. In retrospect, the withdrawal of the subsidies in 1983 was the wrong incentive. Of course, the policy makers were oblivious to the El Nino and indeed it was in 1982 that the first research paper on the relationship between Sri Lankan rainfall and El Nino [8] was published. Shortly thereafter they reversed their decisions at tremendous cost. Two decades later, this mistake should not be repeated.