On Tuesday, the India Meteorological Department made its first forecast for this year’s monsoon: it said that the country as a whole was likely to receive 106 per cent of normal rainfall this season. Based on an interaction with
D SIVANANDA PAI, India’s chief monsoon forecaster, AMITABH SINHA explains the key issues around the IMD’s monsoon forecast.
What exactly does the IMD’s monsoon forecast of “106 per cent of the Long Period Average” mean?
India receives about 116 cm of rainfall every year. A large part of this, 89 cm, comes in the four-month monsoon season from June to September. These numbers are averages of rainfall received over a 50-year period between 1951 and 2001, called the Long Period Average or LPA, and are treated as ‘normal’. Tuesday’s fsorecast, therefore, says that the country as a whole is likely to get 106 per cent of the normal 89 cm of rain this monsoon, which works out to 94 cm of rain.
For the purpose of classification, the IMD considers rainfall between 96 per cent and 104 per cent of the LPA as ‘normal’, between 104 per cent and 110 per cent of LPA as ‘above normal’, and anything above 110 per cent of LPA as ‘excess’. Similarly, rainfall between 90 per cent and 96 per cent of LPA is considered ‘below normal’, and anything less than that as ‘deficient’.
By this categorisation, this year’s monsoon rainfall is likely to be ‘above normal’. Last year, and the year before, India had ‘deficient’ rainfall.
In a normal monsoon year, does the entire country get normal rainfall?
No, says D Sivananda Pai, head of IMD’s Climate Services Division. Even in a year of excess rainfall, there can be pockets that experience drought. In a bad monsoon year, some areas may get good rain. Rainfall is not evenly distributed across regions.
The 89 cm of rainfall in a normal monsoon season is what is called ‘area-weighted average’. It means that if the total amount of rainfall, over every region, through the four-month period, is collected, it would fill a tank of 89 cm height spread over the entire area of the country. It is, therefore, important to note the area under consideration while referring to the rainfall received.
The monsoon, Pai pointed out, is a pretty stable system, and that about 70 per cent of times, monsoon rainfall tends to be ‘normal’. And even in the worst case scenarios, in extreme drought years, rainfall does not go below 70 per cent of ‘normal’. It in fact remains much above that.
In 1918, often recalled as the worst drought of the last century in terms of its geographical spread, rainfall deficiency was only 26 per cent. In much recent memory, 2009 received 22 per cent below normal rainfall. Last year, the country had 14 per cent deficiency, and in the year before, 12 per cent.
Can the IMD forecast how much rain your district or town will get this monsoon season?
No. Certainly not two months in advance. The long range forecast, the one that was issued on Tuesday, can only be done over a large geographical area, and for a prolonged period of time. Over a smaller area, forecasts can be done only for shorter periods of time.
In its first seasonal forecast in April, the IMD, therefore, includes just one number: the probability of rainfall over the country as a whole over the entire four-month period. Only in late May or early June, much closer to the start of the monsoon season, does the IMD come out with its predictions for rainfall in every region and for every month.
What does the IMD base its forecasts on?
The monsoon is impacted by a variety of land and ocean phenomena and their interactions, Pai explained. The first forecasts, in the late 19th century, used to be based only on an assessment of the amount of snow cover over the Eurasian region. Lesser snow cover meant a better monsoon. Progressively, the forecasts became much more sophisticated. By the 1980s, the IMD was using 16 predictors to make its forecast. But soon enough, it was realised that 16 were a few too many. The models were refined further, and currently, the IMD uses five predictors for its April forecast. It relies on two more predictors for its second-stage forecast in June.
The predictors that IMD uses for the April forecast are: temperature difference between surface temperatures in the North Pacific and North Atlantic oceans; surface temperature of the Indian Ocean near the equator; mean sea level pressure in the Pacific Ocean near East Asia; land temperatures over northwest Europe; and the volume of warm water in the equatorial Pacific Ocean.
The IMD uses statistical models to analyse these data and arrive at its forecast. ‘Statistical models’ try to match prevailing conditions with historical records to see how the monsoon had behaved in years when similar conditions had prevailed.
The IMD has also been experimenting with a ‘dynamical’ model. This model makes continuous observation of some selected physical phenomena, and notes how the conditions for monsoon behave over a period of time. It then follows those changes to extrapolate for the future, and comes up with a forecast. Right now, the operational forecast comes out of the statistical model — the dynamical model, even though its results are announced, is being used for research.