Rain gods will be kind this year too

Met dept sees only 4% chance of deficient rainfall.

The India Meteorological Department (IMD) yesterday predicted a normal monsoon this year, which implies that the country could see its second successive year of good rainfall after the severe drought of 2002.

The southern states of Kerala, Karnataka and their adjoining areas, which have had deficient rainfall for the last two years, are likely to get normal rain this year, according to the long-rang monsoon forecast.

The total rainfall during the four-month southwest monsoon season (June-September) this year is forecast to be 100 per cent of the long-period average rainfall of 88 cm. There could be a variation of plus or minus 5 per cent due to prediction model error.

However, the IMD is quiet on when the monsoon might hit the Kerala coast as also on the regional distribution of rainfall. "While the date of onset will be announced around the end of May, the forecast regarding rainfall in different regions will be made around the last week of June, when the IMD comes out with its monsoon forecast update," IMD Director-General S K Srivastav said.

 

Monsoon rainfall (as % of normal)
Year	Rainfall
1998	105
1999	96
2000	92
2001	92
2002	81
2003	102
2004(forecast)	100
Source: IMD

Giving the detailed probability of the likely rainfall pattern, he said there was a 4 per cent chance of deficient rainfall (less than 90 per cent of the long-range average). The probability of rainfall being excess (over 110 per cent of the average) is also 4 per cent.

The probability of near normal rainfall (between 98 per cent and 102 per cent) is 58 per cent and that of above normal (between 103 per cent and 110 per cent) 18 per cent. There was a 16 per cent probability of rainfall remaining between 90 per cent and 97 per cent, in the "below normal" category, he said.

Srivastav said the El Nino (the warming up of the Pacific ocean that causes poor monsoon) was neutral this year, eliminating possibilities of any adverse effects.

This long-range forecast is based on the new eight-parameter statistical model that replaced the old 16-parameter model last year. The old model had given accurate predictions till 1993 but had started showing erratic results thereafter. Its prediction turned out wrong in 2002, necessitating a change in the model.

The new model (with a model error of plus or minus 5 per cent) enabled the IMD to issue its first long-range prediction as early as mid-April based on data till March-end.

The updated monsoon forecast to be issued in the end of June will be based on a 10-parameter model that is more reliable because of a narrower model error of only plus or minus 4 per cent. The IMD will also issue a forecast for the month of July, which is crucial from the agricultural point of view.

Predictions based on the new model last year had proved accurate, Srivastav maintained.