If India’s monsoon forecast turns out to be accurate, the country will be facing its 12th worst drought since 1950 and two consecutive years of the calamity for only the fourth time and the first time in 30 years. However, a private weather service and the meteorological department’s own track record raise some questions on the forecast.
On Tuesday, the Earth System Science Organisation and the India Meteorological Department (IMD) downgraded the operational long-range forecast for the southwest monsoon (June-September) rainfall from its April figure of 93 per cent (with an error margin of five per cent) of the normal to 88 per cent (four per cent). The forecast thus shifted from a below-normal monsoon to a deficient one, which implies a drought.
This makes it the first forecast of a drought before the onset of the monsoon since the department began forecasts in 1988. However, even given the 10 per cent margin of error around pre-monsoon forecasts, the IMD got only three of the last eight monsoon forecasts right, underestimated two and overestimated three, an analysis by The Hindu shows.
“You show me one model which is able to do a better job,” D.S. Pai, Chief Forecaster at the IMD, told The Hindu . There are a large number of unpredictable factors in the development of the monsoon and the longer range the forecast is, the greater the likelihood of error, he said. The IMD has been putting on trial a new supercomputer-driven model which has forecast an even more deficient monsoon, at 86 per cent with a five per cent error margin.
In recent years, private forecaster Skymet Weather Services has emerged as a challenger to the IMD’s projections and has got its forecast right every year since it started in 2012. Skymet has forecast a normal monsoon for 2015 and is sticking to its prediction, G.P. Sharma, vice-president, meteorology, says.
Mapping the monsoon, with all its complexities and the various factors that influence it, is a tough task at the best of times. For example, as meteorologist Eric Holthaus says, should the El Nino phenomenon die out more quickly, or if the Indian Ocean dipole strengthens, both would suggest more rain than is forecast.
Against that background, he says, the India Meteorological Department has been relatively accurate. “The latest June-July-August forecast for the U.S. government’s National Multi-Model Ensemble is also showing below-normal rain for most of India this year,” he says. However, Mr. Holthaus does accept that what matters for Indians is whether or not it will rain enough in their local area, and for that local detail, which is harder to predict, the IMD’s track record is worse.
“We are learning from our mistakes and are making our forecasts faster,” D.S. Pai, Chief Forecaster at the IMD, says.
Both climatological and statistical factors argue against a deficient monsoon, G.P. Sharma, vice-president, meteorology, Skymet Weather Services, says. “If the eastern end of the Indian Ocean is cooler than the western end, and there is an El Nino effect, that is the worst for the monsoon, but conditions are not like that,” he says.
