Follow Us:
Tuesday, October 19, 2021

Make your own forecasts, it’s safer

The Meteorology Department's monsoon forecast on May 25 was as eagerly awaited as the Union Budget. The utility of the forecast of a 12th ...

May 29, 2000

The Meteorology Department’s monsoon forecast on May 25 was as eagerly awaited as the Union Budget. The utility of the forecast of a 12th consecutive normal monsoon, however, is less than nothing, thanks to its objective and methodology.

The forecast is based on observations of 16 parameters and an associative relationship, not necessarily a cause-and-effect phenomenon. Its validation is purely statistical. This is like the punter’s recourse to racehorses’ form book before betting. The true causes of a horse’s performance in a race could be its physical condition on that day, among other factors, making guesswork the punter’s only choice. The weather is unpredictable, somewhat like the horse’s performance, but not entirely so. Scientists say that nature makes up its mind about the weather only three to four weeks in advance. Reliable forecasts are thus possible only for this short duration not the whole of a three-month monsoon. One could, therefore, argue that the associative model is the only method available for monsoon forecasts.

Do we need season-long forecasts and should we set such store by them as we appear to do? Let us list our basic concerns regarding the monsoon. Farmers have the most urgent and immediate questions: when would the monsoon start and how often and how much would it rain, say, in the next fortnight? Crop performance depends on moisture availability, not in the entire growing period but at various stages of growth. Likewise, agricultural operations for the immediate future need to be planned mindful of the pattern of rainfall. The answers also have to be for specific areas, not for the country as a whole.

Thus, what we need are not the government’s season-long, nation-wide guesstimates, but disaggregated, area-specific, short-term forecasts. An extreme example shows how the official forecast could go completely off-target. If an area receives its entire average annual rainfall within a month, most of it will cause floods and run off to the sea, leaving a drought for the remainder of the year. If, however, it receives only 75 per cent of the average rainfall, but is distributed evenly throughout the monsoon, both agriculture and urban life would fare well, with crops receiving adequate and timely water and reservoirs getting filled. The IMD model would, however, term the former a good monsoon and the latter a poor one. The reality is precisely the opposite.

Meteorologists abroad now routinely make highly localised and reliable short-term forecasts using the known cause-and-effect relationships, with probabilities attached. In sharp contrast are our broadcast heavy rainfall warnings. This writer has tracked them over the last several years and almost invariably, they are for areas which have just suffered this fate.

The weather bureau thus provides us `pastcasts’. The Mumbai deluge of mid-May was not forecast. After a similar occurrence few monsoons ago, a weather official said that it was not their job to provide day-to-day forecasts! Even now, the IMD director, while giving us a soothing account of events that have already happened in the South Andaman Sea, admitted that they could not offer specific forecasts for the drought-affected areas.

Given India’s size, the total rainfall for the country as a whole would fall below 90 to 95 per cent of the long-term average only in periods of absolute calamity. Therefore, the current method would almost always forecast a good monsoon, yet large parts of the country could receive below average precipitation in many of these years and face severe water shortages later in the year, as has happened now. Further, these forecasts could lull into inaction those whose job is to be prepared for such hardships ahead of time. Gujarat and Rajasthan administrations did not have to be given a wake-up call in April about the disaster they faced; they could have been prepared six months earlier and escaped much of the hardship.

Soon, we will all discover the naked truth under the Met Department’s new clothes, as Khushwant Singh did last year. He wrote that with officials unable to provide local forecasts, he relied on CNN and Internet for accurate weekly weather forecasts to obtain a bumper potato crop. Your PC with an Internet connection would thus be a far more potent source of information than the supercomputer of Mausam Bhavan.

📣 The Indian Express is now on Telegram. Click here to join our channel (@indianexpress) and stay updated with the latest headlines

For all the latest News Archive News, download Indian Express App.

  • The Indian Express website has been rated GREEN for its credibility and trustworthiness by Newsguard, a global service that rates news sources for their journalistic standards.