Send to a friend

The details you provide on this page will not be used to send unsolicited email, and will not be sold to a 3rd party. See privacy policy.

[NEW DELHI] Indian scientists have developed a new monsoon forecast model that they claim is more accurate and gives better advance warning about possible drought — knowledge that is crucial for the country’s agriculture and economy.

The model, developed by the Indian Meteorological Department (IMD) in Delhi, uses eight land, ocean and wind parameters to predict the extent of the monsoon. It replaces an earlier 16-parameter model, developed in 1988, which came under severe criticism when it failed to predict last year’s drought (see Climate model under fire as rains fail South Asia).

A critical reevaluation by IMD after its forecast went awry last year showed that 10 of the 16 parameters of the old model needed to be discarded and new ones added for a more realistic forecast.

The new model enables a forecast to be made 40 days earlier than previously, and allows scientists to update their predictions halfway through the June-September monsoon season, a time that is crucial for Indian summer crops.

"There is scope for correction, as by [mid-July], the nature of advance of the monsoon will be known, and any disturbing trends in El Niño [which can adversely influence the Indian monsoon performance] will become apparent," says IMD Director General Ranjan Ratnakar Kelkar.

An accurate drought forecast would allow farmers to switch to more resilient crop varieties, and gives authorities more time to prepare for a potentially poor harvest.

Using the new model, the IMD is currently predicting a 21 per cent chance of a drought in India this year. It has also estimated a 39 per cent probability of below-average rainfall in some areas of the country. This contrasts to the previous model, which would have predicted a "normal" monsoon for India this year, and given no indication of a chance of drought.

The new model has retained six of the parameters used in the previous model, which had a relatively stable link with the monsoon, and two additional parameters believed to influence the Indian monsoon: temperatures in northwest Europe in January and in the southern Indian Ocean in March. It was devised using 38 years of data up to 1995, and has been verified for the seven-year period from 1996 to 2002.

The new model is claimed to be 95 per cent accurate. Its true test lies ahead, but Kelkar is confident that India now has a "more realistic and more statistically stable model" for forecasting monsoons.