Poring over the rain

Scientists have been trying to forecast monsoon for over 100 yrs. Yet it baffles them. Will it be any different this year? Reshma Patil probes.

An India-11 is cooped up in Pune, facing their hardest test of the year. They have nothing to do with cricket, though their job is fraught with just as much uncertainty, and the risk of a national embarrassment in case of unexpected failure.
Their computers are whirring in a stone building that locals know as the Shimla office, though it sits at a chaotic, polluted junction in the centre of Pune. The name stuck after the shifting of the headquarters of the India Meteorological Office (IMD) from Shimla to Pune, decades ago.

"From October, we start monitoring global weather changes,'' says M Rajeevan, captain of the team with the stress of predicting rainfall from June-September and issuing a warning if fields will parch in an unanticipated drought like that of 2002. The south-west monsoon is a cryptic phenomenon. "During the last 10 years, the monsoon has been below normal with two droughts,'' says Rajeevan.

So he's peering at intricate global data of melting snow in the Himalayas or the Tibetan plateau and how the world's oceans (especially the Pacific and Indian) are warming, silently influencing the intensity of the coming rains. "I have been getting letters and calls; everyone wants to know whether the rains will be good,'' says IMD director-general R C Bhatia in Delhi.

Bhatia is waiting for Rajeevan's computers and analysts to wade through data before the forecast is announced this month. The computers need six to eight minutes to run a day's forecast analysis. The seasonal monsoon forecast needs computation about 122 days long, but the computer models get cracking from April, so about 180 days of computation must be completed—in 10 different ensembles or weather situations. "We have to interpret data, not just feed it in computers,'' says Rajeevan.

"It's a very tough job,'' says D R Sikka, head of the department of science and technology's climate research programme in Delhi. "There's an average 16 per cent chance of monsoon failure, and failed monsoon years don't come one after another."

In 2005, top meteorologists including Rajeevan, stunned the fraternity with a scientific paper that was also waved in the Lok Sabha. It said India's monsoon forecasting skill has not improved over the last seven decades.

But lead author Sulochana Gadgil, honorary professor at the Centre for Atmospheric and Oceanic Sciences in the Indian Institute of Science, Bangalore, says there is considerable hope. "We know a lot about monsoon extremes now. So five to 10 years hence we'll be able to tell farmers if they can rule out a drought or excess rain," she says.

The total seasonal rainfall has remained stable in India over 130 years, but the frequency and intensity of 'extreme rain events', have been rising since 50 years, reported Pune scientists at the Indian Institute of Tropical Meteorology (IITM) last year.

Another IITM group and an international team found that the prospect of a severe drought is higher if the central Pacific warms more than east Pacific. "The 132-year historical rainfall records reveal that severe droughts in India have always been accompanied by El Nino events. Yet, El Nino events have not always produced severe droughts," reported IITM's Krishna Kumar last year in Science.

El Nino is no obscure headline to Ashok Jain, Director, Contact India Commodities, a futures trading company. The weather may soon be on Jain's investment portfolio. "We expect a legal framework for futures trading in weather derivatives to be approved by the Parliament this year," he says. "Weather derivatives will allow many of our clients to manage weather-related risks like sudden increases in temperature or inadequate rainfall, in a better way."

V Shanmugam, chief economist, MCX, another commodity exchange, says, "We track the monsoon on an ongoing basis because it can create volatility in trading. However, we cannot influence the market with the data we gather."

Records show alternating periods of three-four decades with normal or weak monsoons. But after a good run from 1987-2001, 2002 and 2004 witnessed droughts. "There are no signs of alarm yet for this season,'' says Gadgil. "It's best to wait and watch. Weather signals become clearer by June.''

In Noida, supercomputers that are never switched off are running an experimental forecast generated by a 'dynamical' model, at the National Centre for Medium-Range Weather Forecasting. But it's still at a research stage. Sikka offers veteran wisdom: "Given climate uncertainties, one should neither feel too sorry when the forecast fails, nor too happy when it works."