Why you should care

Because the mystery of rainfall matters in the larger quest to understand global climate.

When D. Sivananda Pai talks about the weather, he sounds like a sober-faced gambler desultorily discussing his winnings. He fixates on the year, beginning in 2003, which was OK. 2004 was bad. The next few years: tolerable. In 2007, he tried something new; luck was on his side, where it’s been ever since. The glee mounts. Things went well in 2008 and 2009, and until 2014. What about 2015? I ask, with near-bated breath. Ah. Things looked bad, for a while. Until they didn’t. Pai was proven right. Cha-ching.

Unfortunately, being right meant bad news for the rest of the region. 2015 was the hottest year on record, and one of only a handful of times the monsoon season fell short of providing farmers with adequate rainfall for their plants. Monsoon refers to the rainy summer months that provide South and Southeast Asia with 70 to 90 percent of the water that crops receive all year. And predicting it — when it begins, the amount of rainfall, its sufficiency — is critical for a country where roughly half the population still works in agriculture, according to the World Bank. It’s especially important in a nation where headlines about rural rainfall are too often stories about farmers committing suicide in the face of drought. (Some 18,000 killed themselves in 2004, the worst year in memory.)

Pai is the guy credited with finally getting the tricky business of monsoon prediction right.


Pai, who lives in Pune and runs the long-range forecasting division of the India Meteorological Department (IMD), is not the sort of weatherman who flails his hands on your television set every day. He is a government-employed meteorologist with a Ph.D. and an expertise in statistics — he is, in fact, the guy credited with finally getting the tricky business of monsoon prediction right. His work has earned him an award of excellence from IMD and a certificate of merit from the Ministry of Earth Sciences, and he’s served on research teams for the World Meteorological Organization.

Statistical modeling, Pai’s terrain, works like this: Relying on a small handful of predictors — five or six, depending on the time of year, including ocean surface temperatures and pressure, the volume of warm water in the Pacific Ocean, etc. — the IMD compares those numbers to historical figures to estimate whether rainfall will come above or below the number of centimeters that mark a healthy versus deficient season. Figuring this out on the subcontinent “isn’t a small scale, local phenomenon,” says Rupa Kumar Kolli, officer in charge of the climate and water department at the World Meteorological Organization in Geneva. Kolli, who has known Pai for decades, says his work is an important puzzle piece in the larger quest to understand the climate — and its shifts. “This is a planetary-scale circulation, involving a tremendous amount of energy,” he says.

Speaking in his government office before the onset of the monsoon, Pai was eager to geek out on the history of his profession. When I asked about his success in statistical modeling, he took me back to 1875, when predictions depended on measuring snowfall in the Himalayas. He spoke for a quarter of an hour before we finally reached his lifetime, which, as it happens, began in the South Indian state of Kerala (where the monsoon generally strikes first). The son of less-than-privileged parents, Pai grew up watching the glowering skyline turn color when the rains were imminent — and when he and his friends could no longer go swimming if the water levels got too high. He developed a nose for predictions early on.

“It’s not that I had done anything special,” he deflects. “It is based on combined wisdom that I am willing to make a prediction.” Pai completed a master’s degree in his home state and then moved on to the prestigious Indian Institutes of Technology, Delhi, for graduate work, before landing in Pune for a doctorate.

Today, he’s recognized the world over as one of the faces of climate prediction in India — but not everyone agrees with his methods, says Andrew Turner, lecturer in monsoon systems at the University of Reading in the United Kingdom. Turner says the international meteorology community is moving in the direction of dynamical models of monsoon prediction, rather than continuing with Pai’s style of statistical methods. Dynamical models don’t rely on empirical comparisons like statistical methods, instead pulling from fluid dynamics and atmospheric data. Turner adds that the U.S. and the U.K. rely on dynamical models, but some countries continue to use statistical methods because they’re cheap. (Even though Pai comes from a stats background, he told me India’s working to improve its dynamical models.)

This year, the monsoon started a bit late, Turner says — four or five days, to be precise. Pai says he’s been proven right; he planned for a delay in Kerala and has come within 5 percentage points of an accurate rainfall. You’d hardly know of a delay from wading through the streets of any major Indian city, where the April and May heat was replaced by standing water fit for a boating expedition. As global conversations about climate change advance and continue, those involving the monsoon are evolving too. The same elements that help to forecast the climate — the El Niño cycle, among other factors — are becoming more erratic, promising to make the game of monsoon prediction even tougher. But therein, of course, lies the thrill for the gambling weather geek.


OZYRising Stars

People who are accelerating our culture and advancing the conversation – for good or for ill. You may not have heard of them yet – but you'll soon need to know 'em.