It’s a weather event so decisive for India’s economy that a former president once called it the nation’s “real” finance minister. But climate change is making the annual monsoon more difficult to forecast, and raising the stakes of getting those predictions wrong.
That’s why researchers at the India Meteorological Department have spent more than a decade fine-tuning a new way to divine when, and how much, rain will fall each year. The National Monsoon Mission, which set out in 2012 to move the nation over to a system that relies less on historical patterns and more on real-time, on-the-ground data gathering, is starting to pay off, potentially saving property, crops, and even lives.

“It was stressful work,” said Madhavan Nair Rajeevan, who led the effort as secretary of the Ministry of Earth Sciences. Indian scientists consider more than 10 existing global climate models — including the state-run Monsoon Mission Climate Forecast System — downloading a vast amount of data to test them out – and narrow it down to four or five that perform best on South Asian weather. “We didn’t give any deadline for the project to finish,” Rajeevan said. “Wherever research was being done, we used to send scientists to get trained.”
Every June, millions of farmers across Asia’s third-largest economy wait anxiously for the monsoon rains to fall and nourish their fields, putting an end to heat waves that have become increasingly extreme as climate change accelerates. Ample rains may boost production of crops like rice, soybeans, corn and sugar cane, helping to lower food prices and aiding the government’s efforts to cool inflation. But abnormal monsoons can be devastating, leading to crop failures and drinking water shortages. Excessive rain can cause floods and landslides.

A monsoon is considered “normal” when rains are 96% to 104% of the long-term average. Last year fell just outside the range, at 106% (the IMD predicted 103% rainfall). The prediction this year is for precipitation to reach 96% and arrive in Kerala on June 4, slightly later than usual.
“Monsoon forecasting is always challenging,” said Mrutyunjay Mohapatra, the IMD’s director general of meteorology. But the weather bureau’s forecasts have become more accurate since switching to a so-called multi-model ensemble forecasting system that analyzes data using a combination of multiple dynamic climate models, he said, instead of relying on a single model previously based on historical patterns.

The agency has also expanded its collection of on-the-ground data, using manned and automatic weather stations, aircraft, ships, Doppler radars, weather balloons, ocean buoys and satellites to gather information on atmospheric temperature, pressure, humidity, wind speed and direction and sea surface temperatures. The data is fed into a supercomputer housed in the Indian Institute of Tropical Meteorology in Pune, which spits out the first long-range monsoon forecast in April for the Monsoon Mission Model and updates that prediction at the end of May. The outlook is then revised every month.
“Dynamic models are simplified representations of some real-world entity, in equations or computer code,” said Suryachandra Rao, the Monsoon Mission’s associate mission director at IITM. “Their main purpose is to mimic some essential features of the study system, in this case, the Earth system.” They are called “dynamic” because they describe how changes will occur as they read real-time observations that are constantly changing. All that input is used to predict how the monsoon will shift and change as it approaches India’s shores.

Predicting weather events with a high level of accuracy has become increasingly important in India, one of the most vulnerable nations to climate change. It allows the country to better prepare — from issuing early heat and rain warnings to coordinating power supplies to guiding farmers on how to protect their crops. Last year, India suffered its hottest March in more than a century, scorching the grain harvest and forcing the government to curb exports. Below-normal rain often prompts the government to start work on drought-relief measures and take steps to prevent a spike in food prices. Stronger showers give the central bank room to cut rates.
The rise in global temperatures has caused the Indian Ocean to warm faster than other oceans, decreasing the difference in temperature between the land and sea, says Rao, which means the monsoon could be delayed further in several years. “Predicting extremes at any scale is difficult,” he said. “Making accurate predictions in the changing climate will become more challenging.”