Every year, billions of people wait for monsoon rains to plant rice and fill rivers. But predicting what the rainy season will be like is very difficult: too many factors affect the weather. Scientists have found a new way: they connected 6242 points on a map where air temperature was measured daily, creating a vast web.
In this web, some points change in tandem, while others change in opposite ways. If point A warms up, point B might cool down a month later. Such connections are called teleconnections. It turned out that some points are especially good at 'sensing' future rains in distant regions: the Amazon, Hainan Island, Africa (the Sahel), and India. Their signals can predict precipitation 4–10 months in advance.
How does it work? For instance, rains in the Amazon are influenced by Rossby waves, which are large atmospheric waves that carry energy. Rains in India are influenced by the East Asian monsoon, while Hainan is affected by El Niño, a warm ocean current. Previously, these connections were known separately, but now they are combined into a single network that both predicts and explains.
The practical benefits are enormous: farmers can decide in advance what to plant, and authorities can prepare for floods or droughts. The atmospheric web is no longer a mystery, and now future rain can be read from it.
For the Sahel, the optimal training window was only 11 years, while for Hainan it was 40 years. This reflects the different rates at which climate links change across regions.