Old mind and the river

Two years ago, Gayathri Vangala, a researcher at the Indian Institute of Technology (IIT) Gandhinagar, and her supervisor, Vimal Mishra, set out to probe a paradox. The Brahmaputra River often defies flood forecasts: the heaviest downpours do not always produce the worst floods. Some of the most damaging floods have occurred without exceptional rainfall, while intense storms have passed with limited impact. The river upends a long-standing assumption: that more rain should mean more inundation.

To investigate, they zoomed outward to view the Brahmaputra Basin as a vast and dynamic system where floods arise from multiple interacting forces and routinely redraw landscapes, disrupt economies and displace millions every other year. The basin is among the most flood-prone in the world. Scientific assessments estimate that floods have affected up to 85% of the basin's agricultural land since 2015 and have impacted more than 7.5 million people.

For decades, flood forecasts have relied on rainfall measurements, upstream river flows and historical flood maps, assuming that rainfall is the primary driver of floods. Yet the Brahmaputra has consistently outpaced those predictions. Floods emerge far from embankments, exceed predicted levels, or unfold without extreme rainfall, as shifting channels and upstream flows combine in unexpected ways.

"Uncertainties in the Brahmaputra's behaviour often overwhelm our existing early warning systems," says Mishra. "But the river is really neither erratic nor unpredictable. The uncertainties come from a complex logic shaped by many forces."