The research was motivated by the flooding in 2008 of the Kosi River in northern India, which caused the displacement of more than a million people. Study of satellite images suggested that a process known as avulsion – whereby river channels fill with sediment, forcing water to spill out into surrounding areas and eventually leading to the relocation of the channel – was critical to the flooding. 

The study’s lead author, Meredith Reitz of the University of Pennsylvania, documented the formation and avulsion of river channels in a bathtub-sized tank in the laboratory. ‘Reducing the scale of the system allows us to speed up time,’ she said. ‘We can observe processes in the lab that we could never see in nature.’

The patterns she observed were similar to those seen in satellite images of the Kosi River flooding. They showed that the formation of a river channel in a delta followed a random path, but once a network of channels had formed, avulsion returned flow to these channels, rather than creating new ones. The frequency of flooding was determined by how long it took for a channel to fill with sediment.

The team was able to use Reitz’s data to construct a simple model for estimating how often catastrophic flooding would occur on real deltas.

May 2010