In 2008, the Kosi River in India abruptly changed course during a rare event called an avulsion. Over a matter of days, the Kosi’s path shifted about 100 kilometers, killing more than 400 people and displacing millions.
What determines where avulsions occur isn’t well known, as data on such river relocations are sparse. Researchers have built their understanding of avulsions mostly from lab-scale experiments, numerical modeling, and studies of the deposits of past avulsions. Now a new study mines almost 50 years of satellite imagery to gain insights into where rivers lurch onto new paths.
“You can think of avulsions as being the earthquakes of river deltas,” said Vamsi Ganti, a geomorphologist at the University of California, Santa Barbara, who led the new work. Avulsions are tied to river deltas as earthquakes are tied to faults, he explained, and just like ground shakers, avulsions are abrupt and can be catastrophic. “The importance of the problem is huge because river jumping events have caused the deadliest floods in recorded human history.” Researchers have even linked the decline of early cities in Mesopotamia to avulsions on the Tigris and Euphrates Rivers.
Changes on the Delta
To root out where avulsions occurred, Ganti’s team tracked changes to rivers in satellite images between 1973 and 2020. Sam Brooke, a postdoctoral researcher working with Ganti at the time, curated much of that data into movies that he watched for hours on end to ensure that the events captured were actual avulsions—sudden shifts—rather than gradual migrations. Once the team built their data set of 113 avulsions, they set about characterizing them.
“The work is beautiful,” said Douglas Jerolmack, a geophysicist at the University of Pennsylvania in Philadelphia. “It’s comprehensive in a way that is rare” because it’s challenging to integrate such measurements globally, he said. The study also validates some of the frameworks and hypotheses that have bubbled up over the past couple of decades.
Avulsions occur because rivers transport and drop sediment. Slowing rivers deposit increasing amounts of sediment, which can eventually choke the flow, forcing the river to find a new path. From the analysis, three regimes of avulsions emerged on the basis of the landscape and physical properties of rivers, the team reported in Science.
Of the 113 avulsions, 33 were on fans, places where a river emerges from a canyon or a valley. These avulsions happened where rivers lose their confinement (where sediment builds up). “That is what people have believed for a long time, but [the new research has] shown it,” Jerolmack said.
For the other 80 observations on rivers meeting the shore of an ocean or lake, the team linked avulsion locations to properties of the rivers’ flows. For each river, the team determined what’s called the backwater length on the basis of the rivers’ slopes and depths. Over its backwater length, a river’s flow responds to the flatness of the ocean or lake and starts slowing down, Jerolmack explained. A river’s backwater can be far from the shore—the Mississippi’s backwater zone stretches some 500 kilometers inland.
For 50 of these 80 avulsions, avulsions occurred within the backwater zone. These rivers tended to be large rivers with a shallow slope, like the Mississippi and the Brahmaputra. Avulsions in the backwater had been documented for such rivers in 2007, but it took time for researchers to understand the physics behind them, Ganti said.
For the other 30 instances on river deltas, avulsions occurred much farther upstream than the backwater length. Ganti’s team had observed this in Madagascar in 2020, but they thought such avulsions were rare. “The data suggests that there is this new regime of avulsions on deltas, which we didn’t think existed except for this weird case of Madagascar,” Ganti said.
Such avulsions occur in rivers with high sediment flows on tropical islands such as Papua New Guinea or in desert environments in countries such as Eritrea and Ethiopia. During floods on these rivers, erosion can travel far upstream, causing avulsions outside the backwater zone. “Communities that have never experienced any avulsion hazards could start to experience them in the future,” Ganti said.
River Reactions to Human Actions
More information is needed on what drives this newly found avulsion regime, Jerolmack said. Lab-scale studies could sweep from the backwater regime, with its low-slope and low-sediment rivers, to the high–sediment supply regime. Tracking how water and sand move along that spectrum could help untangle the physics behind what makes a river switch from one regime to another, he said.
On the basis of the new work, it seems possible that a river’s avulsion regime can also change because of human impacts, Ganti said. For instance, higher sediment loads caused by deforestation could potentially shift a river from the backwater regime to the high–sediment supply regime. Climate change may also shift avulsions upstream too, with more intense flooding events and greater flood variability. And rising sea levels will cause river mouths to retreat, Ganti said, pushing the entire backwater length up the river.
The information drawn from this analysis provides a framework that may allow scientists to predict where avulsions might happen, Ganti said. And that could help communities make engineering decisions about how to protect themselves from these catastrophic events.
—Carolyn Wilke (@CarolynMWilke), Science Writer