What controls when and where rivers abruptly jump course

Global satellite observations of rivers have revealed key factors involved in avulsions, rare events in which a river abruptly abandons its established channel for another, resulting in devastating flooding. The findings reveal how avulsion hazards might respond to climate and land use change – causing upstream communities to be exposed to flooding hazards they never experienced – and they provide a framework to help predict when and where future avulsions might occur. Unlike the gentle meandering migration of a river’s course across floodplains and deltas, river avulsions can be far more dramatic – large rivers can suddenly change course charting new channels across a landscape. While these sporadic events are central to the formation of fluvial landforms like river deltas, avulsions have been responsible for catastrophic historical floods. For example, a 2008 river avulsion at the Kosi River fan in India resulted in flooding that killed hundreds and displaced nearly 3 million people. However, the time between natural avulsions in a river system can range from decades to millennia, and direct observation of the events is rare. As a result, the controls on where avulsions occur and how these events will respond to climate change and human activity is poorly understood. To address these unknowns, Sam Brooke and colleagues used 50 years of satellite imagery and documented 113 avulsion events in rivers across the globe. Consistent with experimental and theoretical conclusions, the authors found that many avulsions are tied to changes in channel slope and sedimentation in the backwater zone of rivers – the most downstream reach of coastal rivers where the flow is slow and channel slope low. However, Brooke et al. discovered a number of avulsions that were located farther upstream of where they would be normally expected. These events occurred most frequently in steep, sediment-rich rivers in tropical and desert environments. According to the authors, these avulsions are likely due to flood-driven erosion, which is expected to increase due to land use and climate change. As a result, this mechanism could make avulsions more common in upstream locations that have previously never experienced them. “Effective placing of river diversions requires a comprehensive understanding of natural avulsion timing and location,” write Paola Passalacqua and Andrew Moodie in a related Perspective. “The ability to predict avulsions is improving and moving toward providing robust predictions that governments and decisionmakers can use to implement interventions at the system scale.”