Top, the the Indian Kosi fan avulsion during the 2008 flood. The photo below was taken 10 days before avulsion.
An interdisciplinary team of physicists and geologists led by Penn has made a major step toward predicting where and how large floods occur on river deltas and alluvial fans.
In a laboratory, researchers created a miniature river delta that replicates flooding patterns seen in natural rivers, resulting in a mathematical model capable of aiding in the prediction of the next catastrophic flood. The results appear in the current issue of Geophysical Research Letters.
Slow deposition of sediment within rivers eventually fills channels, forcing water to spill into surrounding areas and find a new, steeper path. The process is called avulsion. The result, with the proper conditions, is catastrophic flooding and permanent relocation of the river channel.
The goal of the Penn research was to improve prediction of why and where such flooding will occur and to determine how this avulsion process builds deltas and fans over geologic time.
Research was motivated by the Aug. 18, 2008, flooding of the Kosi River fan in northern India, where an artificial embankment was breached and the resulting floodwaters displaced more than a million people. Looking at satellite pictures, scientists from Penn and University of Minnesota Duluth noticed that floodwaters principally filled abandoned channel paths.
Meredith Reitz, lead author of the study and a graduate student in the Department of Physics and Astronomy in Penn’s School of Arts and Sciences, conducted a set of four laboratory experiments to study the avulsion process in detail. Reitz injected a mixture of water and sediment into a bathtub-sized tank and documented the formation and avulsion of river channels as they built a meter-sized delta.
“Reducing the scale of the system allows us to speed up time,” Reiz says. “We can observe processes in the lab that we could never see in nature.”
The laboratory experiments showed flooding patterns that were remarkably similar to the Kosi fan and revealed that flooding and channel relocation followed a repetitive cycle.
One major finding was that the formation of a river channel on a delta followed a random path; however, once a network of channels was formed, avulsion consistently returned flow to these same channels, rather than creating new ones. An additional important finding was that the average frequency of flooding was determined by how long it took to fill a channel with sediment. Researchers constructed a mathematical model incorporating these two ideas, which was able to reproduce the statistical behavior of flooding.
“Avulsions on river deltas and fans are like earthquakes,” said Douglas Jerolmack, director of the Sediment Dynamics Laboratory in the Department of Earth and Environmental Science at Penn and a co-author of the study. “It is impossible to predict exactly where and when they will occur, but we might be able to predict approximately how often they will occur and which areas are most vulnerable.”
The authors derived a simple method for estimating the recurrence interval of catastrophic flooding on real deltas. When used in conjunction with satellite images and topographic maps, this work will allow for enhanced flood hazard prediction. The study was funded by the National Science Foundation and was conducted by Reitz and Jerolmack at Penn and John Swenson of the University of Minnesota Duluth.
Originally published on March 17, 2010