Using a scale-model riverbed 24 feet long and one foot wide, a team of scientists found a new mechanism that could form waterfalls without the commonly thought of external forces such as tectonics, climate, landslides or glaciers.
“We used a model river with flowing water and sediment to examine how waterfalls form,” Joel Scheingross, lead author of a recent article in the scientific journal Nature, said. “We sought to see if waterfalls can form through internal feedbacks alone, the forces between water flow, sediment transport and bedrock incision.”
The experiment, part of Scheingross’ doctoral studies before he became an assistant professor in the Department of Geological Sciences and Engineering at the ΒιΆΉΣ³», was designed to simplify the natural environment so the team could isolate the internal processes, while maintaining dynamic scaling.
They used a synthetic bedrock of polyurethane foam, which simulates uniform bedrock and lacks the differences in rock characteristics known to cause waterfall formation. The foam resists water erosion, and it follows the same erosion scaling law as does natural rock under abrasion by particle impacts such as from sediment.
While the iconic waterfalls of Yosemite are thought to have been formed by external forces, the authors propose that smaller waterfalls further upstream may have self-formed due to internal river dynamics without the action of external forces. The authors cite the beautiful 620-foot high Bridalveil Fall in Yosemite Valley as an example waterfall formed by glaciation, but propose numerous smaller waterfalls upstream of Bridalveil Falls may have formed by sediment transport and erosion. The instabilities between a river’s bedrock erosion, flow and sediment transport increases the washboard characteristics of the riverbed, which can then deepen to gradually form a waterfall.
“Earth scientists have many theories for how waterfalls form under different circumstances in nature,” said Justin Lawrence, a program director in the National Science Foundation’s Division of Earth Sciences, which funded the research. “For example, waterfalls can be produced by the intersection of rock types, changes in sea level, the thrusting of tectonic faults, climate change, landslides and glaciers. But contrary to popular belief that there must be an external trigger, this research team found that waterfalls can form on flat riverbeds as a result of changes in flow, sediment movement and bedrock erosion within a river.”
In the experiments, Scheingross and his colleagues – Michael Lamb (senior scientist on the project and Scheingross’ doctoral advisor) and Brian Fuller at the Division of Geological and Planetary Sciences at the California Institute of Technology in Pasadena, California – scaled the hydraulics and sediment transport on the basis of mountain rivers where waterfalls are common.
In about four hours, with the model riverbed tilted at 19.5 percent (about 11 degrees), the model produced about the equivalent of 50-50,000 years of river incision and evolution. Within the first hour of the experiment, the sediment wore a channel that was about four inches wide, with formation of a slot canyon.
The team ran the experiment, conducted in Lamb’s lab at Caltech, in 20 minute increments so they could stop and take measurements of the bedrock erosion. While the experiment only ran for four hours, though, the whole process of designing, setting up and running the experiment took about six months of full time work. There was about four months of set up, one month of running the experiment and a month of clean up.
“In our paper, we show a proof-of-concept experiment that demonstrates that waterfalls can form in the absence of changes in climate or tectonics, due entirely to the internal dynamics of rivers (the combination of water flow hydraulics, sediment transport and bedrock erosion),” Scheingross said. “This is significant, because if these self-formed waterfalls are widespread, it complicates our ability to use waterfalls to invert for Earth history.”
Scheingross explained that the erosion at waterfalls often causes the waterfalls, once formed, to retreat upstream, so that waterfalls may exist far upriver from their point of creation. This has led some researchers to use the position of waterfalls in river networks, in combination with models for the rate at which waterfalls retreat upstream, to back-calculate the time of a climatic or tectonic change that was thought to form the waterfalls. In this way, waterfalls and other steep sections of river channels have been used to try to understand Earth history.
“Identifying self-formed waterfalls in nature is extremely difficult, because almost every natural landscape is subject to changes in rock type, tectonics, and climate that can produce waterfalls through external forcing,” he said. “This paper shows that waterfalls can self-form, and while we proposed a few examples of where waterfalls may have self-formed in nature, more work needs to be done before we can go out in nature and start distinguishing between self-formed and externally-forced waterfalls.”
Work on self-formation waterfalls is just beginning
Scheingross hopes this paper can promote a community effort to tackle the problem.
“This work is now at its infancy,” Scheingross said. “Our findings suggests that waterfalls that occur in series in mountainous landscapes are more likely to develop from the self-formation mechanism we describe, while single waterfalls that occur in isolation, such as Niagara Falls, are not consistent with the self-formation mechanism.
“I’m in the process of designing new experiments and starting new field work to better constrain the formation and morphologic signature of self-formed waterfalls,” he said.
Funding for the project came mainly from Caltech and the National Science Foundation, with additional support from the Alexander von Humboldt Foundation and NASA.
