April 2, 2022 | David F. Coppedge

Stream Channel Networks Form Quickly

Large complex networks of stream channels do not take
millions of years to form: just thousands or less


One might think that stream channel formation is well understood by now. A new paper shows that is not the case. A study by researchers at the University of Texas at Austin, published in a leading geological journal, proposes a new model that is much quicker than conventional wisdom has thought. The abstract explains why this is an important question.

Source: Swartz et al., Tributary channel networks formed by depositional processes, Nature Geoscience 15, pages 216–221, 28 Feb 2022.

Understanding the detailed structure of landscape topography is important when assessing risks in coastal plain areas susceptible to the combined effects of fluvial, pluvial and coastal flooding. Key to this analysis is the identification and characterization of drainage basins that control surface water flow, but the factors controlling the formation and evolution of drainages in low-relief coastal plains is not well known…. This work presents evidence for the creation and evolution of erosional dendritic channel networks within depositional environments with broad implications for understanding floodplain channelization, partitioning and routing of sediment and water across low-relief landscapes, and timescales and mechanisms of landscape evolution.

With three colleagues, Swartz set out to study how “dendritic channel networks” formed in the Gulf of Mexico coastal plain that lies between the Rio Grande and Mississippi rivers. This area has historically been prone to devastating floods.

The term “dendritic” (tree-like) refers to the observation that channel networks on low-relief landscapes often resemble branching trees. They are fractal, meaning that the branches are self-similar at all scales, whether they form on tiny scales (like on beach sand when the water recedes) or on large scales, covering many miles. Here is a particularly stunning example of dendritic channels I photographed recently on a very flat playa (dry lake) in the Mojave Desert of California:

Dendritic drainage channels in a desert playa, California. All photos c. David Coppedge.

As you can see, the branching patterns become smaller and smaller toward the tips of the “trees” yet have the same basic configuration. How do these patterns come about? This playa fills with a thin sheet of water after heavy rains, but the gypsite surface (gypsum with clay) quickly dries out in summer months, leaving the “dendrites” behind. In some places, one can see faint impressions of older dendrites underneath the top ones, although the major ones appear to last for years. In one location, dendrites had formed in the opposite direction to the primary ones, seeming to indicate changes in the direction of flow.

Dendritic patterns with opposite orientations.

One clue for rapid change can be seen in tire tracks from dirt bikes and jeeps made by drivers seeking the fun of carving “donuts” (circles) into the flat lake bed and driving around, apparently unaware of the very shallow dendritic patterns they are defacing.

Jeep tracks superimposed on natural dendritic channels. Note that older tracks underlie newer ones and are fading away.

I have walked out onto the lake bed and verified that most of the patterns are too shallow to notice; they can only be seen from high above. Most of the photos were taken from a height of 300 to 400 feet. The tire tracks shown above are from ground level, as are the next two photos.

The main trunk channels, shown from ground level, are only about a foot deep. See hat for scale.

The channels become progressively shallower at the tips and are only a few inches long.

Self-similarity of branching dendrites proceeds to the tips. Widest channels are the deepest. Photo from 400 feet.

This example in the desert seems to have the same characteristics of giant low-relief channel networks that Swartz and his team found covering many miles in the Gulf coastal plain: self-similar (fractal) branching at all scales. Indeed, the aerial images they gathered using Lidar have the same appearance as the ones I photographed (see Figs 2 and 3 in the paper).

The self-similarity and scaling for both floodplain drainage networks and the larger continental systems is quantified using Hack’s law, which relates channel length (l) and basin area (a). This analysis reveals a mechanism for the initiation of drainage networks within depositional landscapes on a large spatial scale and highlights the extent of floodplain channel networks on otherwise low-relief coastal landscapes.

How Do the Networks Start?

Judging from the new model, I infer that a dendritic channel begins when flowing stream water pushes aside material in its sinuous path, setting the course for the subsequent flow. This act of “aggradation” (raising the grade of the surface) creates “alluvial ridges” which constrain the direction of flow. The water begins incision, deepening the channel. Increased flow leads to channel avulsion (tearing away; “abrupt channel relocations”) and sapping (bank collapse from undermining by groundwater) as the flow increases and the channel widens. At any spot that permits a break in the path, new channels can branch away, repeating the process but with less flow and energy. The smaller branches decrease in size and length until additional flow is insufficient to branch any farther; the flow at those points run out of steam, while the larger channels get deeper.

We observe that the dendritic drainage basins that govern the coastal landscape have boundaries that are initially set and controlled by sinuous alluvial ridges defining previous courses of modern rivers that were abandoned through the process of channel avulsion. These depositional ridges form topographic highs on an otherwise low-relief coastal plain and define the initial extent and occurrence of the coastal drainages. While the basin boundaries are formed by depositional processes, they exhibit geometric scaling characteristics similar to basins interpreted to have evolved through erosion alone.

The authors note that branching patterns are found both upstream and downstream. Tributary networks, such as in mountains, are “erosional domains” where water flows downhill from numerous sources and collects into large rivers. As the rivers reach flood plains with low relief, they become “depositional domains” where the waters branch out into dendritic drainage basins as seen in the photos above. And yet while depositional domains in lowland rivers are becoming well characterized, “much more poorly understood is the broader structure of the landscape and floodplain adjacent to alluvial rivers.” These are not “featureless terrains devoid of relief” as commonly portrayed.

While the observation of channels in coastal environments is not new, to date there have been few systematic efforts to understand the occurrence and nature of floodplain tributary drainage networks within the terminal coastal plain landscape and how their formational processes might differ from upland channel networks.

How Old Are They?

Swartz’s team takes issue with conventional wisdom that millions of years were required to form the extensive channel networks in the Gulf basin.

These results have numerous implications for landscape dynamics, floodplain drainage pathways, interpretation of the rock record and coastal hydrology. The alluvial ridge divides can form on timescales of 102–104 years [100 to 10,000 years] as a result of river avulsion, and thus the resulting channel networks are constrained to initiate and evolve on the same scales, rather than the millions of years commonly associated with drainage basins.

The role of the small, low-relief channels has been overlooked in previous research on the coastal plain, they say. And yet this aspect has big implications for interpreting the geological record, too:

In addition, these results have significant implications for interpretation of the rock record. The relief generated from the incision of these channels, on average 7–10 m, is equivalent to the overall average elevation of the coastal plain. This relief is similar to or greater than the flow depth of many of the larger fluvial systems that are constructing the coastal plain, such as the Rio Grande, Colorado and Brazos rivers. It therefore provides the space necessary for significant preservation of fluvial stratigraphy in areas with low subsidence rates, a mechanism that has long remained elusive.

Re-Imagining Landscape Evolution

Since “branching, dendritic channel networks are ubiquitous on Earth’s surface and are a primary control on the routing of fluids, sediments and solutes across landscapes,” understanding their origin and evolution (in this sense, development over time) are crucial for understanding landscapes not only on the Earth but also on other planetary bodies, like Titan and Mars. This work may lead to re-interpretations of the nature and time required for features like alluvial fans, canyon systems and river deltas. The authors summarize their findings in the conclusion and point to further research that is needed:

Our analysis illustrates three key results. (1) The drainage divides for channel networks can be constructed through depositional processes, rather than competing erosional processes. (2) The resulting networks are strongly incisional in an otherwise depositional landscape. (3) The channel networks initiating from depositional processes exhibit self-similarity and scaling in line with the larger continental networks…. The stream networks that arise from these processes are of critical importance for understanding coastal hydrology, particularly in the export of nutrients to the coastal ocean. The role of small, low-relief channels in the coastal plain has been overlooked in previous analyses and should be the focus of further study.


(Visited 343 times, 1 visits today)

Leave a Reply