Fluvial geomorphology as a tool for civil engineers

Jim Melvin, P.E., R.L.S.
Senior Staff Engineer
City of Lenexa, Kansas

Although it may come as a surprise to some, fluvial geomorphology can be used as a tool in civil engineering. In conjunction with bio-engineering, fluvial geomorphology can be a powerful tool in the stream restoration process. Most cases of stream restoration are the result of detrimental changes in the stream morphology caused by development or modifications to the natural stream.

As an example, let us examine the Brentwood Reach of the Little Mill Creek Watershed in Lenexa, Kansas.

The particular reach of Little Mill Creek began to experience development in the surrounding area about 30 years ago. Prior to that time the reach was agricultural or pasture ground. Over the last 30 years, the entire basin has developed. Prior to development, the stream was relatively stable, with established vegetation on the banks and overbanks.

During the development of adjacent properties various problems occurred. The banks began to erode at the outside of the bends. The bed began to headcut, causing further bank and bed erosion. The bends started to migrate at a faster rate than normal. What caused all of this? Well, the answer is well known to most engineers. The development increased both the frequency and magnitude of bankfull condition at higher flows, and thus the time of concentration has decreased.

How does geomorphology play into the solution of the problem? The answer is not simple, but geomorphology can help address the problem. In order to understand the problem and apply the right solution, we need to understand the causes of the problem, and how to accurately define the problem. The obvious cause of the problem is the increased flow, but is that all of it? Didn't the stream experience the same flows prior to development? Certainly it did, but not as frequently. The channel forming flow is the bankfull flow—approximately the 1.5-year storm. The bankfull flow prior to development was 370 cfs, and the developed bankfull flow is 570 cfs. When the flow changes, the stream must and will react to that change in flow. Since the bankfull flow increased, the channel will try to get wider, and flatter. We need to know about how much wider, and how much flatter.

Let's start the analysis by doing some measuring.  There are several things we will need to know:

  • The average width of the stream bed. For the Brentwood Reach, the width is 10-12 feet.

  • The ratio of the stream length to the valley length is called the sinuosity.1 It varies between 1.0 and 4.0 for all streams, with a median value of 1.5.1 For the Brentwood Reach, sinuosity is 1.1.

  • The distance between successive curves in opposite directions longitudinally is called the wavelength.1,2 Wavelength usually varies from 7 to 10 times the stream width.1 For the Brentwood Reach, the wavelength is approximately 160 feet.

  • The distance between the outside of two successive curves transversely to the stream is called the amplitude.1,2 Amplitude does not appear to be related to width or wavelength, but seems to be more dependent on bank material. The amplitude of the Brentwood Reach is approximately 80 feet.

  • The ratio of the average radius of curvature of the bends in the channel to the channel width varies only a little from stream to stream and river to river. The median value is 1.5.1,2 For the Brentwood Reach, the radius of curvature is between 20 and 30 feet.

  • The spacing of the riffles and bars. Normal range is between 5 and 7 times the bed width.1 For the Brentwood Reach, the riffle spacing averages 68'.

  • The channel slope varies with the area of the basin. The overall slope of the Brentwood Reach is 0.86%, but slope varies up to 3.0%. In the areas where the slope is near 0.86% the channel is more stable. In the vicinity of the 3.0% slope, there is a headcut. The channel is still reacting in some sections to the steeper slope and appears to be flattening its slope throughout the reach to a value near the average slope.

Basically, the information we have would indicate that the channel is still reacting to the development, but is nearly back to a balanced condition in many areas. No channel is ever truly "stable" over geologic time, and this channel is not quite stable over time in the near term. We need to add some length to the channel in the steeper areas (if possible), cut the slope slightly in the steeper areas (again, if possible), and/or widen the channel by 1 to 3 feet where needed. The additional length is needed to flatten the slope in the steeper section. If we cannot add the additional length, then we need to add some grade controls to flatten the slope locally. This can be done by adjusting the riffle/pool spacing toward the lower end of the range. The geomorphology tells us which direction to go.

In conclusion, fluvial geomorphology can be a powerful tool to help us stabilize stream channels. By analyzing the geomorphology of the channel we can tell whether it is in balance, out of balance, or somewhere in between. The closer you can get to achieving a width, slope, and other channel geometry that correspond to the 1.5-year storm (+/-), the more likely you are to achieve a stable channel.

Jim Melvin can be reached at (913) 477-7663 or at jmelvin@ci.lenexa.ks.us.

1 Fluvial Processes in Geomorphology by Luna B. Leopold, M. Gordon Wolman, and John P. Miller. Dover Publications, Inc., 1995, ISBN: 0486685888.
2 Geomorphology by Robert V. Ruhe. Houghton Mifflin, 1975, ISBN: 039518553X.