Research Topics > Geomorphology: Self-Maintaining Ditches > printer-friendly format

Using Knowledge of Fluvial Processes
to Design Self-Maintaining Agricultural Drainage Ditches
in the Midwestern Region of the USA

Andy Ward and Larry Brown, Professors,
Department of Food, Agricultural and Biological Engineering,
The Ohio State University,
590 Woody Hayes Drive, Columbus, Ohio 43210, U.S.A.
Email: ward.2@osu.edu.

Dan Mecklenburg, Ecological Engineer,
Ohio Department of Natural Resources,
4383 Fountain Square Drive, Columbus, Ohio 43224, U.S.A.
Email: dan.mecklenburg@dnr.state.oh.us.

Andy Ward measuring dimensions	Dan Mecklenberg measuring water depth	Larry Brown in the field

Summary:

The primary goal of the study is to identify how knowledge of fluvial processes might be used to make drainage ditches more self-maintaining and also enhance the ecology of these systems.

Ditch with fluvial features

In many productive agricultural areas in the U.S.A., ditches have been constructed and natural channels deepened and straightened to facilitate the flow of water from agricultural subsurface drainage outlets and to maximize conveyance. Work done periodically to maintain the drainage function typically includes removal of woody vegetation and deposited sediment. Ancillary work includes stabilizing bank slope failures and toe scour.

Before bench removal	After bench removal	Rip-rap restricts wide bench formation

Ditch form is a result of construction and maintenance but also, to verifying degrees, due to fluvial processes that form a meandering channel and benches near the bottom of many ditches. In Northwest Ohio, the channel dimensions formed by fluvial processes are related to drainage area, d₅₀ and d₈₄ of the bed material. The formed channel and adjacent flat vegetated benches are associated with discharges that might occur 10-20 times annually and are strongly influenced by subsurface drainage flow. The new knowledge resulting from this study is being used to retain fluvial features when improving maintenance practices and in the design of ditches.

Subsurface drains	Top-right and bottom-right photos: Unstable low benches that scour during events that exceed the channel full discharge

Site Location and Methods:

The study was conducted in the upper reaches of the Portage River Watershed that is located in a flat (bed slopes generally less than 0.4%) region of Northwest Ohio that was once an ancient lakebed surrounded by glacial moraines. The area was originally known as the Great Black Swamp, and was dominated by deciduous forest. Cleared and drained extensively in the last 150 years it is now dominated by row crop agriculture. Most fields have subsurface drainage as the silty clay soils are poorly drained.

Ditch crossing flat drained fields	Anne Weekes and Andy Ward measuring channel dimension, pattern, and profile.

Features measured were: channel cross-sections along a 100-300 m reach; bed profile; water surface profile; azimuth; top of the bank; and bed material particle size distribution.

Individual ditches and streams are ungaged but daily and annual peak discharge data for a 60 year period of record were obtained for the USGS gauge at Woodville that measures discharge for a 1100 km2 drainage (see figure below). The flow data was used to develop discharge-recurrence interval relationships, based on the annual peaks method, and to determine the mean number of annual events of different magnitudes.

Ditches in the study area

Discharge estimates were obtained using an empirical procedure that relates the peak discharge for a particular recurrence interval to the drainage area, channel bedslope, and the percentage of the area occupied by lakes, ponds, and wetlands. Stage-discharge relationships were obtained using Manning's equation and a mean conceptual geometry for each site.

Results:

Descriptive information on the ditches studied

All reaches exhibited one to four grade breaks up the side slopes (banks) of the ditches. At most sites the dominant grade break was between a small channel and a low depositional bench. (See Figure 1 below.)

Figure 1: Measuring grade breaks

All the benches have coarse material at the elevation of the bed of the fluvial channel. Above the coarse layer there are fine materials interspersed with one to three poorly defined thin sandier layers that might also contain some gravel. (See Figures 2 and 3 below.)

Figure 2	Figure 3

There was good agreement between precipitation, stream gauge and pressure transducer data, ADAPT daily simulation model discharge estimates, empirical discharge predictions, calculations based on Manning's equation and channel full discharge being associated with events that occur 10-20 times annually. (See Figure 4 below.)

Figure 4: Precipitation and stage data

Comparisons of ditches to natural streams

The probable dimensions of the fluvial channel can be empirically determined based on regional studies similar to those that are conducted on natural streams. (See Figures 5 and 6 below.)

Figure 5

Figure 6

The channels exhibit some riffle and pool features. The pattern and profile characteristics of the channels were not as well defined as in natural streams and did not consistently fit expected relationships with bankfull. (See Figures 7 and 8 below.)

Figure 7: Wide benches	Figure 8: Narrow benches

Findings on ditches with stable versus unstable benches

Stable low benches are covered by dense grass and the channel meandered slightly within the confines of the ditch. (See Figure 9 below).

Figure 9

Stable benches form in wider ditches that provide less confinement. (See Figure 10 below.)

Figure 10

The width of the benches and the ratio of the bench width to channel width were significantly different. The mean widths for the stable and unstable bench groups were 8.3 m and 2.9 m respectively. There were also significant differences between the two groups for the width of the flood prone area, the width at twice the maximum channel full depth and the ditch width at three times the average channel mean depth for all the sites. These measurements are indicators of the confinement of the flow and the ability of the flow to dissipate energy as the flow exceeds the channel full stage. (See Figures 11 and 12 below.)

Figure 11: Wide benches	Figure 12: Narrow benches

For the stable benches a ratio of the bench width to channel width at the bench full stage was highly correlated (R² = 0.84) with a linear combination of drainage area and d₈₄ of the bed material. As the drainage area and d₅₀ increased the ratio decreased. (See the table below.)

For the stable benches, the particle size at incipient motion is similar to the measured d₅₀ or between the d₅₀ and d_{84. (See Figures 13 and 14 below.)}

Figure 13

Figure 14

Conclusion

The probable dimensions of the low-flow channel can be empirically determined based on regional studies similar to those that are conducted for natural streams. Similarly, measurement of the study sites begins to suggest that a broad ditch with a total bench width approaching or exceeding the channel width will result in stable benches.

The fluvial channel within a ditch has the potential to provide better habitat for biota in the same way that a bankfull channel is superior to an over-widened trapezoidal channel. It might also be useful in improving water quality particularly for nutrient assimilation in headwater streams.

Acknowledgements
This research was funded by a grant from the Great Lakes Protection Fund and through in kind support from numerous local, state and federal agency. The assistance of Pete Richards (Heidelberg College), Ed Rankin (Ohio EPA), D.J. Mears (Office of the Wood County Engineer), Robert McCall (Ohio State University Extension Service), and Virginie Bouchard (The Ohio State University) is greatly appreciated. Anne Weekes provided excellent leadership to the geomorphology field work.

Ed Rankin, Andy Ward, and Anne Weekes electro-fishing.