Morphological Analyses and Simulated Flood Elevations in a Watershed with Dredged and Leveed Stream Channels, Wheeling Creek, Eastern Ohio

Abstract

The USGS, in cooperation with the Ohio Emergency Management Agency, conducted a study in the Wheeling Creek Basin to (1) evaluate and contrast land-cover characteristics from 2001 with characteristics from 1979 and 1992; (2) compare current streambed elevation, slope, and geometry with conditions present in the late 1980s; (3) look for evidence of channel filling and over widening in selected undredged reaches; (4) estimate flood elevations for existing conditions in both undredged and previously dredged reaches; (5) evaluate the height of the levees required to contain floods with selected recurrence intervals in previously dredged reaches; and (6) estimate flood elevations for several hypothetical dredging and streambed aggradation scenarios in undredged reaches. The amount of barren land in the Wheeling Creek watershed has decreased from 20 to 1 percent of the basin area based on land-cover characteristics from 1979 and 2001. Barren lands appear to have been converted primarily to pasture, presumably as a result of surface-mine reclamation. Croplands also decreased from 13 to 8 percent of the basin area. The combined decrease in barren lands and croplands is approximately offset by the increase in pasture. Stream-channel surveys conducted in 1987 and again in 2006 at 21 sites in four previously dredged reaches of Wheeling Creek indicate little change in the elevation, slope, and geometry of the channel at most sites. The mean change in width-averaged bed and thalweg elevations for the 21 cross sections was 0.1 feet. Bankfull widths, mean depths, and cross-sectional areas measured at 12 sites in undredged reaches were compared to estimates determined from regional equations. The mean percentage difference between measured and estimated bankfull widths was -0.2 percent, suggesting that bankfull widths in the Wheeling Creek Basin are generally about the same as regional averages for undisturbed basins of identical drainage area. For bankfull mean depth and cross-sectional area, the mean percentage differences between the measured and estimated values were -16.0 and -11.2, respectively. The predominantly negative bias in differences between the measured and estimated values indicates that bankfull mean depths and cross-sectional areas in studied reaches generally are smaller than the regional trend. This may be an indication of channel filling and over widening or it may reflect insufficient representation in the regional dataset of basins with characteristics like that of Wheeling Creek. Step-backwater models were constructed for four previously dredged reaches to determine the height of levees required to contain floods with recurrence intervals of 2, 10, 50, and 100 years. Existing levees (all of which are uncertified) were found to contain the 100-year flood at only 20 percent of the surveyed cross sections. At the other 80 percent of the surveyed cross sections, levee heights would have to be raised an average of 2.5 feet and as much as 6.3 feet to contain the 100-year flood. Step-backwater models also were constructed for three undredged reaches to assess the impacts of selected dredging and streambed aggradation scenarios on water-surface elevations corresponding to the 2-, 10-, 50-, and 100-year floods. Those models demonstrated that changes in water-surface elevations associated with a given depth of dredging were proportionately smaller for larger floods due to the fact that more of the flood waters are outside of the main channel. For example, 2.0 feet of dredging in the three study reaches would lower the water-surface elevation an average of 1.30 feet for the 2-year flood and 0.64 feet for the 100-year flood.