Assessing the contribution of legacy sediment and mill dam storage to sediment budgets in the Piedmont of Maryland.

The disparity between watershed erosion rates and downstream sediment delivery has been an important theme in geomorphology for many decades, and the role of floodplains in sediment storage has frequently been cited. In the Piedmont province of the eastern US, upland deforestation and agricultural land use following European settlement led to deposition of thick packages of overbank sediment in valley bottoms, commonly referred to as legacy deposits. Previous authors have cited geomorphic evidence that some of this legacy sediment has been remobilized by lateral migration of channels. Recent interest has focused on the role of historic mill dams as a cause of floodplain aggradation, and on mill-dam deposits as a potentially important source of sediment exported downstream. This thesis seeks to address several outstanding questions: 1) How rapidly is sediment being remobilized from Baltimore County floodplains by channel migration and bank erosion? 2) What proportion of this sediment is derived from legacy deposits? 3) Is the presence of mill dams a necessary prerequisite for aggradation of legacy sediment? 4) How important are mill-dam deposits by comparison with other sediment sources contributing to downstream sediment yields? In order to answer these questions, we measured gross and net bank erosion rates along 25 valley reaches from 18 watersheds with drainage areas between 1.8 and 155 km2 in northern Baltimore County, Maryland, including sites with and without mill dams. We compared stream morphology from LiDAR-based digital elevation data collected in 2005 with channel positions recorded on 1:2400 scale topographic maps from 1959-1961 in order to quantify nearly 50 years of channel change. Sediment samples were collected from stream banks and bars in order to characterize bulk density and particle-size distribution for use in converting volume to mass for comparison with other sediment sources. Average annual lateral migration rates ranged from 0.06 - 0.19 m/year, which represented an annual migration of 3% (0.9-5.5%) channel width across all sites, suggesting that channel dimensions may be used as reasonable predictors of bank erosion rates. Gross bank erosion rates varied from 50 to 310 Mg/km/yr (median = 114) and were positively correlated with drainage area. Measured in-channel deposition accounted for an average of 47% (28-75%) of gross erosion, with deposition increasing in importance at larger drainage areas. Legacy sediments accounted for 6-90% of bank erosion at individual sites, represented about 60% of bank height at most exposures, and accounted for 67% of the measured gross erosion. Extrapolating the results indicated that first- and second-order streams account for 59% of total stream bank erosion from 1026 km2 of northern Baltimore County. After accounting for estimated redeposition, net stream bank sediment yields (74 Mg/km2/yr) constitute 71% of the estimated average Piedmont watershed yield (104 Mg/km2/yr). Field evidence confirmed that legacy sediments existed at sites with and without mill dams. At the reach scale, mill-dam deposits contributed an average of 14.8% more sediment per unit length than typical stream channels, but the cumulative importance of mill-dam deposits as sediment sources decreased with increasing spatial scale. Cross-scale comparisons suggest that potential sediment contributions from mill-dam deposits may peak along low- to mid-basin reaches (10-40 km2). The results suggest mill dams should not be universally regarded as dominant sources of sediment from Piedmont tributaries to the Chesapeake Bay