Spatial and temporal patterns of sedimentation in an infilling reservoir

摘要

River dams are ubiquitous globally, altering the connection of up- and downstream environments. In particular, they store vast amounts of sediment in their upstream reservoirs, many of which are approaching storage capacity. There is much interest in sediment management for these systems, even though quantifying sedimentation rates remains somewhat challenging. This study focuses on one such system, Conowingo Reservoir, upstream of the last and largest dam (Conowingo Dam) on the Susquehanna River before it enters Chesapeake Bay. The trapping capacity of Conowingo Reservoir has decreased since its creation in 1928, potentially increasing fluvial loading to the Bay. In addition, large flood events can scour much more than the average annual sediment load from the Reservoir bottom, delivering exceptionally large sediment loads to the Bay within days. Predicting the downstream impacts of both Reservoir infill and scour events is essential for management; however, these predictions are limited by understanding of the complex spatial and temporal variability of sedimentation. To evaluate this variability, sedimentation rates are determined throughout the Reservoir over seasonal to decadal time scales with the naturally occurring radioisotopes Be-7 and Pb-210 (half-lives 53.3 days and 22.3 years, respectively), as well as anthropogenically produced Cs-137 (half-life 30.7 y) as an independent geochronometer. These rates are placed within the context of sediment character (mud content) and fluvial sediment supply to the Reservoir, as well as its geomorphology. In general, seasonal sedimentation rates scale with fluvial loads and reflect deposition of watershed-derived sediment. Spatial patterns of seasonal sedimentation follow those of river deltas incised by channels, while decadal-scale sedimentation rates increase downstream. Episodic event sedimentation in channels is suggested by the character of Pb-210 profiles, which indicate non steady state sedimentation that makes application of sediment-age dating models challenging in this system. Although sediment mud content has clearly decreased as the Reservoir has filled, observed changes in sedimentation rates may reflect biases of these models rather than physical changes.