BLUFF OR UPLAND SEDIMENT? USING TERRESTRIAL COSMOGENIC NUCLIDES TO DETERMINE SEDIMENT SOURCES IN RIVERS

The Minnesota River and its tributaries have a suspended sediment load that exceeds EPA standards. Lake Pepin, a natural reservoir for the Minnesota/Upper Mississippi River watershed, has a deposition rate of 5-10 times pre-settlement rates. Before mitigation programs can begin, the sediment source must be identified. Initial research focused on two possible sediment sources: (1) bluff erosion of sharply incised incised stream channels and (2) topsoil erosion from flat uplands. Further investigation indicate contributions due to gullying may also be significant. Our goal is to use sediment tracing methods to characterize the origin of river sediments throughout the Minnesota River basin. Several techniques will be employed in an overall pilot study to determine the best combination of methods. These techniques include: cosmogenic nuclides [26Al and 10Be], fallout radionuclides [210 Pb and 137Cs], and sorbed nutrients [N and P]. Differences in lithologic source between the Des Moines and older tills allow the use of lithologic characterization techniques, such as mineral point counts, elemental analysis, and magnetic characterization.

Preliminary work focuses on characterizing the cosmogenic nuclide signatures - 26Al:10Be ratios and concentrations - from end member type locations in a small tributary basin, then transferring that knowledge to more complicated systems. Beauford Creek, our upland source type area, dissects agricultural land. The drainage system was significantly extended in these farm fields, primarily through ditches and drainage tiles, a common practice in the Minnesota River basin. The Le Seuer River, our bank source type location, is a deeply incised river that cuts through Des Moines lobe tills as well as Old Gray tills with steep banks reaching over 30 meters above water level.

Higher cosmogenic nuclide ratios and concentrations are expected in upland soil sediment, and lower ratios and concentrations are expected from bluff erosion. Gullying would complicate sediment sourcing models, with the same 26Al:10Be ratio as soils, and intermediate to low concentrations (similar to a mixture of soil and bluff sediments). Cosmogenic nuclides should provide initial constraints on sediment sources. Further work using fallout radionuclides should further our ability to distinguish between soil and gully sources.