GSA Annual Meeting in Denver, Colorado, USA - 2016

Paper No. 303-2
Presentation Time: 2:05 PM

PALEOFLOOD ANALYSIS OF MEANDER SCARS AND NATURAL LEVEES OF THE UPPER TENNESSEE RIVER BASIN, BLUE RIDGE MOUNTAINS (Invited Presentation)


LEIGH, David S., Department of Geography, The University of Georgia, Geog.-Geol. Building, 210 Field St., Room 204, Athens, GA 30602, dleigh@uga.edu

The humid environment and crystalline bedrock of the Blue Ridge Mountains inhibits preservation of peak-stage slackwater deposits along gorges due to erosion, bioturbation, and lack of rock shelters and caves. Given the paucity of such slackwater deposits, vertical accretion deposits in natural levees and meander scars are used instead to reconstruct paleofloods. Sediment columns were sampled in contiguous one- to three-centimeter intervals down cores or outcrops at several sites along the upper Little Tennessee and Tuckaseegee Rivers. Multi-decadal gaging records and cesium-137 dating are used to calibrate sedimentology to past floods. Relatively high percentages of sand, particularly the coarsest sand fractions, are indicative of large floods. However, sand percentages typically must be normalized to mask effects of upward-fining or upward-coarsening depositional trends. Based on historical analogs, floods with annual recurrence probabilities <0.10 produce distinct sandy flood laminae, whereas smaller floods cannot be discriminated by sedimentology. Multiple stratigraphic records are necessary to capture all past floods, because deposits of large floods may be absent at one site versus another. Relating sand percentages to gaged flood magnitudes produces weakly significant correlations, which appear to be confounded by abundance of mica and inability to precisely sample laminae representing the peak flood stage. Ongoing research seeks to improve correlations with flood magnitudes by using thinner sample slices and segregating minerals for particle size analysis with digital images. In general, the vertical accretion deposits produce a good record of the frequency of large floods, but resolving flood magnitudes remains a challenge.

Backcasting sand sedimentology to the prehistoric late Holocene reveals periods of enhanced flooding at the beginning and end of the Medieval Climate Anomaly. Sedimentation rates during the prehistoric Holocene are about an order of magnitude slower than during historic time, which blurs the prehistoric resolution considerably. Nonetheless, several late Holocene stratigraphic sections, and one other that spans the entire Holocene, suggest that historical floods are much larger than those of the prehistoric Holocene.