CORRELATION OF FLUVIAL TERRACES AND STRATIGRAPHY USING DIGITAL ELEVATION DATA, RIVER LONG PROFILES, AND HISTOGRAM ANALYSIS: AN EXAMPLE FROM THE LICKING RIVER IN NORTHERN KENTUCKY

Fluvial terraces are a cornerstone of landscape evolution studies; their sedimentology, hydrology, and erosion provide a record of evolving climate, base level, and tectonics. Investigations of these changes and other fluvial processes often rely on properly identifying erosional strath vs. depositional fill terraces, terrace stratigraphy, and landform correlation. Stratigraphy and correlation can be complicated by a relatively extensive study area that prevents complete coverage in the field, interpolating between areas of more detailed control, smaller study areas that may not have a complete geomorphic section, or asymmetric preservation of unpaired terraces. Surface profile and histogram analysis of fluvial valleys acquired from digital elevation data (surface models or point clouds) have the potential to solve these issues. The process consists of defining a stream centerline, constructing surface profiles normal to the centerline at a user-defined spacing, histogram analysis of elevations along surface profiles, and plotting long profile distance against histogram peaks; peaks thus represent mean tread surfaces along the stream long profile.

The Licking River, a tributary of the Ohio River and the Pliocene–Early Pleistocene Teays River, has been a major drainage throughout the Quaternary. In northern Kentucky, a succession of terraces has developed during 250 ft of incision in the lower reach since the Pleistocene. Geologic mapping in the 1960s and ‘70s was based on 1:24,000-scale topographic base maps, and resulted in no differentiation of terrace generations. Mapping in 2015-16 was based on 5-ft digital elevation models, but was confined to 7.5-minute quadrangle boundaries, which prevented detection of the entire terrace stratigraphy. The histogram analysis of terrace treads was then performed for 25 km upstream from the mouth. Five regional terrace systems—Qat1 at 500–510 ft, Qlt1 at 530 ft, Qat3 at 560–570 ft, and Qat4 at ~600 ft—and multiple intermediate terraces are clearly distinguished by mean tread elevations relative to the Licking River long profile. This method is applicable to landscape evolution projects based on correlation and stratigraphy of fluvial terraces, but is limited by the availability of accurate elevation data and the scale of landforms of interest.