CORRELATION OF TERRACE TREAD SURFACES USING PYOSP: AN EXAMPLE FROM THE LICKING RIVER, KENTUCKY

River terraces have become a common target for investigations of topographic evolution, surface uplift, base-level adjustments, and climate change. Strath terraces record incision and lateral planation, whereas fill terraces record the river’s transition to aggradation and alluviation. Despite the utility, correlation of terraces is not trivial, in part because of transient incision/deposition, assumptions of surface contemporaneity, and insufficient data points along the length of the watershed. We attempted to address some of these pitfalls using a new Python library, complemented by field mapping and geochronology.

The Licking River flows from the Cumberland Plateau in central Kentucky to the Ohio River in northern Kentucky, serving as a major watershed in the Midcontinent. Recurring intervals of incision, aggradation, impoundment, and avulsion since at least the Early Pleistocene are now preserved along the Licking River Valley as a series of fluvial and lacustrine deposits, meander scars, and paleochannels that provide a terrestrial record of the entire Quaternary. In order to understand the significance of these deposits and landforms, a section of the lower reach of the Licking River was chosen to test the efficacy of fill-and-strath terrace correlation using a new Python library: PyOSP. The library has the capability of classifying a river valley (or any landform), collecting data relevant to the landform, and processing the information using swath analysis, long river and cross-sectional profiles, and histogram analyses. PyOSP classification in the study area was based on topographic position index. Elevation data were further filtered by slope (less than 5°) to clarify terrace tread signatures. Longitudinal swath profiles show a series of terrace tread surfaces that agree with previous mapping; cross swaths and profiles show similar results, but indicate decreasing gradients for alluvial fill terraces and flat surface for one lacustrine terrace. Recent geochronology shows that the lowest stratigraphic terraces date from the Holocene to latest Middle Pleistocene, and provide a minimum age for the now-buried bedrock strath surface. PyOSP provides a standardized solution to terrace identification and correlation, and its utilization will maximize field efficiency and analysis.