Reach Variations in the Alluvial Sedimentation and Stratigraphy That Constrains the Age of Barrier Canyon Style Rock Art in Canyonlands National Park, Utah

The Horseshoe-Barrier Creek drainage of Canyonlands National Park lies in the heart of the rapidly eroding Colorado Plateau and contains the type section of the striking and controversial Barrier Canyon Style rock art. This study has the dual goals of: 1) constraining the hotly debated antiquity of this rock art, whether late Archaic, Paleoindian, or even older; and 2) understanding more deeply the fluvial processes functioning during Quaternary evolution of this drainage, especially exploring whether to approach alluvial records as cycles of cutting and filling versus more stochastic paleoflood sequences. These two inquiries are linked because the terrace chronostratigraphy physically and temporally constrains when the rock art could have been painted on the lower canyon walls along the drainage.

Field mapping, survey, and OSL and radiocarbon geochronology undertaken indicate distinct reach characteristics linked to underlying bedrock. The upper drainage, encountering the Kayenta Formation, is characterized by bedrock knickpoints incising a valley dominated by gravelly Pleistocene strath terraces. In contrast, the middle reach that hosts the rock art is characterized by entrenched meanders in the Navajo Sandstone and a smoothly convex longitudinal profile. The upstream straths appear to converge towards the drainage in this reach and inflate into sandy fill terraces, whereas the overall stratigraphy is dominated by internally complex Holocene paleoflood deposits.

These patterns bespeak a fundamental change in sedimentation and storage as a function of reach geometry controlled by bedrock. There is more short-term storage and preservation in the alcoves and meander bends of the middle Navajo reach, whereas the upper Kayenta catchment records longer-term incision via episodic strath formation and knickpoint incision. Finally, the Barrier Canyon rock art physically must postdate “T2” terrace formation and its subsequent incision, while predating rockfall activity at the type locality. These geochronologic constraints are being produced as this abstract is being written.