IMAGING THE DYNAMICS OF CLIFF RETREAT WITH STRUCTURE-FROM-MOTION 3D RECONSTRUCTIONS, COLORADO PLATEAU

The retreat of cliffbands is a dominant mode of erosion in layered sedimentary rocks containing beds of variable strength. The development of the landscape in these very common settings is the end result of incision by trunk streams followed by cliff retreat across vast areas. For example, the iconic landscapes of the Grand Canyon, Canyonlands National Park, and Monument Valley owe their unique forms to this general process. Once the cliffs have migrated away from the trunk stream floodplain, their retreat is locally controlled. The broad-scale response to incision in these landscapes therefore depends strongly upon the local dynamics of cliff retreat.

Recent theoretical work has shown that the mechanics of these cliffband landscapes emerge from the interaction of resistant rockfall debris with first-order fluvial channels below the cliff. These interactions can have a primary influence on the retreat rate of the cliffband as a whole. Here, we use high-resolution digital terrain models, generated by structure-from-motion (SFM) 3D reconstruction, to examine the mechanisms of erosion and sediment transport on the slopes below cliffs in a natural experiment site on the Colorado Plateau.

Using as a case study a simple, sandstone-over-shale stratigraphy with a continuous gradient in cliffband height, this work uses a combination of digital and field methods to constrain key terms such as the relationship between bedrock fracture spacing and the size distribution of talus, rates of rockfall and talus accumulation below the cliff, rates of in situ weathering and breakdown of this talus as compared to rates of transport and redistribution of the talus, and rates of erosion of the soft rock upon which the talus is deposited. In particular, we use digital terrain from SFM to estimate relative rates of erosion on debris-covered and debris-free portions of the shale slopes below the sandstone cliff using hilltop curvature as a proxy for erosion rate. We use the resulting measurements and relationships to constrain numerical models of cliff-debris-erosion interaction.