DRONE-BASED MEASUREMENT OF STRATIGRAPHIC COLUMNS AND CROSS-SECTIONS

Measurements of stratigraphic columns and cross-sections are commonly hampered by topography, such as ravines and steep cliffs, which make some areas inaccessible and cause gaps in data coverage. Structural tilting can also complicate these measurements. We describe a solution to both problems using drone-based photogrammetry, which we have developed and tested on an exposure previously measured with traditional jacob-staff techniques. We begin by obtaining a series of photographs of an outcrop using a DJI Phantom 3 Advanced quadcopter drone, and merging these into a photomosaic and a 3-D point cloud, using Agisoft Photoscan. Multiple x-y-z coordinates along a dipping bed were obtained with LAStools and processed with an R script (available on the UGA Stratigraphy Lab website) to calculate the strike and dip, which was then used to rotate the entire exposure such that all beds are restored to stratigraphically horizontal positions. In this configuration, the Z-coordinate of any horizon in the point cloud is its stratigraphic elevation, which permits accurate measurement of a stratigraphic column. Comparison of this approach to a traditionally measured section through the Silurian Clinch Formation at the Hagan railroad cut in southwest Virginia produced comparable results. Ginn (2014) reported an average strike of 257° and an average 52.7° dip to the north, which compares well to our drone-based strike of 242° and dip of 53.4°. Differences in strike may be the result of differences in a magnetic vs. geographic reference frame. An interval of strata that Ginn (2014) measured as 35.9 m thick is 33.7 m thick when measured from the drone, and this difference more likely reflects the difficulties of accurate measurements with a Jacob Staff. This drone-based approach has promise not only for the measurement of relatively simple vertical columns, but also for three-dimensional characterization of outcrops that show considerable lateral variation in facies and thicknesses. Similar approaches have potentially wide application across the earth sciences, including volcanology, glaciology, structural geology, and planetary geology.