LITHOLOGY INTERPRETATION USING LIDAR RGB AND INTENSITY VALUES; A CASE STUDY FROM THE EOCENE GREEN RIVER FORMATION, EASTERN UTAH

The Eocene Green River Formation in east-central Utah hosts one of the largest deposits of oil shale in the world. Within the Green River Formation, the Parachute Creek Member is composed of alternating organic rich carbonate mudstones and organic lean siliciclastic deposits. This study focuses on a laterally extensive (~6.4 km long and ~1.4 high) northwest to southeast trending outcrop exposure of the Parachute Creek Member near Evacuation Creek, Utah. Previous detailed sedimentologic and stratigraphic evaluation of this outcrop offers an excellent opportunity to geologically groundtruth remote sensing investigation technologies, such as LiDAR (Light Detection and Ranging) and photogrammetry These remote sensing technologies produce georeferenced, x-y-z “point clouds” representing the surficial expression of an outcrop with high levels of spatial accuracy (±1-2 cm) at scales ranging from centimeters to tens of kilometers. In addition to extremely precise geospatial data, new LiDAR tools return Red-Green-Blue (RGB) as well as default intensity values that are a function of the composition and character of the imaged surface. Recent developments have seen RGB and intensity LiDAR data translated to characterization of various sedimentary components, e.g., siliciclastics, clays, and carbonates. This study uses the RGB and intensity values to highlight the differences between lithologic packages, internal heterogeneity within a single lithologic unit, and distinguish between organic rich and organic lean zones.

A ground-based LiDAR survey using Leica ScanStation C10 scanners produced over 686.9 million points of the study area. Lithologic units were differentiated based on previously measured section and core data and RGB and intensity values were extracted. Organic lean, siliciclastic units resulted in a bimodal distribution of larger values whereas organic rich, carbonate mudstone units showed a unimodal distribution skewed towards lower values. Internal heterogeneity within units shows bimodal distributions with both high and low values. To better analyze these data, shadow and light inhibitors will need to be removed to reduce uncertainty in lithologic interpretation and identification. This proof of concept provides valuable lithologic predictive power using LiDAR outcrop data alone.