61-4 Utilizing Drone-Mounted Multispectral Cameras for Paleontological Surveys, Nebraska (USA)

Session: Advancing Geologic Analysis with Digital Outcrops and Close-Range Remote Sensing Data (Posters)

Presenting Author:

Authors:

Abstract:

Traditional field surveys for paleontological materials often face significant challenges, including remote locations, high transportation costs, difficult terrain, adverse climate conditions, complex logistics, and substantial time commitments. To help mitigate these limitations, remote sensing technologies offer an efficient method for identifying potential fossil-bearing sites across large areas. In this study, a fixed-wing drone was equipped with a S110 multispectral camera system (RGB/NIR/RE) to survey an approximately 500,000 meter² area in northwest Nebraska (USA), known for early Cenozoic fossil vertebrates. Data were collected at a 30-meter altitude, yielding a near-infrared (NIR) pixel resolution of 0.13 m and RGB resolution of 0.036 m. After initial processing two classification techniques—Interactive Supervised Classification (ISC) and Pixel Inspector (PI)—were applied using ArcGIS software. By analyzing the geochemical spectral signatures associated with bone phosphate, several potential fossil localities were identified and grouped into 23 target areas. The ISC method used known fossil vertebrate NIR signatures in the area (1x1 pixel) as a standard to detect similar spectra across the study area. While effective in highlighting general areas of interest, ISC did not define locations of specific sites due to an excessive range in spectra values. In contrast, PI enabled for finer control by distinguishing pixel values “within range” versus “out of range”, facilitating the generation of semi-transparent overlays to visualize fossil hotspots on RGB imagery. PI analysis then generated the 23 target areas which were further ranked on their area. Although both methods showed promise, further refinement is necessary to improve model accuracy. Two key factors were identified: (1) the local background mineralogy—particularly high concentrations of montmorillonite (a clay mineral) in the study area—can interfere with bone phosphate detection, thus understanding local mineralogy is paramount and (2) fossil diagenesis may alter spectral responses, complicating interpretation. This latter issue might be partially addressed through laboratory spectral analyses of fossil samples in varying diagenetic states. In this study depositional patterns were used to resolve some challenges with clay exposures. Overall, this study demonstrates the potential of drone-mounted multispectral imaging systems as powerful tools for paleontological exploration, offering the ability to efficiently survey large areas and prioritize fieldwork in promising locations.

Geological Society of America Abstracts with Program. Vol. 57, No. 6, 2025

doi: 10.1130/abs/2025AM-8532

© Copyright 2025 The Geological Society of America (GSA), all rights reserved.