Submission Type
Poster
Abstract
This study investigated the efficacy of utilizing visible, near-infrared (VNIR), and short-wave infrared (SWIR) imagery for geologic mapping. Samples collected from Kelso Station, CA, provided ground truth for field and digital mapping of the region at a 1:8000 scale using ArcGIS Pro and georeferenced sample sites. Aerial data sets of the Kelso region were gathered from Advanced Spaceborne Thermal Radiometer (ASTER) multispectral data and processed in ENVI utilizing nine bands within the VNIR (0.4-1.0 µm) and SWIR (1.0-3.0 µm) wavelengths. Regions of interest were created using Spectral Angle Mapper (SAM) utilizing three classifications: georeferenced sample lithology, USGS map lithology, and georeferenced sample lithology with varnished alluvium assumption. Comparison to corresponding geologic maps revealed the effectiveness of VNIR and SWIR data in predicting specific rock lithology. In every classification scheme, SAM succeeded in determining carbonate bedrock and quartz sand. Disparities in the recognition of intrusive igneous, metamorphic, and basalt varieties occurred when comparing SAM to geologic maps. These discrepancies between ground truth and remote sensing are likely from physical limitations of the multispectral data: silicate minerals and Bowen’s Reaction Series minerals (except iron) lack absorption features in the VNIR/SWIR wavelengths thus limiting distinct classification. Additionally, VNIR/SWIR wavelengths maintain micron-deep penetration which induces susceptibility to surficial alteration and varnishing processes. The results of the study concluded that lithology-classified ASTER data using SAM is valuable in the identification of carbonate and aeolian quartz sand but maintained low accuracy in distinguishing basalt, intrusive igneous, and metamorphic rocks.
Recommended Citation
Wing, Allison, "188-Efficacy and Limitations of Geologic Remote Sensing: Comparing Ground Truth to Spectrally Identified Lithology, Mojave National Preserve, California" (2024). GREAT Day Posters. 48.
https://knightscholar.geneseo.edu/great-day-symposium/great-day-2024/posters-2024/48
188-Efficacy and Limitations of Geologic Remote Sensing: Comparing Ground Truth to Spectrally Identified Lithology, Mojave National Preserve, California
This study investigated the efficacy of utilizing visible, near-infrared (VNIR), and short-wave infrared (SWIR) imagery for geologic mapping. Samples collected from Kelso Station, CA, provided ground truth for field and digital mapping of the region at a 1:8000 scale using ArcGIS Pro and georeferenced sample sites. Aerial data sets of the Kelso region were gathered from Advanced Spaceborne Thermal Radiometer (ASTER) multispectral data and processed in ENVI utilizing nine bands within the VNIR (0.4-1.0 µm) and SWIR (1.0-3.0 µm) wavelengths. Regions of interest were created using Spectral Angle Mapper (SAM) utilizing three classifications: georeferenced sample lithology, USGS map lithology, and georeferenced sample lithology with varnished alluvium assumption. Comparison to corresponding geologic maps revealed the effectiveness of VNIR and SWIR data in predicting specific rock lithology. In every classification scheme, SAM succeeded in determining carbonate bedrock and quartz sand. Disparities in the recognition of intrusive igneous, metamorphic, and basalt varieties occurred when comparing SAM to geologic maps. These discrepancies between ground truth and remote sensing are likely from physical limitations of the multispectral data: silicate minerals and Bowen’s Reaction Series minerals (except iron) lack absorption features in the VNIR/SWIR wavelengths thus limiting distinct classification. Additionally, VNIR/SWIR wavelengths maintain micron-deep penetration which induces susceptibility to surficial alteration and varnishing processes. The results of the study concluded that lithology-classified ASTER data using SAM is valuable in the identification of carbonate and aeolian quartz sand but maintained low accuracy in distinguishing basalt, intrusive igneous, and metamorphic rocks.