Exploratory Assessment of Aerial Gamma Radiometrics across the Conterminous United States

Abstract

Core Ideas US aerial gamma radiometrics were analyzed with legacy soil data. Relationships for aerial gamma and soils data differed by physiography and geology. Gamma radiometrics could best map soil texture within relatively flat landscapes. A US map where aerial radiometrics is most useful for soil mapping is provided. Gamma radiometrics have recently been connected with soil properties at field scales (10 m) and thus may have potential to characterize soil across regional scales using data from preexisting aerial surveys. However, the relationship between γ radiometrics and soil properties across these larger spatial extents (100–1000 km) has been largely unexplored, particularly within the United States. Thus, the overall purpose of this work was to test the effectiveness of aerial γ radiometrics in modeling soil properties across the conterminous United States. After discovering that up to 19% of aerial γ radiometric variability could be explained by variations in physiography and parent material type, linear regression models were created between soil properties (clay and sand contents, cation exchange capacity, CaCO3 equivalent, and pH) and aerial γ dose rate based on physiography and parent material type. Overall, aerial γ radiometrics was most frequently suitable for mapping clay content within flat landscapes of unconsolidated and sedimentary parent materials. Suitability maps for all investigated soil properties are presented. The importance of aerial γ radiometrics relative to other covariates was then assessed. The predictive power of aerial γ radiometrics, relative to other covariates, was mixed due to a non‐optimized sampling scheme. Results suggest that aerial γ radiometrics can be useful in predicting soil properties across large spatial extents after considering both physiography and parent material type. However, a true assessment of the utility of aerial γ radiometrics across the United States requires a soil sampling strategy that represents its feature space.