Digital Soil Organic Carbon Assessment in the Southeastern U.S.

Human-induced impacts and biogeochemical cycling lead to profound variation of carbon storage in soils modulated by geographic and ecological factors. The objective of this study was to assess soil carbon storage across a large region in the southeastern U.S.. We used soil carbon data in the topsoil (0-20 cm) and aggregated 0-100 cm depth from the U.S. National Soil Survey Database (Natural Resource and Conservation Service, NRCS). Environmental covariate sets were assembled by STEP-AWBH variables (S: soils, T: topography, E: ecology, P: parent, A: atmosphere/climate, W: water, B: biota, and H: human) as predictor variables to develop digital soil models. We used partial least squares regression and ensemble regression trees and various error metrics to assess the prediction performance of soil carbon. Environmental predictor variables that allowed inference on soil carbon stocks in the topsoil and subsoil were identified. Results suggest that grid-based digital soil carbon stock models provide improved capabilities compared to traditional soil carbon polygon maps to characterize the variation of soil carbon across the landscape.