Importance and strength of environmental controllers of soil organic carbon changes with scale

• DOI: 10.1016/j.geoderma.2020.114472
• Publication Date: 2020-06-23T05:22:51Z
• Authors: K. Adhikari; U. Mishra; P.R. Owens; Z. Libohova; S.A. Wills; W.J. Riley; F.M. Hoffman; D.R. Smith
• Keywords: Soil carbon; Spatial heterogeneity; Earth system science

Spatial heterogeneity in environmental factors on the land surface moderates exchanges of water, energy, and greenhouse gases between atmosphere. However, appropriately representing this earth system models remains a critical scientific challenge. We used large dataset (n = 31) soil-forming factors, field observations soil organic carbon (SOC) 6213), machine-learning algorithm (Cubist) to analyze scaling behavior SOC across conterminous United States. found that various are significant predictors stocks at different spatial scales. Out 31 we investigated, only 13 were scales ranging from 100 m 50 km. Overall, topographic variables had higher influence finer scales, whereas climatic more important coarser The model performance worsened with increasing scale or resolution prediction (R2 0.38–0.65). strength controls (median regression coefficient) weakened scale, represented them using mathematical functions 0.38–0.98). Both mean variance decreased linearly soils Fitted linear accounted for 81% 82% variability SOC, respectively. also relationships among high-order moments 0.51–0.97). Improved understanding their controllers can improve benchmarking may eventually representation biogeochemistry.