Premium
Layer‐Specific Analysis and Spatial Prediction of Soil Organic Carbon Using Terrain Attributes and Erosion Modeling
Author(s) -
Dlugoß Verena,
Fiener Peter,
Schneider Karl
Publication year - 2010
Publication title -
soil science society of america journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.836
H-Index - 168
eISSN - 1435-0661
pISSN - 0361-5995
DOI - 10.2136/sssaj2009.0325
Subject(s) - subsoil , soil carbon , topsoil , environmental science , soil science , topographic wetness index , tillage , kriging , digital soil mapping , bulk density , terrain , spatial variability , geostatistics , soil horizon , erosion , hydrology (agriculture) , soil map , soil water , geology , remote sensing , mathematics , digital elevation model , agronomy , geography , statistics , paleontology , cartography , geotechnical engineering , biology
High‐resolution soil organic C (SOC) maps are a major prerequisite for many environmental studies dealing with C stocks and fluxes. Especially in hilly terrain, where SOC variability is most pronounced, high‐quality data are rare and costly to obtain. In this study, factors and processes influencing the spatial distribution of SOC in three soil layers (<0.25, 0.25–0.50, and 0.5–0.90 m) in a sloped agricultural catchment (4.2 ha) were statistically analyzed, utilizing terrain parameters and results from water and tillage erosion modeling (with WaTEM/SEDEM). Significantly correlated parameters were used as covariables in regression kriging (RK) to improve SOC mapping for different input data densities (6–38 soil cores ha −1 ) and compared with ordinary kriging (OK). In general, patterns of more complex parameters representing soil moisture and soil redistribution correlated highest with measured SOC patterns, and correlation coefficients increased with soil depth. Analogously, the relative improvement of SOC maps produced by RK increased with soil depth. Moreover, an increasing relative improvement of RK was achieved with decreasing input data density. Hence, the expected decline of interpolation quality with decreasing data density could be reduced, especially for the subsoil layers, by incorporating soil redistribution and wetness index patterns in RK. The optimal covariable differed among the soil layers. This indicates that bulk SOC patterns derived from topsoil SOC measurements might not be appropriate in sloped agricultural landscapes; however, generally more complex covariables, especially patterns of soil redistribution, exhibit a great potential to improve subsoil SOC mapping.