Interpolating Soil Properties Using Kriging Combined with Categorical Information of Soil Maps

• Published in: Soil Science Society of America journal Vol. 70; no. 4; pp. 1200 - 1209
• Main Authors: Liu, T.L, Juang, K.W, Lee, D.Y
• Format: Journal Article
• Language: English
• Published: Madison Soil Science Society 01.07.2006; American Society of Agronomy
• Subjects: accuracy; Clay; equations; estimation; kriging; Physical properties; Sand; Silt; simulation models; soil chemical properties; soil physical properties; Soil properties; Soil sciences; Soil types; Soils; spatial data; spatial variation; Variance analysis; Taiwan
• DOI: 10.2136/sssaj2005.0126
• ISSN: 0361-5995

Kriging interpolation is frequently used for mapping soil properties in the analysis and interpretation of spatial variation of soil. Mapping quality could affect the performance of site-specific management. Soil map-delineation in existing soil maps showing abrupt changes at the boundaries between different soil types can provide valuable categorical information for interpreting variation in soil properties. In this study, map units were used to group sampled observations, and the variation in soil properties was separated into two parts: (i) between soil types (i.e., soil type effect) and (ii) within each soil type (i.e., residual). A kriging model combined with soil map-delineation, taking into account the variation components of soil type effect and residual, was proposed. A comparison of performance of kriging combined with soil map-delineation (KSMD) and ordinary kriging (OK) was performed to assess the feasibility of KSMD for improving the interpolation of soil properties. Real data of soil properties (sand, silt, and clay contents; pH; and Mehlich-3 Ca and P) in a field of midwestern Taiwan were used for illustration. The analysis of variance table revealed that the contribution of soil types is a significant source of the spatial variation of soil. The spatial variation components of soil type effect and residual were thus determined for KSMD. When comparing KSMD and OK, the mean errors of KSMD and OK estimations were similar. However, decreases in estimation of imprecision for KSMD relative to OK (DIP %) for the 127 validation locations for sand, silt, and clay contents; pH; and Mehlich-3 Ca and P were 34, 40, 48, 20, 42, and 3%, respectively. These results suggested that the proposed KSMD method could use the information of soil map-delineation to increase precision of the interpolation of soil properties.