A new digital soil mapping approach based on the adjacency effect.

Area of Science:

• Environmental Science
• Soil Science
• Geospatial Analysis

Background:

• Accurate soil mapping is essential for effective agriculture, land management, and environmental conservation.
• Digital Soil Mapping (DSM) is a widely adopted method, but its accuracy can be further improved.
• Incorporating spatial context, such as neighborhood effects, is a promising avenue for enhancing DSM.

Purpose of the Study:

• To develop and evaluate a novel DSM strategy that integrates the neighborhood effect of environmental covariates (ECs).
• To improve the mapping accuracy of key soil properties: organic carbon (OC), cation exchange capacity (CEC), bulk density (BD), and pH.
• To compare the performance of the proposed method against a conventional DSM approach.

Main Methods:

• Utilized 18 ECs from the Soil Landscapes of Canada dataset and satellite imagery for soil property modeling in southern Canada.
• Compared a conventional DSM approach with a novel strategy incorporating neighboring ECs via Inverse Distance Weighting (IDW).
• Employed Gaussian Process Regression (GPR) for both modeling strategies.

Main Results:

• The proposed strategy significantly reduced mean absolute errors for OC (32%), CEC (36%), BD (28%), and pH (14%).
• Improved R-squared values and decreased high-error areas across all modeled soil properties.
• Reduced mean uncertainty in soil property modeling by 3.4% to 3.9%.

Conclusions:

• Considering the neighborhood effect of ECs is crucial for enhancing DSM accuracy.
• The proposed IDW-based neighborhood approach offers a more nuanced and accurate method for soil property modeling.
• This improved soil mapping has substantial implications for sustainable land management, precision agriculture, and environmental conservation.