A new method for selecting sites for soil sampling, coupling global weighted principal component analysis and a...

Area of Science:

• Soil Science
• Geostatistics
• Environmental Science

Background:

• Analyzing spatial patterns of soil properties is crucial for landscape management.
• Effective soil sampling strategies are needed to capture toposequence variability.
• Existing methods may not adequately address local soil property variations.

Purpose of the Study:

• To develop and present a hybrid methodology for optimized soil sampling stratification.
• To improve the analysis of local soil property variability and environmental factor influence.
• To ensure representative sampling locations across diverse landscapes.

Main Methods:

• Coupling of global weighted principal component analysis (GWPCA) for variance capture.
• Application of cost-constrained conditioned Latin hypercube algorithm (cLHC) for optimized sampling location selection.
• Integration of methods to analyze local variability and environmental influences.

Main Results:

• The hybrid methodology successfully optimizes sampling stratification by analyzing local soil variability.
• GWPCA effectively captures maximum local variances from global auxiliary datasets.
• cLHC optimizes the selection of representative sampling locations, suppressing sampling in less representative areas.

Conclusions:

• The developed hybrid methodology provides an optimized approach to soil sampling.
• It enhances the representation of soil properties by accounting for local structures and spatial autocorrelation.
• The method is effective in stratifying geographical space for adequate soil property representation, as demonstrated in Ghana's Guinea savannah zone.