Soil Buffering Capacity Titration Data from Six African Countries

543 soil samples from Ivory Coast, Ghana, Malawi, Zimbabwe, Zambia, and Mozambique were analyzed using a multidimensional titration framework. Random Forest models predicted three buffering indices from chemical properties with R² up to 0.73, while MIR spectroscopy models achieved R² up to 0.81. The dataset likely contains titration-derived indices and associated soil property measurements.

Use Cases

• Predicting soil buffering capacity based on organic matter, pH, cation exchange capacity, and texture
• Comparing machine learning (Random Forest, XGBoost) and spectroscopy (MIR-PLSR) models for soil property prediction
• Classifying soils into low, medium, and high buffering capacity categories for lime-rate calibration
• Analyzing the relationship between different buffering indices (BCP, BCL, BCI) and soil chemical properties

Strengths

• Includes 543 soil samples from six distinct African countries
• Provides three complementary buffering indices derived from titration curves
• Model performance metrics are reported (R² up to 0.81, RMSE up to 3.20)

Limitations

• Column-level documentation is absent; field semantics must be inferred after download
• Row count is unknown, which may limit suitability assessment
• The dataset is very small (14.9 KB), suggesting limited scope

Provenance

Source

figshare

Collection Method

Soil samples analyzed via controlled calcium hydroxide titration and MIR spectroscopy.

Freshness

Last updated 2026-05-22 05:33:48; freshness should be verified

Geography

Ivory Coast, Ghana, Malawi, Zimbabwe, Zambia, Mozambique