iSDAsoil: Machine Learning-Predicted Soil Properties for Africa

iSDAsoil provides predicted soil property maps for the entire African continent. The resource includes over 20 different soil property predictions at two depths (0-20 and 20 50cm), generated using machine learning models trained on more than 100,000 analyzed soil samples and remote sensing data. It was created by Innovative Solutions for Decision Agriculture (iSDA).

Use Cases

• Model crop suitability based on predicted soil properties.
• Assess soil health and degradation risks for conservation planning.
• Validate or improve regional soil maps using the provided model error estimates.
• Analyze relationships between satellite covariates and soil characteristics.

Strengths

• Covers the entire African continent, providing continent-wide consistency.
• Includes predictions for over 20 different soil properties.
• Models were trained on a dataset of over 100,000 analyzed soil samples.
• Provides data at two standard depth intervals (0-20cm and 20-50cm).

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Last update date is unknown; freshness unverified.

Provenance

Source

Innovative Solutions for Decision Agriculture (iSDA)

Collection Method

Generated using machine learning models coupled with remote sensing data and a training set of analyzed soil samples.

Time Range

null

Freshness

Last updated date is unknown.

Geography

African continent