Africa Soil Information Service (AfSIS) Soil Chemistry

Sub-Saharan Africa soil samples from 19 countries collected between 2009 and 2013. The dataset pairs infrared spectral data with conventional soil property measurements, released by the Africa Soil Information Service (AfSIS) project. It is intended to form a training set for machine learning models to predict soil properties via spectroscopy.

Use Cases

• Train machine learning models to predict soil properties based on infrared spectral data.
• Analyze soil chemistry variation across 19 Sub-Saharan African countries.
• Calibrate and validate less expensive spectral measurement techniques against conventional lab tests.
• Support agricultural planning and food security research with georeferenced soil data.

Strengths

• Data collected from 19 countries using a statistically sound sampling scheme.
• Pairs two measurement types: conventional soil tests and infrared spectroscopy.
• Project conducted over a defined period from 2009 to 2018, with this release covering Phase I (2009-2013).

Limitations

• Row count and dataset size are unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Last update date is unknown; freshness unverified.

Provenance

Source

Africa Soil Information Service (AfSIS) project, organization QED.

Collection Method

Georeferenced samples collected and analyzed using conventional soil testing and infrared diffuse reflectance spectroscopy.

Time Range

2009-2013 (Phase I).

Freshness

Data collection ended in 2018; this release covers Phase I (2009-2013).

Geography

19 countries in Sub-Saharan Africa.