Pulse Brain · Growing Health Evidence Index
Tier 3 — Observational / field trialPeer-reviewed

Performance of linear mixed models and random forests for spatial prediction of soil pH

Mirriam Makungwe, Lydia M. Chabala, Benson H. Chishala, R. M. Lark

Geoderma · 2021

Read source ↗ All evidence

Summary

This paper evaluates two contrasting statistical approaches—linear mixed models and random forests—for predicting soil pH spatially. The work, conducted in Zambia, assesses the relative strengths of traditional geostatistical methods versus machine learning in capturing soil spatial variability. As suggested by the title and journal scope, the study contributes methodological insight into soil characterisation for agricultural and environmental management.

UK applicability

Spatial soil pH prediction methods are relevant to UK soil survey and precision agriculture applications. However, transferability of findings depends on whether climatic, geological, and soil type differences between Zambia and UK contexts affect model performance; direct applicability would require local validation.

Key measures

Soil pH predictions; model accuracy metrics (as suggested by predictive modelling comparison); spatial prediction performance

Outcomes reported

The study compared the performance of linear mixed models and random forests as methods for spatial prediction of soil pH across a study region. The research evaluated predictive accuracy, spatial patterns, and model utility for soil characterisation.

Theme
Measurement & metrics
Subject
Soil health assessment & monitoring
Study type
Research
Study design
Field trial
Source type
Peer-reviewed study
Status
Published
Geography
Zambia
System type
Other
DOI
10.1016/j.geoderma.2021.115079
Catalogue ID
SNmov5j98g-lyxqpt

Topic tags

Pulse AI · ask about this record

Dig deeper with Pulse AI.

Pulse AI has read the whole catalogue. Ask about this record, its theme, or how the findings apply to UK farming and policy — every answer cites the underlying studies.