Pulse Brain · Growing Health Evidence Index
Tier 4 — Narrative / commentaryPeer-reviewed

Spatial cross-validation is not the right way to evaluate map accuracy

Alexandre M.J.‐C. Wadoux, G.B.M. Heuvelink, Sytze de Bruin, D.J. Brus

Ecological Modelling · 2021

Read source ↗ All evidence

Summary

This paper argues that spatial cross-validation is an inappropriate method for evaluating map accuracy in ecological modelling applications. The authors, affiliated with prominent spatial statistics research groups, contend that the technique introduces systematic biases in accuracy assessment and propose considerations for selecting more suitable validation strategies. The work contributes to methodological best practice in spatial model evaluation relevant to environmental and agricultural mapping.

UK applicability

UK soil and land use mapping programmes—including those supporting precision agriculture and natural capital assessment—rely on spatial models that require rigorous validation. This paper's methodological critique is directly applicable to evaluating accuracy of UK-based soil maps, habitat distribution models, and nutrient status assessments used in farm advisory services.

Key measures

Map accuracy assessment methods, spatial cross-validation performance, model validation approaches

Outcomes reported

The study critiques spatial cross-validation as a validation approach for map accuracy assessment, examining methodological limitations in how ecological and spatial models are evaluated.

Theme
Measurement & metrics
Subject
Measurement methods & nutrient profiling
Study type
Commentary
Study design
Commentary
Source type
Peer-reviewed study
Status
Published
Geography
International
System type
Other
DOI
10.1016/j.ecolmodel.2021.109692
Catalogue ID
SNmov5j4tp-gpy9tk

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.