Remote sensingDirect CHEERSFull text reviewed

Satellite-based multi-annual yield models for major food crops at the household field level for nutrition and health research: A case study from the Nouna HDSS, Burkina Faso

Schwarz et al. · 2023 · International Journal of Applied Earth Observation and Geoinformation

DOI: 10.1016/j.jag.2023.103203 ↗

Open the published paper ↗People, places & projects ↓

01 / Study designRemote sensing

02 / ContributionEmpirical and methodological evidence

03 / Interpretation

Read the study design and limitations alongside its findings.

Read the limitations ↓

Summary

Develops satellite-based yield models at individual household-field level in Nouna HDSS.

Full text reviewed

The summary describes the published study and the sources listed below.

Scientific notes: 2026-11-11 · Research connections: 12 November 2026

Key findings

Multi-year models achieved R² 0.30–0.62 and normalized RMSE about 12–16%, with less apparent overfitting than single-year models.

Study design & setting

Study design and methods

Population, setting or evidence base

Nouna HDSS, Burkina Faso. Crop-specific satellite yield-model development. 411 yield measurements and 1,027 field boundaries; five crops over 2018,2020 and 2021.

Methods

LASSO regression combines Sentinel-2 monthly vegetation indices and accumulated CHIRPS rainfall; repeated five-fold cross-validation selects model tuning; FAO statistics provide a plausibility comparison.

Limits to interpretation

Limitations

Small crop-specific datasets, uncertain ground measurements and vegetation signals, and only three sampled years restrict transferability. National statistical agreement is not independent field-level validation.

Research relevance

Enables linking household-field food production estimates with household health and nutrition research.

People, places & projects