What question did this study set out to answer?

The study aims to evaluate the effectiveness of remote sensing techniques in estimating rice yields in Niamey, Niger.

February 19, 2026

Assessment of the Potential of Medium‐Resolution Satellite Images for Estimating Rice Yields in Semiarid Irrigation Systems in Niamey

Puntos clave

The study aims to evaluate the effectiveness of remote sensing techniques in estimating rice yields in Niamey, Niger.
Utilized Landsat imagery to derive the normalized difference vegetation index (NDVI).
Mapped rice fields using a support vector machine classifier.
Developed exponential regression models to correlate NDVI with yield statistics.
Assessed accuracy and reliability of yield and area estimations.
NDVI showed a strong relationship with rice yield, with R2 values exceeding 0.91.
Estimated yields ranged from 5.61 to 6.15 t/ha, closely matching reported values.
Achieved accuracies of 94.05% to 94.87% in yield estimation.
Area estimation reliability was moderate with R2 = 0.58, while production had strong predictive power with R2 = 0.93.

Resumen

ABSTRACT Rice underpins global food security; however, in semiarid regions such as Niger, production is limited by climate variability, arable land and urban expansion, making reliable yield estimates essential for guiding agricultural policies and enhancing food security. This study aims to evaluate the effectiveness of remote sensing in estimating rice yields in the Niamey region during four dry seasons (2022–2025). Landsat imagery was used to derive the normalised difference vegetation index (NDVI), which was correlated with yield statistics from ‘Office National des Aménagements Hydro‐Agricole’. Rice fields were mapped using a support vector machine classifier, and exponential regression models were developed. The results show a good NDVI‐yield relationship, with R 2 exceeding 0.91 across all seasons. Estimated yields ranged from 5.61 to 6.15 t/ha, closely matching reported values of 5.32–5.85 t/ha and achieving accuracies of 94.05%–94.87%. However, cultivated areas were underestimated by 12–40 ha due to mixed pixels, fragmented boundaries and heterogeneous plot configurations. Area estimation showed moderate reliability ( R 2 = 0.58); production performed strongly ( R 2 = 0.93), whereas yield ( R 2 = −0.90) exhibited poor explicative power due to flooding impacts and management variability. Medium‐resolution satellite imagery enables rapid, cost‐effective rice monitoring, irrigation management support, food‐security planning and land‐use policy development across the Sahel.

Me gusta

Guardar