Abstract
Rice yield prediction at high spatial resolution is essential to support precision management and sustainable intensification in irrigated systems. While many remote sensing studies provide yield estimates at the field scale, pixel-level predictions are required to characterize within-field variability. This study assesses the potential of multitemporal Sentinel-2 imagery and machine learning to estimate rice yield at pixel level in the Albufera rice area (Valencia, Spain). Yield data from combine harvester maps were collected for ‘JSendra’ and ‘Bomba’ Japonica varieties over five growing seasons (2020–2024) and linked to 10 m Sentinel-2 bands in the visible, near-infrared (NIR) and short-wave infrared (SWIR) regions. Random Forest (RF) and XGBoost (XGB) models were trained with 2020–2023 data and independently validated in 2024. XGB systematically outperformed RF, achieving at 110 and 130 DAS (days after showing), R2 values of 0.74 and 0.85 and RMSE values of 0.63 and 0.28 t·ha−1 for ‘JSendra’ and ‘Bomba’. Prediction accuracy increased as the season progressed, and models using all spectral bands clearly outperformed configurations based only on spectral indices, confirming the dominant contribution of NIR reflectance. Spatial error analysis revealed errors at field edges and headlands, while central pixels were more accurately predicted. Overall, the proposed approach provides accurate, spatially explicit rice yield maps that capture within-field variability and support both end-of-season yield estimation and early season forecasting, enabling the identification of potentially low-yield zones to support targeted management decisions.
Abstract from the paper’s OpenAlex record; open copy licensed cc-by. Read at the source ↗
The text the model read, with the source record’s inline markup (italics, sub- and superscripts) rendered and its character entities decoded; quote offsets refer to the raw text.
Claims 11
Claim · model-written summary
XGBoost outperformed Random Forest for rice yield prediction.
- Taxon
- Rice
Oryza sativat0 - Outcome
- outperformede0
- Direction
- outperformede0
- Comparator
- RFe0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“XGB systematically outperformed RF”
t0“Rice Yield Prediction”
Claim · model-written summary
For JSendra, XGBoost achieved R2 of 0.74 at 110 DAS.
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“XGB systematically outperformed RF, achieving at 110 and 130 DAS (days after showing), R2 values of 0.74 and 0.85 and RMSE values of 0.63 and 0.28 t·ha−1 for ‘JSendra’ and ‘Bomba’”
t0“Rice Yield Prediction”
Claim · model-written summary
For Bomba, XGBoost achieved R2 of 0.85 at 130 DAS.
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“XGB systematically outperformed RF, achieving at 110 and 130 DAS (days after showing), R2 values of 0.74 and 0.85 and RMSE values of 0.63 and 0.28 t·ha−1 for ‘JSendra’ and ‘Bomba’”
t0“Rice Yield Prediction”
Claim · model-written summary
For JSendra, XGBoost RMSE was 0.63 t·ha−1.
- Taxon
- RiceOryza sativat0
- Material
- JSendrae0
- Outcome
- RMSEe0
- Value
- 0.63e0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“RMSE values of 0.63 and 0.28 t·ha−1 for ‘JSendra’ and ‘Bomba’”
t0“Rice Yield Prediction”
Claim · model-written summary
For Bomba, XGBoost RMSE was 0.28 t·ha−1.
- Taxon
- RiceOryza sativat0
- Material
- Bombae0
- Outcome
- RMSEe0
- Value
- 0.28 t·ha−1e0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“RMSE values of 0.63 and 0.28 t·ha−1 for ‘JSendra’ and ‘Bomba’”
t0“Rice Yield Prediction”
Claim · model-written summary
Prediction accuracy increased as the season progressed.
- Taxon
- RiceOryza sativat0
- Outcome
- Prediction accuracye0
- Direction
- increasede0
- Comparator
- the season progressede0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“Prediction accuracy increased as the season progressed”
t0“Rice Yield Prediction”
Claim · model-written summary
Models using all spectral bands outperformed spectral-index configurations.
- Taxon
- RiceOryza sativat0
- Outcome
- outperformede0
- Direction
- outperformede0
- Comparator
- configurations based only on spectral indicese0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“models using all spectral bands clearly outperformed configurations based only on spectral indices”
t0“Rice Yield Prediction”
Claim · model-written summary
NIR reflectance made the dominant contribution to model prediction.
- Taxon
- RiceOryza sativat0
- Outcome
- dominant contribution of NIR reflectancee0
- Value
- NIR reflectancee0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“confirming the dominant contribution of NIR reflectance”
t0“Rice Yield Prediction”
Claim · model-written summary
Prediction errors occurred at field edges and headlands.
- Taxon
- RiceOryza sativat0
- Outcome
- errors at field edges and headlandse0
- Value
- field edges and headlandse0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“Spatial error analysis revealed errors at field edges and headlands”
t0“Rice Yield Prediction”
Claim · model-written summary
Central pixels were predicted more accurately.
- Taxon
- RiceOryza sativat0
- Outcome
- accurately predictede0
- Direction
- more accuratelye0
- Comparator
- central pixelse0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“central pixels were more accurately predicted”
t0“Rice Yield Prediction”
Claim · model-written summary
The approach generated spatial rice yield maps capturing within-field variability.
- Taxon
- RiceOryza sativat0
- Outcome
- spatially explicit rice yield mapse0
- Value
- within-field variabilitye0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“the proposed approach provides accurate, spatially explicit rice yield maps that capture within-field variability”
t0“Rice Yield Prediction”