The objective of this study is to address the capabilities of multi-temporal optical images to estimate the fine-scale yield variability of wheat, over a study site located in southwestern France. The methodology is based on the Landsat-8 and Sentinel-2 satellite images acquired after the sowing and before the harvest of the crop throughout four successive agricultural seasons, the reflectance constituting the input variables of a statistical algorithm (random forest). The best performances are obtained when the Normalized Difference Vegetation Index (NDVI) is combined with the yield maps collected during the crop rotation, the agricultural season 2014 showing the lower level of performances with a coefficient of determination (R2
) of 0.44 and a root mean square error (RMSE) of 8.13 quintals by hectare (q.h−1
) (corresponding to a relative error of 12.9%), the three other years being associated with values of R2
close or upper to 0.60 and RMSE lower than 7 q.h−1
(corresponding to a relative error inferior to 11.3%). Moreover, the proposed approach allows estimating the crop yield throughout the agricultural season, by using the successive images acquired from the sowing to the harvest. In such cases, early and accurate yield estimates are obtained three months before the end of the crop cycle. At this phenological stage, only a slight decrease in performance is observed compared to the statistic obtained just before the harvest.
This is an open access article distributed under the Creative Commons Attribution License
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited