A national, remote-sensing-based model for predicting field-scale crop yield
Randall Donohue and colleagues have now published the C-Crop model that Digiscape is using in the GrainCast project to forecast grain yields at multiple scales.
Existing agricultural grain yield models predict yield at the field scale, or at regional scales (like districts and countries), but not both with consistent accuracy. C-Crop is a scalable, satellite-based crop yield model. It is calibrated locally and so has field-scale accuracy.
The full paper in Field Crops Research describes C-Crop with its calibration and validation: you can find it here.