Prediction of a Wheat Crop Yield Map by Using Post-anthesis Radiometrical Data

Because yield maps provide only synthetic information on the whole crop cycle, it appears necessary to supplement them with more analytical tools enabling to explain the spatial variations in measured yield and to make diagnosis. Here we tested two candidate methods: simulation modelling and radiometric mapping of green leaf area index (GLAI) after anthesis, on a 15-ha wheat-cropped field with marked soil and topographic heterogeneity under a temperate climate. When aggregated over the major soil units within the field, GLAI maps and modelling could mimic their overall effects on yield. However, they both were poor predictor of short-range (10 m) yield variability, even when the model was re-initialized with the GLAI data measured after anthesis. Possible reasons include sampling and spatial joint biases, and that GLAI should be taken several times during grain filling.