A simple approach using DataFarming NDVI. Three maps of the same paddock from early June, July and August, using three management zones and a 100 m scale to keep things simple. I used the 5 July image for variable rate urea, with the aim of supporting yield potential rather than evening up the paddock (note: the evening up approach can be taken if your confidence P, S, pH & sodicity are not highly variable). The supporting yield approach meant that blue areas received the highest nitrogen rate and red areas the lowest. Before making the prescription, I did the basics, checked that weeds and soil type were not confounding the NDVI signal, then measured tiller numbers in each of the three target zones as a proxy for likely head density. Combined with assumptions around grains per head and grain weight, this helped estimate yield potential and nitrogen demand. Paddock was sown to Rio wheat on 5 May. Great day to sow wheat 😉.