Landsat Historic Yield Pattern Assessment of Field and Yield Prediction Mapping
Historic Assessment of Field Slideshow
The simplest way to use Landsat to assess historic yield patterns of a field is to view imagery with the field boundary overlain for many seasons (the amount of correct imagery will vary per field; the field below is in a Landsat overlap zone so there is more imagery than most locations); when assessing data, ignore pixels near the field boundary that average values. For corn, different indices can be used (MSAVI has been shown to be be better than others) but NDVI or NNIR based on reflectance or digital number will both work well in showing relative patterns of predicted yield. For corn, imagery should be from V12 to just prior to the onset of tassel which can be determined based on viewing imagery. For soybeans and other crops, NIR solely should be used; imagery should be from R2 to R6. Depending on the crop, Landsat needs to be applied in particular ways. In the graphics below, lighter shades of gray are higher values and correlate to higher yield; the field shows overall consistent spatial patterns of higher and lower areas of predicted yield. Images transition every 4 seconds (or click arrows or dots).