(Open Management Zones folder in Store to access more background information.)
Landsat Yield Quantity Solely Zones
Related Page: Yield Quantity & Variability Zones
Landsat can correlate high enough to yield to be used to develop management zones for different purposes. For any crop, careful attention needs to applied to select imagery that correlates well enough to future yield patterns. Of course there can be events that occur after the time of imagery, such as hail damage in a higher area of crop condition, that can lower the correlation to yield.
The example below is of a field that, although it does show noticeable variability, is spatially stable enough (wet and dry season yield patterns are highly correlated) to be a candidate for management zones strictly based on yield patterns (variability does not need to be included as it is in Yield Quantity & Variability Zones). If a field shows more variability than the example below, Productivity Zones should be used (and are usually the zones that should be used). It is important to only use pixels from within the field boundary and extend those valid values to the field boundary (pixels that average in surfaces from outside the field should not be used). A management zone map is shown at the bottom of the page. (All maps are symbolized from higher crop condition [darker green] to lower crop condition [lighter green] from ± 3 standard deviations from the mean).
2011 Soybeans
2009 Soybeans
2007 Soybeans
2006 Corn
Management Zones
The histogram and zones shown below are based on natural breaks classification. Natural breaks is a clustering classification method designed to best group similar values while maximizing differences between the groups. A clustering classification, such as natural breaks or fuzzy classification, can be used to group data or manual classification based on the histogram can be used. Zones should ideally each include a main modal area, or an apparent divisions of a main modal area, of the frequency histogram (which natural breaks did well overall in this case). Groupings commonly correlate to the amounts of major topographic areas (but do not necessarily need to correlate to topographic areas as there are other fixed or static factors that can affect yield patterns).
The map at the below left is based on an average of normalized equally weighted maps from the four seasons (maps are equally weighted yet retain their own unique distribution). The map in the middle was produced after classifying with the four natural breaks represented in the above histogram. The map on the right is based on the map in the middle, but is more coherent because relatively small areas have been dissolved into larger surrounding areas. 