|
Grand Rapids, MI | From what I have seen in the field, and worked with some clients on, I think Carl Hobbs is on the $$. I use yield maps, elevations, history of the field (former hedgelines, field divisions, etc.) imagery (if availble), soil type maps, and producer knowledge.
One thing I will note...when looking at yield maps, I would strongly discourage you from averaging years together. There is a place for this, but IMHO it is not when making management zones. You WANT to see the diversity in the field. If you have a very dry year or a very wet year (or ideally both) they will show you changes in soil types often much better than soil type maps. Use these differences to draw your zones.
I am not in agreement, however, with dloc's ideas of one sample per grid. I would personally prefer to see multiple cores from throughout the region. IMHO there is just too much possibility to get a non-representative single spot in that zone. I also think interpolation between these management zones is a very bad idea. Now if you have a 20 acre management zone, which is pretty homogeneous from year to year and want to break it down into smaller grids, I can see interpolation. | |
|
Grid vs Zone sampling
