Winter is no time to rest for the precision farmer.
It’s time for the year-end analysis of yield data, which needs to be done every year.
The first step for any yield analysis is getting the data from the yield monitor. I am surprised by the number of farmers who have told me they still have four or more years of yield data on a card.
“Could you help me look at my yield maps?” they ask.
I will admit that recent technological advances have made it more difficult to access data from the combine, but that is not an excuse for leaving valuable yield data in the field.
There are many places where farmers can get help accessing their data. The yield monitor manufacturer, the equipment manufacturer, the local co-operative, a local consultant and the local seed salesperson all have resources to read and process the data coming from the combine.
The second step is cleaning the yield data, which is one of the forgotten elements of precision analysis.
Everybody has seen those red points at the edges of field that represent low yields. Do we really have lower yields around the field edges? It’s unlikely.
We have also seen those dark green points representing extremely high yields scattered randomly through the field.
Did you really get 450 bushels per acre of wheat in that one spot?
Or maybe you’ve seen that one crop yield point across the road in the neighbour’s wheat field?
All of these are candidates for cleanup.
The most common method to clean up yield data is to pick a low yield value and delete all the points below that. Then pick a high yield value that is unreasonable and delete all of the yield points above that.
This can be valuable, but there are better techniques.
Those low yields points on the ends of rows may have something to do with the lag time setting, which is the time it takes between when the grain is cut and when it hits the mass flow sensor.
Without this lag time setting, the calculated yield value would be 60 feet down the field from where it was actually harvested.
Low yield points at the edge of the field can also be caused by an incorrect header height calibration.
The flow of data from the mass flow sensor should stop when the header is lifted at the end of the row.
However, before that happens, the full flow of grain hitting the mass flow sensor may have already stopped. Leftover kernels continue hitting the pressure plate, which calculates a yield of half a bushel per acre.
As for the ridiculously high values, they often happen when the combine stops in the field. A lot of grain is hitting the mass flow sensor and it is all associated with that one point.
The next several points are often suspiciously low before a new batch of grain hits the mass flow sensor.
Incorrect swaths are another major cause of errors.
If the last swath in a field is actually a half swath and the operator does not change the swath setting, the yield monitor system measures the grain and calculates the bushels per acre based on a full swath. This results in half the yield per acre of what was actually in the field.
Farmers have a variety of options for solving these problems, but many of them need to be done during harvest, such as checking lag time setting and calibrations and entering correct swath widths.
For now, we can look at the data included in the yield file and delete individual yield points if they are determined to be inaccurate.
Combine speed is one of the most valuable attributes to use for cleanup. Look in the yield file for extremely low combine speeds or rapid speed changes.
Yield values are most inaccurate in these two situations.
The good news is that producers do not have to clean up yield points by themselves.
The U.S. Department of Agriculture has released Yield Editor software that is highly recommended for cleaning up yield data. It is available for a free download.