About Yield Monitor Calibration and Yield Map Post-Calibration
It is challenging to keep a yield monitor properly calibrated throughout the harvest season, but it's important to have yield maps with accurate quantities if you base variable rate application on yield map values. (A clean yield map based on a monitor that has not been properly calibrated still has many uses though, as will be described.) Post-calibration can produce maps with yield quantities and variability based on your specifications. You should review yield quantities on a yield map, then make a decision whether to have amounts and/or variability adjusted to meet your specifications.
Challenging to have yield monitor properly calibrated throughout the season
No matter how well calibrated, impact-based yield monitors inherently cannot produce data that are the same as actual yield amounts on a point-by-point basis (Colvin and Arslan [1999], pg. 3-4). This is so because the mechanics of the yield monitor system smoothes data values. Smoothing and yield monitor data accuracy are described in Colvin and Arslan (1999). In their experiment, “corn kernels between 60 and 70 ft. from the edge of the field were painted blue” then were harvested; after the combined reached the blue kernels “it took 20 feet before blue kernels were measured, 50 feet to reach a peak, and 100 feet to get the majority through the machine”. Over an entire field, however, yield errors from a properly calibrated monitor tend to average themselves out and the field average of yield monitor data can become close to average actual yield.
Calibrating an impact-based yield monitor system is a multiple-step process. In the majority of systems, calibrating the weight is the most time-consuming part. Multiple loads may need to be weighed (the necessary amount of loads can differ depending on the type of system) on a certified scale (e.g. weigh wagon or commercial scale) for each crop many times during the season as the yield monitor can become more and more out of calibration throughout the season (Watermeier, [2001]; Grisso et al. [2009], pg. 5). Calibration can be a challenge (Cowan, 2000, pg. 5) and getting loads weighed often can cause logistical problems (Casady et al., 1998, pg. 3) and can be very time-consuming; as a result, there are many yield monitor datasets that have not been calibrated properly. That being said, calibrated yield data has been shown to be important to apply to variable rate technology. Barker (2008) used calibrated yield monitor data to develop management zones for variable rate application (VRA) of phosphorus and potassium and saved $88.04 per acre compared to uniform application.
Clean yield maps from yield monitors that have not been properly calibrated still have uses
Whether a yield monitor is or is not properly calibrated, if cleaned, the resulting map will coherently show relatively higher and lower areas of yield, which can be useful. (Areas of higher and lower yield on a map from a properly calibrated monitor will, however, more closely represent actual differences of yield amounts.) One way of applying clean yield monitor data that are not properly calibrated is to use a map to define similar yielding zones by classifying with a clustering method such as natural breaks or manually defining boundaries and, if necessary, assigning yield values to the zones based on knowledge and experience (not based on the yield monitor data). Determining if zones of higher and lower yield are located in the same areas over many seasons is important in defining management zones (it is important to know if a field is spatially stable). Too much stop and go driving or a very inconsistent traveling speed hinders the usefulness of yield monitor data more than if it is not properly calibrated (to a reasonable extent). Another option of handling cleaned yield monitor data that are not properly calibrated is to have the yield map post-calibrated for yield average and variability as described and offered here. If you are basing inputs on crop removal, it is important to study the yield quantity and variability on a clean map to determine if post-calibration should occur.
References
Barker, J. 2008. Today's Higher Fertilizer Prices Show Even Greater Savings for Precision Agriculture. Ohio Ag Manager. The Ohio State University Extension. Cited at: http://ohioagmanager.osu.edu/uncategorized/todays-higher-fertilizer-prices-show-even-greater-savings-for-precision-agriculture/.
Cassady, W., Pfost, D., Ellis, C., and K. Shannon. 1998. Precision agriculture: yield monitors. Water Quality; WQ 451. Published by University Extension, University of Missouri – System.
Colvin, T.S., and S. Arslan. 1999. Yield monitor accuracy. Site-Specific Management Guidelines: SSMG-9.
Cowan, T. 2000. Precision agriculture and site-specific management: current status and emerging policy issues. CRS Report for Congress; received through the CRS web. Order Code RL30630.
Grisso, R., Alley, M., and P. McClellan. 2009. Precision farming tools: yield monitor. Virginia Cooperative Extension; 442-502.
Watermeier, N. 2001. Yield monitor calibration tips—making the most from your data. The Ohio State University Extension; ANR-8-01.