Understanding Yield Variations
If you have ever uploaded the same data set into more than one digital tool, you may have noticed variations in the yield figures that are displayed. It is important to understand that variation in yield between different software programs or your monitor does not necessarily mean that the data is wrong or that the software is broken.
If the difference in yield displayed between digital tools is less than 10%, this variation can likely be explained by a few differences in how digital tools process data.
- Data cleaning rules: Each digital tool has its own unique way of cleaning data as it is imported. Through this process, some tools clean or remove more data points than others.
- Area used in the yield calculation: Yield is typically displayed in bushels per acre or kilograms per hectare. To determine the field size, some digital tools use the field boundary area, while others use the area covered by the combine during harvest. In most cases, there is very little difference between these areas, but where there is a large discrepancy, this could explain yield variation. For example, Granular Insights uses the geographic area that was harvested in the yield calculation.
- Treatment of data points: Data imported into a digital tool is typically aggregated to make it easier to visualize and analyze. The degree to which this data is aggregated varies between tools and is often driven by the features the tools was designed to support. For example, Granular Insights handles data as a raster (a collection of points) to help quickly display different map layers. On the other hand, AgStudio treats data as a series of points, allowing for maximum flexibility when using the data for different agronomic workflows.
These three differences are the most common drivers in yield variation between digital tools. However, it is important to note that there may be other differences unique to certain software and monitors.
Our team is working hard to ensure your data is displayed accurately and consistently, so that it is most useful to you. If you see differences in yield greater than 10% or your data doesn't look right for any reason, we are here to help.