Yield estimates can be a tricky task—wandering through a cornfield in mid-July, unsure if you’re standing in the right spot to get an accurate representation of what harvest might bring. Then there’s the challenge of estimating soybean yield, which often feels even more uncertain. But what if I told you we can now get highly accurate yield forecasts for both corn and soybeans in the middle of the growing season—using imagery? Just imagine the power that kind of information gives you heading into harvest. 

myFS Agronomy’s Yield Forecast Tool, developed by IntelinAir, has been used over the past several years by growers, agronomists and ag retailers to track crop performance and guide in-season decisions. It combines a wide range of data layers—including weather data, planting and harvest information, in-season applications, crop performance alerts, aerial imagery, and historical field performance—to form a complete picture of how the crop is progressing.  

Every two weeks during the growing season, the tool analyzes these layers and generates yield estimates, confidence ranges, and trend direction for each enrolled field. Forecasting begins in early July and becomes increasingly refined, with the goal of delivering the most accurate prediction roughly eight weeks before harvest. On average, the tool achieves 94.3% accuracy for corn and 93.1% for soybeans, making it a valuable in-season decision-making resource. 

Over the past five seasons, prediction accuracy has steadily improved, with 2024 standing out as the most accurate year yet. For soybeans, field-level prediction accuracy averaged 93.1%, peaking at 94.3% in 2024 and dipping to 91.8% in 2020. Soybean models begin forecasting yield at about 475 growing degree days (GDD), with the most accurate predictions occurring between 2,200 and 2,700 GDD at 93.69%, and the least accurate between 740 and 1,135 GDD at 92.65%. 

Corn predictions have shown even stronger performance, with an average field-level accuracy of 94.3%. Accuracy peaked at 94.9% in 2024 and was lowest in 2020 at 93.7%. Like soybeans, corn yield predictions begin at 475 GDD, with the highest accuracy seen between 2,200 and 2,700 GDD at 96.1%. Accuracy is slightly lower in the early growth stages, between 475 and 1,000 GDD at 94.2%. 

In terms of yield range, the models perform best within mid-range yields. For corn, fields yielding 250–275 bu/ac had the highest accuracy at 98%, while fields in the 175–200 bu/ac range had the lowest at 90.9%. For soybeans, the 60–80 bu/ac range showed the strongest performance at 93.97% accuracy, while accuracy dropped slightly to 88.11% in the 40–60 bu/ac range. These trends highlight the value of advanced modeling and the key role that growth stage and yield range play in predictive accuracy for both crops. 

An example of this can be seen in a field enrolled during the 2024 growing season. On an August 27th flight, yield forecast estimated an average yield of 250 bu/ac. That field was harvested on November 1st with a final dry yield of 257 bu/ac—just a 3% margin of error. That kind of precision isn’t just impressive—it’s useful. 

Aerial Imagery – Provided by Haley Brokate, Illini FS

Yield estimates provide more than just a number. They help growers make smarter decisions by allowing them to plan logistics and storage ahead of harvest, prioritize scouting or input adjustments, strengthen conversations with landlords or lenders, and enhance marketing strategies with real-time field data. Yield prediction through aerial imagery isn’t just a neat tool—it’s becoming a reliable, in-season decision-support system that adds value long before the grain is in the bin.

*Information in this article was gained from my experience using myFS Agronomy and from Conner Schmidt with IntelinAir, Inc.