CropProphet forecasts for the 2017 U.S. corn crop began on May 7 following a period of beta testing, and the forecasts are now being updated daily on the CropProphet website. The CropProphet models ingest the latest weather, satellite, and crop health information each day to make forecasts of yield and production for spatial scales ranging from counties to the national average. Upgrades for 2017 include a new high-quality temperature dataset and a new version of NASA’s MODIS satellite data, leading to improved yield and production forecasts; these changes will be discussed in more detail in future commentary.

The U.S. corn yield forecast has remained nearly steady so far this month (see figure below). The most likely final yield is currently 169.6 bushels per acre (bpa), which is very close to the long-term upward trend line as estimated by CropProphet and just 1 bpa below the May 10 USDA estimate. It is normal for the USDA estimate to be higher than the CropProphet forecast at this time of year, because the USDA methodology assumes that June will not be very dry (Westcott and Jewison 2013). The CropProphet models do not make any assumptions about future weather, and therefore the early-season CropProphet forecast allows for the possibility that damaging drought could emerge.


The large uncertainty in yield forecasts made at this time of year is reflected by the wide confidence intervals shown in the year-to-date yield forecast chart (see figure above). According to historical tests of CropProphet performance, the U.S. corn yield is 50% likely to be between 164 and 175 bpa, and it is 90% likely to fall in the range 138-185 bpa. It is useful to note that there is a higher chance of a large yield shortfall than of a large yield surplus, because the effects of weather are non-linear and asymmetric; in other words, very bad weather has a greater yield impact (in bushels per acre) than very good weather.

At this early stage in the season, the most significant predictor of the overall size of the U.S. corn crop at harvest is simply the estimated planted acreage. According to the March 31 USDA prospective plantings report, producers intended to plant 90.0 million acres of corn, which is less than last year (93.8 million acres) and less than the 5-year average (92.9 million acres); the figure below shows the expected acreage change compared to 2016. Owing to the modest acreage deficit, CropProphet expects U.S. production to be about 200 million bushels (1.5%) below trend.

Looking ahead to the beginning of summer, a relatively cool and moist weather pattern is expected to prevail in the next 4 weeks, and this would be favorable for corn prospects. CropProphet’s weather outlook component indicates that corn yields are likely to increase in many areas, and production gains are expected from Nebraska to the Great Lakes (see figure below). The forward-looking weather outlook component of CropProphet is one of its most valuable features, as CropProphet users are able to anticipate and objectively quantify the likely crop-specific impacts of weather before it happens.

The next update to the corn commentary will occur on June 6.



Westcott and Jewison, 2013: Weather Effects on Expected Corn and Soybean Yields, USDA-ERS, Feed Outlook No. (FDS-13G-01), July 2013.