The WASDE Report on October 9th - Why did USDA's corn number rise?

Key takeaways

USDA's October 9 Crop Production report raised both crops. Corn yield rose 2.7 bu/ac to 181.2, putting production at about 16.0 billion bushels, the second largest on record. Soybean yield rose to a record 53.1 bu/ac, for 4.56 billion bushels. Harvested area was unchanged for both crops.

Soybeans: close on every line. Our estimate as of October 5 was 52.53 bu/ac on 85.81 million harvested acres, for 4.51 billion bushels. That is 0.57 bu/ac (1.1%) below the official yield, 0.08% below on area and 1.1% below on production.

Corn: the official number moved up, ours did not. We held at 178.31 bu/ac, essentially where we were on September 7 at 178.56. USDA went from 178.5 to 181.2. The yield gap is now 2.9 bu/ac, or 1.6%.

Acreage is still the bigger corn disagreement. We carry 86.27 million harvested acres against the official 88.51 million. Together with the yield gap, our production estimate of 15.38 billion bushels sits about 4% below USDA, and just under the low end of USDA's own historical confidence range of 15.5 to 16.6 billion bushels.

What changed in USDA's numbers this month?

NASS attributed the increases to three inputs: farmer-reported yields from the early harvest, objective yield measurements as more samples reached the lab, and satellite-modeled yields. Record corn yields are now forecast in Arkansas, Iowa, Kentucky, Mississippi, Ohio and Tennessee. Soybean gains were concentrated in the eastern Corn Belt, while western states came in lower on early-season dryness.

The corn number also came in above most pre-report trade estimates, so this was a surprise to the market as well as to us.

One detail is worth pausing on. NASS explicitly cited satellite modeling as supporting evidence for this month's revisions, in both corn and soybeans. Satellite observation is no longer an alternative to the official process. It is part of it. The open question is cadence: a monthly report against a crop that is observed continuously.

Why did USDA's corn number rise while ours held?

In September we wrote that kernel weight was the last variable still in motion, and that it would decide the final number. October suggests the crop filled better than its canopy looked.

Our model reads the plant through its canopy. Late in a hot season, canopies senesce early, and a model can read that as continued stress even when most of the grain has already been filled. The farmer reports coming off the combine say yields are better than expected, especially in the east. We think our late-season signal leaned too heavily on that canopy decline, and we are testing that now.

There is a second, more structural factor. Four of the six states with record corn yields this month, Arkansas, Kentucky, Mississippi and Tennessee, sit outside the 12 states we break out in our live report. Early harvest in the south is where farmers had actual yields to report first, and part of the national lift came from there.

What about the calls we put on the record in September?

We said we would publish whichever way these resolved, so here they are.

Indiana. In September we read 214.9 bu/ac against USDA's 200 and said harvest would settle it. Our own reading has since come down to 201.1, essentially to USDA's September figure. Our eastern reading was running hot, and the model corrected as the crop matured. That one goes to USDA.

Iowa. We sat 6 bu/ac below USDA's 219 in September. USDA now forecasts a record Iowa crop, so that gap has widened. That one also goes to USDA, at least for now.

Ohio. We read Ohio above the official figure in September. USDA now forecasts a record Ohio crop, a move in our direction.

What do the satellites show about harvest pace?

As of October 4, corn harvest was 23% complete, 4 points behind the five-year average, with Iowa 13 points behind. Soybean harvest was 25% complete, 8 points behind. NASS pointed to a wet September in the western Corn Belt.

Our phenology read on October 7 adds a nuance. In corn, almost every tracked state is fully mature, with only Ohio and Michigan still showing meaningful grain fill. The corn delay is about field access, not crop readiness. Soybeans are different: Missouri and Kansas still show more than half the crop in grain fill, and Iowa and Ohio about a third. In parts of the belt, the soybean crop is still finishing.

What is still unresolved?

Corn acreage. The 2.2 million acre gap has held since August. USDA's area is anchored on FSA certified acres and did not change this month. Ours is a harvested-for-grain measurement from orbit. Resolving which is closer is now our main open question for the season.

The final yield. USDA had 82% of its corn objective yield samples processed this month, against 90% a year ago, and harvest is running late. October figures have historically moved before the final January report. We will publish the comparison again when they do.

Our numbers stay on the record, weekly and state by state, in the SatYield Live Report. For the background on this season, see our September analysis and our August comparison.

Frequently asked questions

What did the October 2026 USDA report say about U.S. corn and soybeans?

USDA raised corn yield to 181.2 bu/ac from 178.5 in September, with production near 16.0 billion bushels. Soybean yield rose to a record 53.1 bu/ac, with production at 4.56 billion bushels. Harvested acreage was unchanged for both crops.

How close was SatYield's satellite estimate?

On soybeans, our October 5 estimate of 52.53 bu/ac and 4.51 billion bushels was within about 1.1% of the official figures, and within 0.1% on harvested area. On corn, our 178.31 bu/ac was 1.6% below the official yield, and our production of 15.38 billion bushels was about 4% below, mainly because we carry 2.2 million fewer harvested acres.

Does USDA use satellite data?

Yes. In its October briefing, NASS cited satellite-modeled yields alongside farmer surveys and objective yield field measurements as supporting this month's corn and soybean revisions.

Why is the 2026 harvest running behind?

A wet September, especially in the western Corn Belt, kept farmers out of fields. Satellite phenology shows most corn is mature and waiting on field access, while some soybean acres in states like Missouri and Kansas are still filling grain.

See how SatYield tracks the U.S. crop week by week, state by state. Explore the SatYield AI Agent.

A closing note in my own voice. September was a good month for us and October was a humbler one, and both belong on the record. Soybeans came in where we said they would. Corn did not, and Indiana and Iowa both went USDA's way. What I take from it is specific: the acreage question is still open for us an need further work.