US Winter Wheat: See The Yield Before USDA Publishes A Number
Treefera’s US winter wheat yield forecast is a weekly, county-level prediction across HRW, SRW and white winter wheat, built from satellite imagery and weather reanalysis. It runs from October planting through August harvest - and for seven of those months USDA publishes no winter wheat yield estimate at all.
Five weeks before USDA's first revision. Seven months before USDA's first estimate.
In April 2022, USDA had published no winter wheat yield estimate. La Niña drought persisted across the Southern Plains. Treefera's model, running since October, began correcting sharply: −0.66 bu/ac in a single week. That correction began five weeks before the May WASDE published its first winter wheat figure. Treefera finished within 1 bu/ac of the NASS actual, with verified uncertainty.
In 2023, Treefera's average error across all 45 vintages was 1.47 bu/ac. WASDE's was 5.76, nearly four times less accurate. The edge does not shrink as USDA incorporates more data. At grain fill in June, Treefera's MAE is 2.20 against WASDE's 2.67. At harvest in August, 2.17 against 2.72. The advantage holds through the season.
18%
more accurate than USDA WASDE in the in-season overlap window, averaged across five strictly out-of-sample years
7
months before USDA publishes its first winter wheat yield estimate. Treefera is not just earlier, it is the only source
5
weeks ahead of USDA in the 2022 La Niña drought. The model was already correcting in the right direction before the market had any official number
USDA's first winter wheat yield estimate lands in May, and its error holds near 2.7 bu/ac through to harvest. Treefera's error falls to 2.13 bu/ac over the same window.
USDA publishes no winter wheat yield estimate between October planting and the May WASDE. Across a 5-year hindcast (2019-2023), both Treefera and WASDE were scored against NASS final actuals at each crop stage. At heading in May, Treefera's error was 2.42 bu/ac against WASDE's 2.71. At grain fill in June - the window that sets kernel weight - it was 2.20 against 2.67. At maturity in July, 2.13 against 2.72. At harvest in August, with USDA holding months of survey data, 2.17 against 2.72. Across the full overlap window, the means are 2.22 bu/ac and 2.72 bu/ac.
The 5-year average hides real spread. In 2023, Treefera's mean error across all 45 weekly vintages was 1.47 bu/ac against WASDE's 5.76. WASDE was closer in 2020 and 2022, and both years are published rather than excluded. The same model specification runs across all 5 hindcast years with no year-specific tuning. WASDE never enters the model as an input. Only vintage-level scoring against NASS finals exposes that spread.
Trusted for billion-dollar supply chain and risk management decisions.
Four reasons Treefera delivers information advantage
The only source for seven months
From October planting through late April, USDA publishes no winter wheat yield estimate. Treefera's forecast runs throughout, 45 weekly vintages from October through August harvest. When the market has no official signal, Treefera clients do.
Validated across five years, same model
18% lower in-season MAE than WASDE, averaged across 2019-2023. In 2023, Treefera 1.47 vs WASDE 5.76. Same model specification across all five years, no tuning. WASDE is a scoring benchmark only, never a model input.
Independent of government reporting
Built on satellite imagery and weather reanalysis, not federal survey data. Available through government shutdowns and data blackout periods. When USDA has published nothing on winter wheat yield for seven months, Treefera has been running weekly.
Phenologically weighted precision
Stress during flowering in May registers very differently to the same stress in March. Treefera's model weights every weather input by crop stage. A standalone weather correction signal isolates in-season weather impact from the trend baseline.
Insights
Recent News
FAQ
The May WASDE, roughly seven months after October planting - and that first figure is trend-based, with no field survey data in it. USDA then revises monthly through to the January WASDE, which is the final number the market settles on. From October through late April there is no official US winter wheat yield estimate of any kind. Treefera’s forecast runs weekly throughout that window.
Clear Data. Clear Decisions.
Walk through the model, five years of backtested results and the current 2026 signal - and assess whether it fits your workflow.
Request a demo