Which weather model is most accurate in your city?

The Spread — the forecast spread — archives what every major model predicted, then scores those forecasts against what actually happened. No vibes, no marketing: bias and typical miss, by city and by forecast lead, on a rolling window. Pick a city:

• Atlanta, Georgia ECMWF IFS leads at 1.59°F (1-day)
• Boston, Massachusetts ECMWF IFS leads at 1.62°F (1-day)
• Buffalo, New York DWD ICON leads at 1.44°F (1-day)
• Charlotte, North Carolina ECMWF AIFS (AI) leads at 1.26°F (1-day)
• Chicago, Illinois ECMWF AIFS (AI) leads at 1.67°F (1-day)
• Dallas, Texas DWD ICON leads at 1.39°F (1-day)
• Denver, Colorado ECMWF IFS leads at 1.85°F (1-day)
• Houston, Texas NOAA AIGFS (AI) leads at 1.51°F (1-day)
• Los Angeles, California ECMWF IFS leads at 1.95°F (1-day)
• Miami, Florida DWD ICON leads at 1.27°F (1-day)
• Minneapolis, Minnesota ECMWF IFS leads at 2.13°F (1-day)
• New Orleans, Louisiana NOAA AIGFS (AI) leads at 1.49°F (1-day)
• New York, New York GEM (Canada) leads at 1.6°F (1-day)
• Oklahoma City, Oklahoma
• Orlando, Florida ECMWF AIFS (AI) leads at 1.2°F (1-day)
• Phoenix, Arizona DWD ICON leads at 1.35°F (1-day)
• Salt Lake City, Utah ECMWF IFS leads at 1.47°F (1-day)
• San Francisco, California ECMWF IFS leads at 1.8°F (1-day)
• Seattle, Washington UKMO (UK) leads at 1.37°F (1-day)
• Washington, District of Columbia ECMWF AIFS (AI) leads at 1.33°F (1-day)

How the scoring works

Each model's 1-, 3-, and 7-day-ahead forecasts are captured from its actual archived runs (nothing is reconstructed after the fact), then verified against ERA5 reanalysis once the day has happened. Bias is the signed average error — positive means the model ran warm. The typical miss (MAE) is the average absolute error. Cities join these pages as they join The Spread's archive.

The Spread itself is in closed beta — these verification pages are public while the rest of the product is gated.