Which weather model is most accurate in Atlanta?

Outlook Model accuracy

Atlanta, Georgia · daily high temperature in °C · trailing 30 days · refreshed at most every 6 hours

Every day, The Spread archives what each major weather model predicted for Atlanta at 1-, 3-, and 7-day leads, then scores those forecasts against the ATL observations once the day has happened. This page covers the trailing 30 days (2026-08-04 to 2026-09-02), for daily high temperature.

One day out, too close to call; at three days, too close to call; at seven days, NOAA GFS (1.84°C).

The largest standing bias belongs to GEM (Canada), which has run cool by about 2.64°C at 3-day lead over this window.

The Spread's own bias-corrected blend scored a 0.59°C typical miss at one day. Its corrections come from the same window it is scored on (in-sample), so read it as a ceiling, not a promise.

The scoreboard

Typical miss (MAE) in °C with the signed bias in parentheses — positive bias means the model ran warm. Days = scored days behind the 1-day number.

Model	1-day	3-day	7-day	Days
ECMWF IFS	1.0° (-0.8°)	1.9° (-0.7°)	2.3° (-1.1°)	30
NOAA GFS	1.2° (+1.2°)	1.5° (+0.9°)	1.8° (+1.2°)	30
NOAA AIGFS (AI)	1.2° (-1.1°)	1.8° (-1.8°)	2.1° (-1.9°)	30
UKMO (UK)	1.3° (-1.3°)	1.5° (-1.0°)	—	30
DWD ICON	1.4° (-1.2°)	1.5° (-0.8°)	—	30
GEM (Canada)	1.9° (-1.9°)	2.7° (-2.6°)	3.3° (-3.1°)	30
ECMWF AIFS (AI)	2.4° (-2.4°)	2.5° (-2.5°)	2.7° (-2.6°)	30
NBM (US blend)	1.6° (-1.5°)	1.8° (-1.7°)	2.3° (-2.1°)	30
Blended (bias-corrected)	0.6° (+0.0°)	1.0° (+0.0°)	1.1° (+0.0°)	30

Method, honestly

The Spread's Atlanta point is the Atlanta Hartsfield-Jackson station (ATL), 14 km from downtown, used for both forecasts and verifications. Forecasts come from each model's actual archived runs via Open-Meteo's Previous Runs API — what the model said at the time, not a reconstruction. Truth is ATL observations — the station's own daily highs, a thermometer rather than a grid, and no correction applies: the thermometer is the truth outright. Rankings cover the models themselves; the two consensus products at the bottom of the table — NBM, which is itself a blend of models, and The Spread's own bias-corrected blend — are shown for comparison, never ranked or named "most accurate" here. The blend's corrections come from the same trailing window it is scored on (in-sample), so treat its score as a ceiling. A rolling 30-day window means these numbers move — that is the point.

Receipts — was the forecast right?

One page per settled date: what every model said as the day approached, what happened, and who was closest.

Tue Sep 1 10 days out the models said 29–31; it hit 34. ECMWF led (off 0.1°C); UKMO ran 2°C cold.
Mon Aug 31 9 days out the models said 28–33; it hit 34. GFS led (off 0.1°C); GEM ran 3°C cold.
Sun Aug 30 8 days out the models said 27–36; it hit 33. GFS led (off 0.4°C); GEM ran 3°C cold.
Sat Aug 29 7 days out the models said 26–35; it hit 32. GFS led (off 0.4°C); GEM ran 5°C cold.
Fri Aug 28 6 days out the models said 22–33; it hit 34. GFS led (off 0.6°C); AIFS ran 3°C cold.
Thu Aug 27 5 days out the models said 24–33; it hit 32. AIGFS led (off 1.2°C); UKMO ran 4°C cold.
Wed Aug 26 4 days out the models said 30–34; it hit 33. AIGFS led (off 0.3°C); AIFS ran 2°C cold.
Tue Aug 25 3 days out the models said 28–34; it hit 34. ECMWF led (off 0.1°C); AIFS ran 3°C cold.
Wed Aug 19 13 days out the models said 29–37; it hit 35. UKMO led (off 0.3°C); GEM ran 3°C cold.
Tue Aug 18 13 days out the models said 30–35; it hit 34. GEM led (off 0.4°C); ECMWF ran 2°C hot.
Mon Aug 17 13 days out the models said 27–34; it hit 37. ICON led (off 0.1°C); GEM ran 2°C cold.
Sun Aug 16 13 days out the models said 24–33; it hit 36. AIGFS led (off 0.0°C); GEM ran 2°C cold.
Sat Aug 15 13 days out the models said 31–36; it hit 34. GFS led (off 0.2°C); UKMO ran 1°C hot.
Fri Aug 14 13 days out the models said 30–34; it hit 34. GEM led (off 0.2°C); GFS ran 2°C hot.
Thu Aug 13 13 days out the models said 30–39; it hit 33. ECMWF led (off 1.1°C); UKMO ran 2°C hot.
Wed Aug 12 13 days out the models said 28–33; it hit 32. ECMWF led (off 0.1°C); UKMO ran 2°C hot.
Tue Aug 11 13 days out the models said 26–40; it hit 31. ICON led (off 0.6°C); GFS ran 3°C hot.
Mon Aug 10 13 days out the models said 30–33; it hit 32. AIFS led (off 0.3°C); GFS ran 4°C hot.
Sun Aug 9 13 days out the models said 31–36; it hit 33. UKMO led (off 0.1°C); GEM ran 2°C cold.
Sat Aug 8 12 days out the models said 30–34; it hit 32. GEM led (off 0.2°C); UKMO ran 2°C hot.
Fri Aug 7 11 days out the models said 28–35; it hit 29. ICON led (off 0.2°C); GFS ran 4°C hot.
Thu Aug 6 10 days out the models said 31–38; it hit 30. AIGFS led (off 0.3°C); AIFS ran 2°C cold.
Wed Aug 5 9 days out the models said 30–36; it hit 30. UKMO led (off 0.2°C); ECMWF ran 3°C cold.
Tue Aug 4 8 days out the models said 28–35; it hit 30. AIGFS led (off 0.3°C); UKMO ran 3°C cold.
Mon Aug 3 7 days out the models said 26–32; it hit 28. AIFS led (off 0.3°C); ICON ran 2°C hot.
Sun Aug 2 6 days out the models said 25–33; it hit 31. GFS led (off 0.3°C); ECMWF ran 3°C cold.
Sat Aug 1 5 days out the models said 26–35; it hit 30. UKMO led (off 0.3°C); AIFS ran 3°C cold.
Fri Jul 31 4 days out the models said 30–34; it hit 32. ECMWF led (off 0.2°C); GEM ran 2°C cold.
Thu Jul 30 3 days out the models said 29–35; it hit 32. ECMWF led (off 0.2°C); GEM ran 3°C cold.

Other cities

Berlin, State of Berlin Boston, Massachusetts Buffalo, New York Charlotte, North Carolina Chicago, Illinois Dallas, Texas Denver, Colorado Detroit, Michigan Houston, Texas London, England Los Angeles, California Miami, Florida Milan, Lombardy Minneapolis, Minnesota New Orleans, Louisiana New York, New York Oklahoma City, Oklahoma Orlando, Florida Paris, Île-de-France Region Philadelphia, Pennsylvania Phoenix, Arizona Salt Lake City, Utah San Antonio, Texas San Diego, California San Francisco, California Seattle, Washington St. Louis, Missouri Tampa, Florida Washington, District of Columbia

The Spread — the forecast spread — is in closed beta. These verification pages are public while the rest is gated.