Which weather model is most accurate in Phoenix?

Outlook Model accuracy

Phoenix, Arizona · 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 Phoenix at 1-, 3-, and 7-day leads, then scores those forecasts against the PHX 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, ECMWF IFS has been most accurate (typical miss 0.81°C); at three days, too close to call; at seven days, too close to call.

The largest standing bias belongs to NOAA AIGFS (AI), which has run cool by about 1.74°C at 3-day lead over this window.

The Spread's own bias-corrected blend scored a 0.76°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	0.8° (-0.4°)	1.2° (+0.0°)	1.6° (-0.1°)	30
DWD ICON	1.1° (-0.7°)	1.5° (-0.4°)	—	30
NOAA GFS	1.1° (-0.4°)	1.2° (-0.5°)	2.0° (-0.7°)	30
ECMWF AIFS (AI)	1.5° (-1.2°)	1.7° (-1.5°)	2.3° (-1.6°)	30
UKMO (UK)	1.6° (-1.3°)	1.8° (-0.4°)	—	30
NOAA AIGFS (AI)	1.6° (-1.4°)	1.9° (-1.7°)	2.1° (-1.6°)	30
GEM (Canada)	1.8° (-1.2°)	2.1° (-1.4°)	1.7° (-0.8°)	30
NBM (US blend)	1.2° (-0.9°)	1.4° (-0.9°)	1.7° (-0.9°)	30
Blended (bias-corrected)	0.8° (+0.0°)	1.1° (-0.0°)	1.4° (+0.0°)	30

Method, honestly

The Spread's Phoenix point is the Phoenix Sky Harbor station (PHX), 5.7 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 PHX 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 12 days out the models said 40–43; it hit 38. GFS led (off 0.1°C); UKMO ran 2°C cold.
Mon Aug 31 11 days out the models said 40–44; it hit 33. ICON led (off 0.1°C); AIFS ran 2°C cold.
Sun Aug 30 10 days out the models said 40–45; it hit 41. GFS led (off 0.0°C); AIFS ran 6°C cold.
Sat Aug 29 9 days out the models said 41–46; it hit 45. ECMWF led (off 0.8°C); AIFS ran 5°C cold.
Fri Aug 28 8 days out the models said 41–46; it hit 44. GFS led (off 0.1°C); AIFS ran 3°C cold.
Thu Aug 27 7 days out the models said 41–46; it hit 45. ECMWF led (off 0.2°C); UKMO ran 2°C cold.
Wed Aug 26 6 days out the models said 41–46; it hit 47. ECMWF led (off 0.3°C); GEM ran 3°C cold.
Tue Aug 25 5 days out the models said 42–46; it hit 46. ECMWF led (off 1.0°C); GEM ran 3°C cold.
Mon Aug 24 4 days out the models said 39–44; it hit 42. ICON led (off 0.2°C); UKMO ran 3°C cold.
Sun Aug 23 3 days out the models said 39–45; it hit 45. ICON led (off 0.1°C); AIFS ran 3°C cold.
Mon Aug 17 13 days out the models said 38–42; it hit 45. GFS led (off 0.2°C); GEM ran 2°C cold.
Sun Aug 16 13 days out the models said 33–43; it hit 43. GFS led (off 0.3°C); UKMO ran 4°C cold.
Sat Aug 15 13 days out the models said 31–42; it hit 42. GFS led (off 0.3°C); GEM ran 2°C cold.
Fri Aug 14 13 days out the models said 36–41; it hit 38. ECMWF led (off 0.2°C); UKMO ran 2°C cold.
Thu Aug 13 13 days out the models said 38–41; it hit 36. AIGFS led (off 0.6°C); ICON ran 2°C cold.
Wed Aug 12 13 days out the models said 36–41; it hit 35. AIFS led (off 0.5°C); ECMWF ran 5°C hot.
Tue Aug 11 13 days out the models said 36–40; it hit 39. AIGFS led (off 0.3°C); GFS ran 1°C hot.
Mon Aug 10 13 days out the models said 31–43; it hit 42. ICON led (off 0.0°C); UKMO ran 2°C cold.
Sun Aug 9 13 days out the models said 32–37; it hit 44. GFS led (off 0.2°C).
Sat Aug 8 12 days out the models said 35–40; it hit 44. GEM led (off 0.1°C); ECMWF ran 2°C hot.
Fri Aug 7 11 days out the models said 38–40; it hit 43. GEM led (off 0.1°C); ECMWF ran 2°C hot.
Thu Aug 6 10 days out the models said 38–42; it hit 40. ECMWF led (off 0.4°C); GEM ran 3°C hot.
Wed Aug 5 9 days out the models said 39–42; it hit 40. ECMWF led (off 0.2°C); GEM ran 2°C hot.
Tue Aug 4 8 days out the models said 36–44; it hit 42. AIFS led (off 0.1°C); GEM ran 2°C hot.
Mon Aug 3 7 days out the models said 43–44; it hit 46. GEM led (off 0.1°C); AIGFS ran 2°C cold.
Sun Aug 2 6 days out the models said 43–45; it hit 46. ICON led (off 0.1°C); ECMWF ran 1°C hot.
Sat Aug 1 5 days out the models said 41–47; it hit 46. ICON led (off 0.1°C).
Fri Jul 31 4 days out the models said 45–47; it hit 45. AIGFS led (off 0.1°C); ECMWF ran 2°C hot.
Thu Jul 30 3 days out the models said 44–46; it hit 45. AIFS led (off 0.1°C); AIGFS ran 1°C cold.

Other cities

Atlanta, Georgia 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 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.