Raw model output on the native 0.25° grid — no MOS, no bias correction, no post-processing. Scores understate operational forecast quality. Read the fairness statement →

CastCheck methodology v0.2 · data through 2026-08-30

All stations (mean of daily station errors) · ECMWF AIFS Single · lead day 0

ALL · aifs_single · lead 0

Target day = initialization date + 0 days. This page is a permanent link: the numbers change as history accumulates, the address and the methodology version do not.

Scores use the latest model version segment, aifs-single-gpid5, starting 2026-08-01 — days before that ran a different version of the model and are excluded rather than silently pooled. Of the scored days, 2 carry a QC flag on the observation.

JSON for this page CSV of the daily errors Cite this page

Every number for this combination

All four windows × both initializations × both interpolations × both variables. Errors are computed in °C and shown in °F; CI is a 95 % moving-block bootstrap interval. MAE debiased and skill, debiased remove the constant offset of the window first, so a station that the model always reads a degree too warm is not scored as if it had no skill — compare the two skill columns to see how much of the error is a fixed offset. n QC next to n counts days whose observation carries a QC flag.

VariableInitMethod Windown MAE °F MAE debiased Bias °F RMSE °F ±1 °F±2 °F±3 °F Skill Skill, debiased Period
tmax00Zbilinear30d30 2 QC3.41 [3.25, 3.56]0.39-3.06 [-3.22, -2.90]3.99 [3.83, 4.14]11% [9%, 14%]27%47%——2026-08-01 → 2026-08-30
tmax00Zbilinear90d30 2 QC3.41 [3.22, 3.61]0.39-3.06 [-3.26, -2.88]3.99 [3.80, 4.19]11% [9%, 14%]27%47%——2026-08-01 → 2026-08-30
tmax00Zbilinear365d30 2 QC3.41 [3.23, 3.63]0.39-3.06 [-3.29, -2.88]3.99 [3.80, 4.21]11% [8%, 14%]27%47%——2026-08-01 → 2026-08-30
tmax00Zbilinearall30 2 QC3.41 [3.23, 3.61]0.39-3.06 [-3.25, -2.88]3.99 [3.80, 4.19]11% [9%, 14%]27%47%——2026-08-01 → 2026-08-30
tmax00Znearest30d30 2 QC3.44 [3.27, 3.60]0.42-2.97 [-3.16, -2.76]4.04 [3.86, 4.21]12% [10%, 15%]30%46%——2026-08-01 → 2026-08-30
tmax00Znearest90d30 2 QC3.44 [3.24, 3.67]0.42-2.97 [-3.22, -2.73]4.04 [3.82, 4.27]12% [9%, 16%]30%46%——2026-08-01 → 2026-08-30
tmax00Znearest365d30 2 QC3.44 [3.24, 3.69]0.42-2.97 [-3.25, -2.74]4.04 [3.82, 4.30]12% [9%, 16%]30%46%——2026-08-01 → 2026-08-30
tmax00Znearestall30 2 QC3.44 [3.23, 3.66]0.42-2.97 [-3.21, -2.73]4.04 [3.82, 4.27]12% [9%, 16%]30%46%——2026-08-01 → 2026-08-30
tmin00Zbilinear30d30 2 QC2.12 [1.90, 2.31]0.38+0.09 [-0.07, 0.22]2.76 [2.45, 3.02]34% [31%, 36%]57%74%——2026-08-01 → 2026-08-30
tmin00Zbilinear90d30 2 QC2.12 [1.91, 2.35]0.38+0.09 [-0.13, 0.23]2.76 [2.46, 3.06]34% [31%, 36%]57%74%——2026-08-01 → 2026-08-30
tmin00Zbilinear365d30 2 QC2.12 [1.89, 2.34]0.38+0.09 [-0.17, 0.23]2.76 [2.42, 3.05]34% [31%, 37%]57%74%——2026-08-01 → 2026-08-30
tmin00Zbilinearall30 2 QC2.12 [1.91, 2.35]0.38+0.09 [-0.14, 0.24]2.76 [2.45, 3.08]34% [31%, 36%]57%74%——2026-08-01 → 2026-08-30
tmin00Znearest30d30 2 QC2.22 [1.99, 2.42]0.39+0.06 [-0.12, 0.20]2.91 [2.56, 3.18]30% [28%, 32%]56%73%——2026-08-01 → 2026-08-30
tmin00Znearest90d30 2 QC2.22 [2.00, 2.46]0.39+0.06 [-0.18, 0.21]2.91 [2.59, 3.24]30% [28%, 33%]56%73%——2026-08-01 → 2026-08-30
tmin00Znearest365d30 2 QC2.22 [1.96, 2.45]0.39+0.06 [-0.23, 0.22]2.91 [2.52, 3.23]30% [28%, 33%]56%73%——2026-08-01 → 2026-08-30
tmin00Znearestall30 2 QC2.22 [2.00, 2.47]0.39+0.06 [-0.20, 0.23]2.91 [2.56, 3.25]30% [28%, 33%]56%73%——2026-08-01 → 2026-08-30

+ model too warm · − model too cold · bias interval includes zero · ★ lowest MAE · = not distinguishable from the leader · ▼ significantly worse · n < 30 greyed and unranked · CI is a 95 % moving-block bootstrap interval.

Daily error, last 90 days

Daily forecast minus observed error in °F for ECMWF AIFS Single at ALL, lead day 0, 00Z bilinear tmax+5+2.50-2.5-52026-08-01: -2.41 °F2026-08-02: -2.89 °F2026-08-03: -3.78 °F2026-08-04: -2.58 °F2026-08-05: -3.43 °F2026-08-06: -2.72 °F2026-08-07: -3.01 °F2026-08-08: -2.4 °F2026-08-09: -2.61 °F2026-08-10: -2.45 °F2026-08-11: -2.93 °F2026-08-12: -3.35 °F2026-08-13: -3.11 °F2026-08-14: -3.76 °F2026-08-15: -2.13 °F2026-08-16: -2.8 °F2026-08-17: -3.1 °F2026-08-18: -3.12 °F2026-08-19: -2.91 °F2026-08-20: -3.18 °F2026-08-21: -3.64 °F2026-08-22: -3.38 °F2026-08-23: -3.11 °F2026-08-24: -3.38 °F2026-08-25: -2.63 °F2026-08-26: -2.89 °F2026-08-27: -4.05 °F2026-08-28: -3.92 °F2026-08-29: -3.5 °F2026-08-30: -2.69 °F2026-08-012026-08-30°F, forecast − observed
Forecast − observed, °F, 00Z bilinear daily maximum. The dashed line is zero; the thick line is a centred 7-day mean, drawn only where the whole window is present. Gaps are days with no score, never interpolation. Hover a point for its date and value.
The same series as a table (30 days, newest first) · download CSV
Climatological dayError °F
2026-08-30-2.69
2026-08-29-3.50
2026-08-28-3.92
2026-08-27-4.05
2026-08-26-2.89
2026-08-25-2.63
2026-08-24-3.38
2026-08-23-3.11
2026-08-22-3.38
2026-08-21-3.64
2026-08-20-3.18
2026-08-19-2.91
2026-08-18-3.12
2026-08-17-3.10
2026-08-16-2.80
2026-08-15-2.13
2026-08-14-3.76
2026-08-13-3.11
2026-08-12-3.35
2026-08-11-2.93
2026-08-10-2.45
2026-08-09-2.61
2026-08-08-2.40
2026-08-07-3.01
2026-08-06-2.72
2026-08-05-3.43
2026-08-04-2.58
2026-08-03-3.78
2026-08-02-2.89
2026-08-01-2.41

Error distribution

Distribution of the signed daily error in °F for ECMWF AIFS Single at ALL, lead day 0-5 to -4.5 °F: 0 days-4.5 to -4 °F: 1 day-4 to -3.5 °F: 5 days-3.5 to -3 °F: 10 days-3 to -2.5 °F: 10 days-2.5 to -2 °F: 4 days-2 to -1.5 °F: 0 days-1.5 to -1 °F: 0 days-1 to -0.5 °F: 0 days-0.5 to 0 °F: 0 days0 to 0.5 °F: 0 days0.5 to 1 °F: 0 days1 to 1.5 °F: 0 days1.5 to 2 °F: 0 days2 to 2.5 °F: 0 days2.5 to 3 °F: 0 days3 to 3.5 °F: 0 days3.5 to 4 °F: 0 days4 to 4.5 °F: 0 days4.5 to 5 °F: 0 daysdashed: P50 |e| 3.05 · P90 |e| 3.76 °F-50+5 °Fn = 30 days
All scored days for 00Z bilinear daily maximum, 20 equal bins. The dashed marks are the 50th and 90th percentiles of the absolute error, mirrored on both sides. Hover a bar for its bin and count.
Bin counts as a table
From °FTo °FDays
-5.00-4.500
-4.50-4.001
-4.00-3.505
-3.50-3.0010
-3.00-2.5010
-2.50-2.004
-2.00-1.500
-1.50-1.000
-1.00-0.500
-0.50+0.000
+0.00+0.500
+0.50+1.000
+1.00+1.500
+1.50+2.000
+2.00+2.500
+2.50+3.000
+3.00+3.500
+3.50+4.000
+4.00+4.500
+4.50+5.000

Month by month

00Z bilinear daily maximum, newest first.

Monthn MAE °FBias °F
2026-08303.06 -3.06

Availability

ModelScored period DaysCoverage
ECMWF AIFS Single2026-08-01 → 2026-08-30 30
ECMWF AIFS Single2026-08-01 → 2026-08-30 30

Compared with other models

Paired bootstrap on the days both models cover (90d, 00Z, bilinear, daily maximum). Negative means ECMWF AIFS Single has the lower MAE. "Distinguishable" means the 95 % interval excludes zero.

Other modelCommon days MAE difference °FDistinguishable
Aurora (GFS init) 2 -2.31 [-2.38, -2.23] yes
Aurora (IFS init) 3 -1.40 [-1.48, -1.23] yes
FourCastNet v2 (GFS init) 2 -0.87 [-1.07, -0.67] yes
FourCastNet v2 (IFS init) 3 -0.91 [-1.36, -0.69] yes
NCEP GFS 30 +0.77 [0.61, 0.90] yes
GraphCast (GFS init) 2 -2.11 [-2.20, -2.01] yes
GraphCast (IFS init) 28 -1.18 [-1.28, -1.08] yes
ECMWF IFS HRES 25 +0.19 [0.07, 0.28] yes
Pangu-Weather (GFS init) 2 -0.24 [-0.47, -0.02] yes
Pangu-Weather (IFS init) 3 +0.06 [-0.26, 0.52] no

Cite this page

ShortCastCheck, ALL aifs_single lead 0, methodology v0.2, accessed 2026-08-31, https://castcheck.zifanzhang.com/station/ALL/model/aifs_single/lead/0/

LongCastCheck (2026). Station-level verification of raw weather-model 2 m temperature forecasts: ALL (All stations (mean of daily station errors)), model aifs_single, lead day 0. Methodology version 0.2, schema version 0.1, data through 2026-08-30. Accessed 2026-08-31. https://castcheck.zifanzhang.com/station/ALL/model/aifs_single/lead/0/

Accessed2026-08-31 (data through 2026-08-30; page built 2026-08-31T03:52:06+00:00)

JSON CSV all scores methodology v0.2 licence