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) · GraphCast (IFS init) · lead day 9

ALL · graphcast_ifs · lead 9

Target day = initialization date + 9 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, GRAP_v100, starting 2026-01-01 — days before that ran a different version of the model and are excluded rather than silently pooled. Of the scored days, 3 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
tmax00Zbilinear30d21 2 QC6.14 [5.72, 6.55]1.82-4.75 [-6.05, -3.46]7.17 [6.79, 7.68]4% [2%, 6%]13%23%-0.42-0.692026-08-10 → 2026-08-30
tmax00Zbilinear90d21 2 QC6.14 [5.63, 6.63]1.82-4.75 [-6.19, -3.27]7.17 [6.66, 7.80]4% [1%, 6%]13%23%-0.42-0.692026-08-10 → 2026-08-30
tmax00Zbilinear365d27 3 QC6.57 [5.90, 7.66]2.23-4.70 [-5.80, -3.37]7.68 [6.92, 8.83]5% [2%, 7%]12%21%-0.14-0.392026-01-10 → 2026-08-30
tmax00Zbilinearall27 3 QC6.57 [5.88, 7.77]2.23-4.70 [-5.90, -3.33]7.68 [6.91, 8.94]5% [3%, 7%]12%21%-0.14-0.392026-01-10 → 2026-08-30
tmax00Znearest30d21 2 QC6.17 [5.73, 6.60]1.85-4.73 [-6.06, -3.41]7.22 [6.81, 7.73]5% [3%, 7%]14%22%-0.43-0.722026-08-10 → 2026-08-30
tmax00Znearest90d21 2 QC6.17 [5.63, 6.67]1.85-4.73 [-6.21, -3.23]7.22 [6.68, 7.83]5% [3%, 7%]14%22%-0.43-0.722026-08-10 → 2026-08-30
tmax00Znearest365d27 3 QC6.58 [5.91, 7.64]2.25-4.68 [-5.80, -3.34]7.71 [6.96, 8.83]5% [3%, 7%]13%21%-0.14-0.402026-01-10 → 2026-08-30
tmax00Znearestall27 3 QC6.58 [5.90, 7.77]2.25-4.68 [-5.92, -3.27]7.71 [6.96, 8.95]5% [3%, 7%]13%21%-0.14-0.402026-01-10 → 2026-08-30
tmax12Zbilinear30d21 2 QC6.60 [6.32, 6.91]1.08-5.99 [-6.29, -5.65]7.76 [7.43, 8.11]6% [4%, 8%]11%19%-0.53+0.002026-08-10 → 2026-08-30
tmax12Zbilinear90d21 2 QC6.60 [6.32, 6.91]1.08-5.99 [-6.30, -5.55]7.76 [7.44, 8.11]6% [4%, 8%]11%19%-0.53+0.002026-08-10 → 2026-08-30
tmax12Zbilinear365d26 3 QC6.63 [6.22, 7.03]1.56-5.77 [-6.18, -4.98]7.87 [7.40, 8.40]6% [4%, 8%]13%20%-0.20-0.002026-01-10 → 2026-08-30
tmax12Zbilinearall26 3 QC6.63 [6.29, 7.05]1.56-5.77 [-6.21, -5.02]7.87 [7.42, 8.42]6% [5%, 8%]13%20%-0.20-0.002026-01-10 → 2026-08-30
tmax12Znearest30d21 2 QC6.71 [6.38, 7.05]1.10-5.95 [-6.26, -5.61]7.90 [7.50, 8.29]5% [4%, 7%]12%20%-0.55-0.022026-08-10 → 2026-08-30
tmax12Znearest90d21 2 QC6.71 [6.35, 7.03]1.10-5.95 [-6.27, -5.52]7.90 [7.51, 8.30]5% [3%, 7%]12%20%-0.55-0.022026-08-10 → 2026-08-30
tmax12Znearest365d26 3 QC6.75 [6.32, 7.14]1.57-5.74 [-6.14, -4.96]8.00 [7.48, 8.51]6% [4%, 7%]13%20%-0.22-0.012026-01-10 → 2026-08-30
tmax12Znearestall26 3 QC6.75 [6.40, 7.18]1.57-5.74 [-6.17, -5.00]8.00 [7.55, 8.50]6% [4%, 7%]13%20%-0.22-0.012026-01-10 → 2026-08-30
tmin00Zbilinear30d21 2 QC3.52 [3.17, 3.95]1.76+0.74 [-0.45, 1.90]4.56 [4.10, 5.08]22% [17%, 26%]34%54%+0.08-0.542026-08-10 → 2026-08-30
tmin00Zbilinear90d21 2 QC3.52 [3.15, 3.95]1.76+0.74 [-0.51, 2.07]4.56 [4.07, 5.09]22% [16%, 26%]34%54%+0.08-0.542026-08-10 → 2026-08-30
tmin00Zbilinear365d27 2 QC4.11 [3.30, 5.48]2.21+0.83 [-0.25, 2.15]5.70 [4.26, 7.65]21% [16%, 25%]34%51%+0.23-0.312026-01-10 → 2026-08-30
tmin00Zbilinearall27 2 QC4.11 [3.30, 5.69]2.21+0.83 [-0.33, 2.11]5.70 [4.28, 7.85]21% [15%, 25%]34%51%+0.23-0.312026-01-10 → 2026-08-30
tmin00Znearest30d21 2 QC3.54 [3.19, 3.99]1.76+0.77 [-0.42, 1.94]4.59 [4.11, 5.11]19% [15%, 23%]35%54%+0.08-0.542026-08-10 → 2026-08-30
tmin00Znearest90d21 2 QC3.54 [3.15, 4.00]1.76+0.77 [-0.48, 2.12]4.59 [4.09, 5.12]19% [15%, 23%]35%54%+0.08-0.542026-08-10 → 2026-08-30
tmin00Znearest365d27 2 QC4.12 [3.31, 5.47]2.21+0.86 [-0.22, 2.17]5.68 [4.29, 7.57]19% [15%, 22%]34%51%+0.23-0.302026-01-10 → 2026-08-30
tmin00Znearestall27 2 QC4.12 [3.32, 5.66]2.21+0.86 [-0.30, 2.16]5.68 [4.29, 7.74]19% [14%, 22%]34%51%+0.23-0.302026-01-10 → 2026-08-30
tmin12Zbilinear30d21 2 QC3.31 [2.97, 3.62]1.20-0.14 [-0.99, 0.76]4.38 [4.00, 4.74]23% [19%, 26%]42%57%+0.14-0.052026-08-10 → 2026-08-30
tmin12Zbilinear90d21 2 QC3.31 [2.97, 3.62]1.20-0.14 [-1.18, 0.92]4.38 [3.97, 4.74]23% [19%, 27%]42%57%+0.14-0.052026-08-10 → 2026-08-30
tmin12Zbilinear365d26 2 QC3.79 [3.11, 4.87]1.57+0.09 [-0.81, 1.16]5.04 [4.12, 6.30]19% [11%, 25%]38%53%+0.26-0.132026-01-10 → 2026-08-30
tmin12Zbilinearall26 2 QC3.79 [3.11, 5.16]1.57+0.09 [-0.83, 1.01]5.04 [4.12, 6.57]19% [9%, 25%]38%53%+0.26-0.132026-01-10 → 2026-08-30
tmin12Znearest30d21 2 QC3.36 [2.99, 3.68]1.21-0.10 [-0.97, 0.79]4.44 [4.03, 4.83]20% [17%, 24%]42%57%+0.12-0.052026-08-10 → 2026-08-30
tmin12Znearest90d21 2 QC3.36 [3.00, 3.69]1.21-0.10 [-1.17, 0.96]4.44 [4.00, 4.84]20% [17%, 24%]42%57%+0.12-0.052026-08-10 → 2026-08-30
tmin12Znearest365d26 2 QC3.86 [3.15, 4.99]1.58+0.12 [-0.77, 1.19]5.18 [4.16, 6.57]18% [12%, 23%]38%52%+0.25-0.152026-01-10 → 2026-08-30
tmin12Znearestall26 2 QC3.86 [3.14, 5.30]1.58+0.12 [-0.82, 1.05]5.18 [4.16, 6.89]18% [11%, 23%]38%52%+0.25-0.152026-01-10 → 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 GraphCast (IFS init) at ALL, lead day 9, 00Z bilinear tmax+10+50-5-102026-08-10: -3.61 °F2026-08-11: -5.87 °F2026-08-12: -3.92 °F2026-08-13: -6.78 °F2026-08-14: -5.86 °F2026-08-15: -6.8 °F2026-08-16: -6.16 °F2026-08-17: -6.43 °F2026-08-18: -5.71 °F2026-08-19: -5.69 °F2026-08-20: -2.86 °F2026-08-21: +0.09 °F2026-08-22: +0.54 °F2026-08-23: -3.17 °F2026-08-24: -3.8 °F2026-08-25: -5.34 °F2026-08-26: -5.31 °F2026-08-27: -2.15 °F2026-08-28: -4.69 °F2026-08-29: -8.44 °F2026-08-30: -7.71 °F2026-08-102026-08-30°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 (21 days, newest first) · download CSV
Climatological dayError °F
2026-08-30-7.71
2026-08-29-8.44
2026-08-28-4.69
2026-08-27-2.15
2026-08-26-5.31
2026-08-25-5.34
2026-08-24-3.80
2026-08-23-3.17
2026-08-22+0.54
2026-08-21+0.09
2026-08-20-2.86
2026-08-19-5.69
2026-08-18-5.71
2026-08-17-6.43
2026-08-16-6.16
2026-08-15-6.80
2026-08-14-5.86
2026-08-13-6.78
2026-08-12-3.92
2026-08-11-5.87
2026-08-10-3.61

Error distribution

Distribution of the signed daily error in °F for GraphCast (IFS init) at ALL, lead day 9-10 to -9 °F: 1 day-9 to -8 °F: 1 day-8 to -7 °F: 3 days-7 to -6 °F: 4 days-6 to -5 °F: 6 days-5 to -4 °F: 1 day-4 to -3 °F: 5 days-3 to -2 °F: 2 days-2 to -1 °F: 1 day-1 to 0 °F: 0 days0 to 1 °F: 2 days1 to 2 °F: 0 days2 to 3 °F: 1 day3 to 4 °F: 0 days4 to 5 °F: 0 days5 to 6 °F: 0 days6 to 7 °F: 0 days7 to 8 °F: 0 days8 to 9 °F: 0 days9 to 10 °F: 0 daysdashed: P50 |e| 5.34 · P90 |e| 7.64 °F-100+10 °Fn = 27 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
-10.00-9.001
-9.00-8.001
-8.00-7.003
-7.00-6.004
-6.00-5.006
-5.00-4.001
-4.00-3.005
-3.00-2.002
-2.00-1.001
-1.00+0.000
+0.00+1.002
+1.00+2.000
+2.00+3.001
+3.00+4.000
+4.00+5.000
+5.00+6.000
+6.00+7.000
+7.00+8.000
+8.00+9.000
+9.00+10.000

Month by month

00Z bilinear daily maximum, newest first.

Monthn MAE °FBias °F
2026-08214.81 -4.75
2026-0165.42 -4.54

Availability

ModelScored period DaysCoverage
GraphCast (IFS init)2026-01-10 → 2026-08-30 27
GraphCast (IFS init)2026-01-10 → 2026-08-30 27
GraphCast (IFS init)2026-01-10 → 2026-08-30 26
GraphCast (IFS init)2026-01-10 → 2026-08-30 26

Compared with other models

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

Other modelCommon days MAE difference °FDistinguishable
ECMWF AIFS Single 21 +1.12 [0.53, 2.09] yes
NCEP GFS 21 +1.46 [0.66, 2.41] yes
ECMWF IFS HRES 16 +1.88 [-0.03, 2.79] no
Persistence (baseline) 21 +1.81 [1.45, 2.31] yes

Cite this page

ShortCastCheck, ALL graphcast_ifs lead 9, methodology v0.2, accessed 2026-08-31, https://castcheck.zifanzhang.com/station/ALL/model/graphcast_ifs/lead/9/

LongCastCheck (2026). Station-level verification of raw weather-model 2 m temperature forecasts: ALL (All stations (mean of daily station errors)), model graphcast_ifs, lead day 9. Methodology version 0.2, schema version 0.1, data through 2026-08-30. Accessed 2026-08-31. https://castcheck.zifanzhang.com/station/ALL/model/graphcast_ifs/lead/9/

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