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 6

ALL · graphcast_ifs · lead 6

Target day = initialization date + 6 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
tmax00Zbilinear30d24 2 QC5.66 [5.20, 6.14]0.73-5.34 [-5.85, -4.87]6.56 [6.08, 7.06]6% [4%, 8%]14%23%-0.43+0.492026-08-07 → 2026-08-30
tmax00Zbilinear90d24 2 QC5.66 [5.17, 6.18]0.73-5.34 [-5.88, -4.85]6.56 [6.04, 7.09]6% [4%, 8%]14%23%-0.43+0.492026-08-07 → 2026-08-30
tmax00Zbilinear365d30 3 QC5.73 [5.25, 6.20]0.97-5.08 [-5.70, -4.34]6.77 [6.17, 7.33]7% [5%, 10%]15%25%-0.10+0.532026-01-07 → 2026-08-30
tmax00Zbilinearall30 3 QC5.73 [5.26, 6.19]0.97-5.08 [-5.76, -4.31]6.77 [6.17, 7.36]7% [5%, 10%]15%25%-0.10+0.532026-01-07 → 2026-08-30
tmax00Znearest30d24 2 QC5.79 [5.29, 6.31]0.73-5.29 [-5.82, -4.79]6.76 [6.19, 7.31]8% [6%, 9%]14%24%-0.47+0.492026-08-07 → 2026-08-30
tmax00Znearest90d24 2 QC5.79 [5.25, 6.35]0.73-5.29 [-5.85, -4.77]6.76 [6.15, 7.35]8% [5%, 10%]14%24%-0.47+0.492026-08-07 → 2026-08-30
tmax00Znearest365d30 3 QC5.84 [5.35, 6.34]0.97-5.04 [-5.70, -4.30]6.93 [6.30, 7.49]8% [6%, 10%]15%26%-0.12+0.532026-01-07 → 2026-08-30
tmax00Znearestall30 3 QC5.84 [5.34, 6.33]0.97-5.04 [-5.71, -4.23]6.93 [6.32, 7.53]8% [6%, 10%]15%26%-0.12+0.532026-01-07 → 2026-08-30
tmax12Zbilinear30d24 2 QC5.64 [5.10, 6.17]0.71-5.35 [-5.91, -4.77]6.54 [5.92, 7.09]6% [4%, 8%]14%22%-0.43+0.512026-08-07 → 2026-08-30
tmax12Zbilinear90d24 2 QC5.64 [5.02, 6.22]0.71-5.35 [-5.96, -4.71]6.54 [5.84, 7.14]6% [4%, 8%]14%22%-0.43+0.512026-08-07 → 2026-08-30
tmax12Zbilinear365d29 3 QC5.53 [4.99, 6.11]1.06-4.93 [-5.80, -3.82]6.49 [5.89, 7.08]7% [5%, 9%]15%24%-0.08+0.472026-01-07 → 2026-08-30
tmax12Zbilinearall29 3 QC5.53 [4.98, 6.12]1.06-4.93 [-5.80, -3.74]6.49 [5.90, 7.06]7% [5%, 9%]15%24%-0.08+0.472026-01-07 → 2026-08-30
tmax12Znearest30d24 2 QC5.79 [5.15, 6.36]0.75-5.30 [-5.90, -4.66]6.75 [6.06, 7.34]8% [5%, 10%]16%22%-0.47+0.482026-08-07 → 2026-08-30
tmax12Znearest90d24 2 QC5.79 [5.08, 6.42]0.75-5.30 [-5.93, -4.59]6.75 [5.95, 7.39]8% [5%, 10%]16%22%-0.47+0.482026-08-07 → 2026-08-30
tmax12Znearest365d29 3 QC5.66 [5.05, 6.31]1.08-4.89 [-5.80, -3.80]6.68 [6.02, 7.33]8% [5%, 10%]16%24%-0.11+0.462026-01-07 → 2026-08-30
tmax12Znearestall29 3 QC5.66 [5.03, 6.30]1.08-4.89 [-5.76, -3.72]6.68 [6.01, 7.30]8% [5%, 10%]16%24%-0.11+0.462026-01-07 → 2026-08-30
tmin00Zbilinear30d24 2 QC2.62 [2.49, 2.77]0.73+0.59 [0.29, 0.93]3.37 [3.17, 3.59]27% [24%, 29%]47%64%+0.24+0.312026-08-07 → 2026-08-30
tmin00Zbilinear90d24 2 QC2.62 [2.48, 2.82]0.73+0.59 [0.25, 0.96]3.37 [3.17, 3.63]27% [24%, 29%]47%64%+0.24+0.312026-08-07 → 2026-08-30
tmin00Zbilinear365d30 2 QC3.02 [2.54, 3.94]0.94+0.92 [0.38, 1.81]4.00 [3.25, 5.17]25% [20%, 28%]44%60%+0.37+0.502026-01-07 → 2026-08-30
tmin00Zbilinearall30 2 QC3.02 [2.54, 4.03]0.94+0.92 [0.35, 1.82]4.00 [3.24, 5.25]25% [19%, 28%]44%60%+0.37+0.502026-01-07 → 2026-08-30
tmin00Znearest30d24 2 QC2.67 [2.50, 2.85]0.74+0.62 [0.31, 0.96]3.45 [3.20, 3.69]25% [23%, 28%]47%65%+0.22+0.302026-08-07 → 2026-08-30
tmin00Znearest90d24 2 QC2.67 [2.50, 2.89]0.74+0.62 [0.25, 1.01]3.45 [3.21, 3.73]25% [22%, 29%]47%65%+0.22+0.302026-08-07 → 2026-08-30
tmin00Znearest365d30 2 QC3.10 [2.57, 4.10]0.95+0.94 [0.38, 1.84]4.17 [3.31, 5.45]24% [20%, 27%]43%61%+0.35+0.492026-01-07 → 2026-08-30
tmin00Znearestall30 2 QC3.10 [2.56, 4.19]0.95+0.94 [0.37, 1.83]4.17 [3.29, 5.57]24% [20%, 27%]43%61%+0.35+0.492026-01-07 → 2026-08-30
tmin12Zbilinear30d24 2 QC2.59 [2.40, 2.80]0.70+0.35 [-0.02, 0.75]3.35 [3.08, 3.63]28% [25%, 31%]47%65%+0.25+0.342026-08-07 → 2026-08-30
tmin12Zbilinear90d24 2 QC2.59 [2.39, 2.82]0.70+0.35 [-0.07, 0.79]3.35 [3.05, 3.65]28% [24%, 31%]47%65%+0.25+0.342026-08-07 → 2026-08-30
tmin12Zbilinear365d29 2 QC2.93 [2.44, 3.77]0.92+0.76 [0.06, 1.84]3.92 [3.14, 5.07]26% [18%, 30%]45%60%+0.37+0.482026-01-07 → 2026-08-30
tmin12Zbilinearall29 2 QC2.93 [2.45, 3.91]0.92+0.76 [0.06, 1.93]3.92 [3.14, 5.23]26% [18%, 30%]45%60%+0.37+0.482026-01-07 → 2026-08-30
tmin12Znearest30d24 2 QC2.64 [2.39, 2.89]0.70+0.37 [-0.02, 0.76]3.43 [3.11, 3.74]27% [24%, 30%]47%65%+0.23+0.342026-08-07 → 2026-08-30
tmin12Znearest90d24 2 QC2.64 [2.38, 2.90]0.70+0.37 [-0.07, 0.83]3.43 [3.09, 3.77]27% [24%, 30%]47%65%+0.23+0.342026-08-07 → 2026-08-30
tmin12Znearest365d29 2 QC3.00 [2.45, 3.93]0.94+0.78 [0.07, 1.87]4.07 [3.20, 5.33]25% [19%, 29%]44%61%+0.36+0.472026-01-07 → 2026-08-30
tmin12Znearestall29 2 QC3.00 [2.46, 4.05]0.94+0.78 [0.07, 1.94]4.07 [3.19, 5.53]25% [19%, 29%]44%61%+0.36+0.472026-01-07 → 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 6, 00Z bilinear tmax+10+50-5-102026-08-07: -5.18 °F2026-08-08: -5.89 °F2026-08-09: -4.31 °F2026-08-10: -4.35 °F2026-08-11: -5.55 °F2026-08-12: -5.4 °F2026-08-13: -3.66 °F2026-08-14: -6.38 °F2026-08-15: -3.46 °F2026-08-16: -3.87 °F2026-08-17: -5.15 °F2026-08-18: -5.76 °F2026-08-19: -5.54 °F2026-08-20: -5.26 °F2026-08-21: -5.67 °F2026-08-22: -5.57 °F2026-08-23: -7.07 °F2026-08-24: -7.14 °F2026-08-25: -5.14 °F2026-08-26: -5.66 °F2026-08-27: -6.31 °F2026-08-28: -6.19 °F2026-08-29: -4.62 °F2026-08-30: -4.96 °F2026-08-072026-08-30°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 (24 days, newest first) · download CSV
Climatological dayError °F
2026-08-30-4.96
2026-08-29-4.62
2026-08-28-6.19
2026-08-27-6.31
2026-08-26-5.66
2026-08-25-5.14
2026-08-24-7.14
2026-08-23-7.07
2026-08-22-5.57
2026-08-21-5.67
2026-08-20-5.26
2026-08-19-5.54
2026-08-18-5.76
2026-08-17-5.15
2026-08-16-3.87
2026-08-15-3.46
2026-08-14-6.38
2026-08-13-3.66
2026-08-12-5.40
2026-08-11-5.55
2026-08-10-4.35
2026-08-09-4.31
2026-08-08-5.89
2026-08-07-5.18

Error distribution

Distribution of the signed daily error in °F for GraphCast (IFS init) at ALL, lead day 6-10 to -9 °F: 0 days-9 to -8 °F: 0 days-8 to -7 °F: 3 days-7 to -6 °F: 3 days-6 to -5 °F: 12 days-5 to -4 °F: 7 days-4 to -3 °F: 3 days-3 to -2 °F: 1 day-2 to -1 °F: 1 day-1 to 0 °F: 0 days0 to 1 °F: 0 days1 to 2 °F: 0 days2 to 3 °F: 0 days3 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| 5.22 · P90 |e| 6.45 °F-100+10 °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
-10.00-9.000
-9.00-8.000
-8.00-7.003
-7.00-6.003
-6.00-5.0012
-5.00-4.007
-4.00-3.003
-3.00-2.001
-2.00-1.001
-1.00+0.000
+0.00+1.000
+1.00+2.000
+2.00+3.000
+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
2026-08245.34 -5.34
2026-0164.06 -4.06

Availability

ModelScored period DaysCoverage
GraphCast (IFS init)2026-01-07 → 2026-08-30 30
GraphCast (IFS init)2026-01-07 → 2026-08-30 30
GraphCast (IFS init)2026-01-07 → 2026-08-30 29
GraphCast (IFS init)2026-01-07 → 2026-08-30 29

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 24 +1.49 [1.16, 1.83] yes
NCEP GFS 24 +1.96 [1.42, 2.45] yes
ECMWF IFS HRES 19 +1.74 [0.96, 2.37] yes
Persistence (baseline) 24 +1.71 [0.90, 2.47] yes

Cite this page

ShortCastCheck, ALL graphcast_ifs lead 6, methodology v0.2, accessed 2026-08-31, https://castcheck.zifanzhang.com/station/ALL/model/graphcast_ifs/lead/6/

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 6. 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/6/

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