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 5

ALL · graphcast_ifs · lead 5

Target day = initialization date + 5 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
tmax00Zbilinear30d25 2 QC5.41 [4.95, 5.85]0.62-5.07 [-5.43, -4.67]6.21 [5.67, 6.67]5% [3%, 6%]13%23%-0.38+0.592026-08-06 → 2026-08-30
tmax00Zbilinear90d25 2 QC5.41 [4.88, 5.88]0.62-5.07 [-5.46, -4.59]6.21 [5.57, 6.70]5% [3%, 6%]13%23%-0.38+0.592026-08-06 → 2026-08-30
tmax00Zbilinear365d31 3 QC5.38 [4.96, 5.83]0.79-4.80 [-5.35, -4.12]6.25 [5.75, 6.70]6% [4%, 7%]14%25%-0.05+0.642026-01-06 → 2026-08-30
tmax00Zbilinearall31 3 QC5.38 [4.96, 5.81]0.79-4.80 [-5.32, -4.07]6.25 [5.74, 6.69]6% [4%, 7%]14%25%-0.05+0.642026-01-06 → 2026-08-30
tmax00Znearest30d25 2 QC5.56 [5.08, 6.03]0.66-5.03 [-5.45, -4.57]6.44 [5.87, 6.93]7% [4%, 8%]16%22%-0.42+0.562026-08-06 → 2026-08-30
tmax00Znearest90d25 2 QC5.56 [5.02, 6.08]0.66-5.03 [-5.49, -4.47]6.44 [5.78, 6.97]7% [4%, 9%]16%22%-0.42+0.562026-08-06 → 2026-08-30
tmax00Znearest365d31 3 QC5.52 [5.07, 6.01]0.83-4.77 [-5.36, -4.08]6.44 [5.90, 6.96]7% [5%, 9%]16%24%-0.07+0.622026-01-06 → 2026-08-30
tmax00Znearestall31 3 QC5.52 [5.07, 5.98]0.83-4.77 [-5.32, -4.02]6.44 [5.92, 6.93]7% [5%, 8%]16%24%-0.07+0.622026-01-06 → 2026-08-30
tmax12Zbilinear30d25 2 QC5.73 [5.34, 6.17]0.75-5.44 [-5.88, -5.06]6.52 [6.08, 6.95]5% [4%, 7%]13%21%-0.46+0.502026-08-06 → 2026-08-30
tmax12Zbilinear90d25 2 QC5.73 [5.29, 6.24]0.75-5.44 [-5.92, -5.00]6.52 [6.00, 7.02]5% [4%, 7%]13%21%-0.46+0.502026-08-06 → 2026-08-30
tmax12Zbilinear365d30 2 QC5.54 [4.99, 6.06]0.89-5.17 [-5.76, -4.43]6.39 [5.89, 6.88]7% [5%, 10%]15%24%-0.10+0.592026-01-06 → 2026-08-30
tmax12Zbilinearall30 2 QC5.54 [4.96, 6.02]0.89-5.17 [-5.73, -4.39]6.39 [5.84, 6.84]7% [5%, 11%]15%24%-0.10+0.592026-01-06 → 2026-08-30
tmax12Znearest30d25 2 QC5.87 [5.43, 6.37]0.78-5.39 [-5.89, -4.96]6.72 [6.22, 7.19]7% [4%, 9%]15%21%-0.49+0.482026-08-06 → 2026-08-30
tmax12Znearest90d25 2 QC5.87 [5.38, 6.46]0.78-5.39 [-5.92, -4.87]6.72 [6.16, 7.28]7% [3%, 9%]15%21%-0.49+0.482026-08-06 → 2026-08-30
tmax12Znearest365d30 2 QC5.66 [5.06, 6.24]0.91-5.13 [-5.77, -4.38]6.57 [6.00, 7.12]8% [5%, 10%]17%24%-0.13+0.582026-01-06 → 2026-08-30
tmax12Znearestall30 2 QC5.66 [5.05, 6.21]0.91-5.13 [-5.71, -4.35]6.57 [5.98, 7.07]8% [5%, 10%]17%24%-0.13+0.582026-01-06 → 2026-08-30
tmin00Zbilinear30d25 2 QC2.47 [2.33, 2.63]0.59+0.52 [0.23, 0.83]3.19 [2.97, 3.41]27% [24%, 29%]51%67%+0.22+0.362026-08-06 → 2026-08-30
tmin00Zbilinear90d25 2 QC2.47 [2.31, 2.70]0.59+0.52 [0.19, 0.87]3.19 [2.95, 3.47]27% [23%, 30%]51%67%+0.22+0.362026-08-06 → 2026-08-30
tmin00Zbilinear365d31 2 QC2.79 [2.37, 3.50]0.81+0.88 [0.35, 1.74]3.75 [3.04, 4.75]27% [23%, 29%]48%63%+0.39+0.552026-01-06 → 2026-08-30
tmin00Zbilinearall31 2 QC2.79 [2.38, 3.60]0.81+0.88 [0.29, 1.79]3.75 [3.06, 4.87]27% [24%, 29%]48%63%+0.39+0.552026-01-06 → 2026-08-30
tmin00Znearest30d25 2 QC2.48 [2.30, 2.66]0.59+0.54 [0.24, 0.86]3.25 [2.99, 3.48]29% [25%, 32%]53%68%+0.21+0.352026-08-06 → 2026-08-30
tmin00Znearest90d25 2 QC2.48 [2.27, 2.72]0.59+0.54 [0.19, 0.92]3.25 [2.96, 3.54]29% [25%, 33%]53%68%+0.21+0.352026-08-06 → 2026-08-30
tmin00Znearest365d31 2 QC2.85 [2.36, 3.65]0.81+0.89 [0.34, 1.73]3.88 [3.08, 4.99]27% [22%, 31%]49%63%+0.38+0.552026-01-06 → 2026-08-30
tmin00Znearestall31 2 QC2.85 [2.38, 3.75]0.81+0.89 [0.30, 1.79]3.88 [3.11, 5.13]27% [22%, 31%]49%63%+0.38+0.552026-01-06 → 2026-08-30
tmin12Zbilinear30d25 2 QC2.42 [2.23, 2.60]0.65+0.31 [-0.03, 0.62]3.14 [2.89, 3.38]29% [25%, 32%]52%68%+0.23+0.292026-08-06 → 2026-08-30
tmin12Zbilinear90d25 2 QC2.42 [2.21, 2.65]0.65+0.31 [-0.07, 0.66]3.14 [2.86, 3.42]29% [24%, 32%]52%68%+0.23+0.292026-08-06 → 2026-08-30
tmin12Zbilinear365d30 2 QC2.67 [2.28, 3.32]0.86+0.65 [0.07, 1.54]3.55 [2.97, 4.39]28% [24%, 32%]49%65%+0.40+0.502026-01-06 → 2026-08-30
tmin12Zbilinearall30 2 QC2.67 [2.30, 3.41]0.86+0.65 [0.02, 1.58]3.55 [2.98, 4.54]28% [24%, 32%]49%65%+0.40+0.502026-01-06 → 2026-08-30
tmin12Znearest30d25 2 QC2.45 [2.22, 2.65]0.66+0.34 [-0.01, 0.65]3.21 [2.91, 3.47]28% [25%, 32%]53%69%+0.22+0.282026-08-06 → 2026-08-30
tmin12Znearest90d25 2 QC2.45 [2.20, 2.70]0.66+0.34 [-0.07, 0.70]3.21 [2.88, 3.52]28% [24%, 33%]53%69%+0.22+0.282026-08-06 → 2026-08-30
tmin12Znearest365d30 2 QC2.72 [2.28, 3.44]0.84+0.66 [0.09, 1.51]3.66 [3.00, 4.58]27% [23%, 31%]50%65%+0.38+0.512026-01-06 → 2026-08-30
tmin12Znearestall30 2 QC2.72 [2.31, 3.53]0.84+0.66 [0.05, 1.56]3.66 [3.02, 4.73]27% [23%, 31%]50%65%+0.38+0.512026-01-06 → 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 5, 00Z bilinear tmax+10+50-5-102026-08-06: -4.63 °F2026-08-07: -5.34 °F2026-08-08: -4.7 °F2026-08-09: -4.23 °F2026-08-10: -3.63 °F2026-08-11: -4.2 °F2026-08-12: -6.01 °F2026-08-13: -4.22 °F2026-08-14: -4.63 °F2026-08-15: -4.93 °F2026-08-16: -4.42 °F2026-08-17: -5.01 °F2026-08-18: -5.21 °F2026-08-19: -6.33 °F2026-08-20: -5.5 °F2026-08-21: -5.36 °F2026-08-22: -4.6 °F2026-08-23: -6.1 °F2026-08-24: -6.03 °F2026-08-25: -5.01 °F2026-08-26: -5.65 °F2026-08-27: -5.64 °F2026-08-28: -6.31 °F2026-08-29: -4.69 °F2026-08-30: -4.36 °F2026-08-062026-08-30°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 (25 days, newest first) · download CSV
Climatological dayError °F
2026-08-30-4.36
2026-08-29-4.69
2026-08-28-6.31
2026-08-27-5.64
2026-08-26-5.65
2026-08-25-5.01
2026-08-24-6.03
2026-08-23-6.10
2026-08-22-4.60
2026-08-21-5.36
2026-08-20-5.50
2026-08-19-6.33
2026-08-18-5.21
2026-08-17-5.01
2026-08-16-4.42
2026-08-15-4.93
2026-08-14-4.63
2026-08-13-4.22
2026-08-12-6.01
2026-08-11-4.20
2026-08-10-3.63
2026-08-09-4.23
2026-08-08-4.70
2026-08-07-5.34
2026-08-06-4.63

Error distribution

Distribution of the signed daily error in °F for GraphCast (IFS init) at ALL, lead day 5-10 to -9 °F: 0 days-9 to -8 °F: 0 days-8 to -7 °F: 0 days-7 to -6 °F: 5 days-6 to -5 °F: 9 days-5 to -4 °F: 13 days-4 to -3 °F: 2 days-3 to -2 °F: 0 days-2 to -1 °F: 2 days-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| 4.93 · P90 |e| 6.03 °F-100+10 °Fn = 31 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
-8.00-7.000
-7.00-6.005
-6.00-5.009
-5.00-4.0013
-4.00-3.002
-3.00-2.000
-2.00-1.002
-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
2026-08255.07 -5.07
2026-0163.66 -3.66

Availability

ModelScored period DaysCoverage
GraphCast (IFS init)2026-01-06 → 2026-08-30 31
GraphCast (IFS init)2026-01-06 → 2026-08-30 31
GraphCast (IFS init)2026-01-06 → 2026-08-30 30
GraphCast (IFS init)2026-01-06 → 2026-08-30 30

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 25 +1.51 [1.02, 1.92] yes
NCEP GFS 25 +1.85 [1.50, 2.28] yes
ECMWF IFS HRES 18 +1.90 [1.29, 2.29] yes
Persistence (baseline) 25 +1.48 [0.52, 2.30] yes

Cite this page

ShortCastCheck, ALL graphcast_ifs lead 5, methodology v0.2, accessed 2026-08-31, https://castcheck.zifanzhang.com/station/ALL/model/graphcast_ifs/lead/5/

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 5. 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/5/

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