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

Atlanta Hartsfield · GraphCast (IFS init) · lead day 5

KATL · 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, 1 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 1 QC5.04 [3.33, 6.73]2.24-5.02 [-6.73, -3.31]5.82 [4.04, 7.37]12% [0%, 29%]12%20%-0.55+0.302026-08-06 → 2026-08-30
tmax00Zbilinear90d25 1 QC5.04 [3.13, 6.96]2.24-5.02 [-6.96, -3.10]5.82 [3.79, 7.56]12% [0%, 29%]12%20%-0.55+0.302026-08-06 → 2026-08-30
tmax00Zbilinear365d31 1 QC5.31 [3.54, 6.80]2.18-5.29 [-6.80, -3.50]6.00 [4.28, 7.36]10% [0%, 25%]10%16%-0.36+0.432026-01-06 → 2026-08-30
tmax00Zbilinearall31 1 QC5.31 [3.66, 6.76]2.18-5.29 [-6.76, -3.63]6.00 [4.48, 7.30]10% [0%, 25%]10%16%-0.36+0.432026-01-06 → 2026-08-30
tmax00Znearest30d25 1 QC4.99 [3.31, 6.67]2.23-4.98 [-6.66, -3.30]5.78 [4.01, 7.32]12% [0%, 29%]12%20%-0.54+0.312026-08-06 → 2026-08-30
tmax00Znearest90d25 1 QC4.99 [3.10, 6.86]2.23-4.98 [-6.86, -3.09]5.78 [3.77, 7.51]12% [0%, 29%]12%20%-0.54+0.312026-08-06 → 2026-08-30
tmax00Znearest365d31 1 QC5.30 [3.55, 6.78]2.19-5.30 [-6.78, -3.53]6.00 [4.28, 7.34]10% [0%, 25%]10%16%-0.36+0.432026-01-06 → 2026-08-30
tmax00Znearestall31 1 QC5.30 [3.65, 6.78]2.19-5.30 [-6.78, -3.64]6.00 [4.45, 7.29]10% [0%, 25%]10%16%-0.36+0.432026-01-06 → 2026-08-30
tmax12Zbilinear30d25 1 QC5.55 [3.85, 7.25]2.16-5.55 [-7.25, -3.85]6.19 [4.37, 7.79]0% [0%, 0%]8%24%-0.71+0.332026-08-06 → 2026-08-30
tmax12Zbilinear90d25 1 QC5.55 [3.63, 7.53]2.16-5.55 [-7.53, -3.63]6.19 [4.17, 8.02]0% [0%, 0%]8%24%-0.71+0.332026-08-06 → 2026-08-30
tmax12Zbilinear365d30 1 QC5.70 [4.09, 7.27]1.94-5.70 [-7.27, -4.09]6.24 [4.60, 7.71]0% [0%, 0%]7%20%-0.49+0.482026-01-06 → 2026-08-30
tmax12Zbilinearall30 1 QC5.70 [4.11, 7.24]1.94-5.70 [-7.24, -4.11]6.24 [4.70, 7.68]0% [0%, 0%]7%20%-0.49+0.482026-01-06 → 2026-08-30
tmax12Znearest30d25 1 QC5.53 [3.83, 7.20]2.16-5.53 [-7.20, -3.83]6.18 [4.34, 7.79]0% [0%, 0%]8%24%-0.71+0.332026-08-06 → 2026-08-30
tmax12Znearest90d25 1 QC5.53 [3.62, 7.52]2.16-5.53 [-7.52, -3.62]6.18 [4.15, 8.02]0% [0%, 0%]8%24%-0.71+0.332026-08-06 → 2026-08-30
tmax12Znearest365d30 1 QC5.70 [4.08, 7.24]1.98-5.70 [-7.24, -4.08]6.25 [4.60, 7.72]0% [0%, 0%]7%20%-0.49+0.462026-01-06 → 2026-08-30
tmax12Znearestall30 1 QC5.70 [4.12, 7.21]1.98-5.70 [-7.21, -4.12]6.25 [4.70, 7.69]0% [0%, 0%]7%20%-0.49+0.462026-01-06 → 2026-08-30
tmin00Zbilinear30d25 1 QC1.96 [1.15, 2.50]1.97+0.03 [-1.19, 1.06]2.43 [1.62, 2.85]32% [8%, 67%]60%80%+0.17+0.172026-08-06 → 2026-08-30
tmin00Zbilinear90d25 1 QC1.96 [0.88, 2.48]1.97+0.03 [-1.34, 1.09]2.43 [1.15, 2.82]32% [8%, 80%]60%80%+0.17+0.172026-08-06 → 2026-08-30
tmin00Zbilinear365d31 1 QC2.24 [1.38, 3.01]2.23-0.06 [-1.32, 1.15]2.98 [1.95, 4.04]35% [12%, 65%]58%74%+0.45+0.482026-01-06 → 2026-08-30
tmin00Zbilinearall31 1 QC2.24 [1.46, 3.19]2.23-0.06 [-1.34, 1.14]2.98 [2.03, 4.21]35% [14%, 67%]58%74%+0.45+0.482026-01-06 → 2026-08-30
tmin00Znearest30d25 1 QC1.99 [1.20, 2.52]1.98-0.07 [-1.30, 0.96]2.46 [1.62, 2.91]32% [8%, 67%]60%84%+0.16+0.162026-08-06 → 2026-08-30
tmin00Znearest90d25 1 QC1.99 [0.91, 2.50]1.98-0.07 [-1.44, 0.98]2.46 [1.17, 2.90]32% [8%, 80%]60%84%+0.16+0.162026-08-06 → 2026-08-30
tmin00Znearest365d31 1 QC2.30 [1.47, 3.09]2.25-0.20 [-1.47, 1.02]3.01 [1.97, 4.02]29% [8%, 60%]58%77%+0.44+0.482026-01-06 → 2026-08-30
tmin00Znearestall31 1 QC2.30 [1.52, 3.26]2.25-0.20 [-1.47, 1.00]3.01 [2.05, 4.24]29% [8%, 60%]58%77%+0.44+0.482026-01-06 → 2026-08-30
tmin12Zbilinear30d25 1 QC1.83 [1.16, 2.26]1.82-0.13 [-1.22, 0.72]2.33 [1.51, 2.85]24% [4%, 53%]64%88%+0.22+0.232026-08-06 → 2026-08-30
tmin12Zbilinear90d25 1 QC1.83 [0.90, 2.25]1.82-0.13 [-1.40, 0.76]2.33 [1.35, 2.85]24% [3%, 65%]64%88%+0.22+0.232026-08-06 → 2026-08-30
tmin12Zbilinear365d30 1 QC2.05 [1.33, 2.74]2.01-0.56 [-1.92, 0.62]2.88 [1.74, 4.02]23% [6%, 50%]63%87%+0.46+0.502026-01-06 → 2026-08-30
tmin12Zbilinearall30 1 QC2.05 [1.33, 2.75]2.01-0.56 [-2.14, 0.60]2.88 [1.73, 4.17]23% [6%, 50%]63%87%+0.46+0.502026-01-06 → 2026-08-30
tmin12Znearest30d25 1 QC1.87 [1.18, 2.31]1.85-0.23 [-1.35, 0.64]2.38 [1.52, 2.96]28% [12%, 55%]64%84%+0.21+0.222026-08-06 → 2026-08-30
tmin12Znearest90d25 1 QC1.87 [0.95, 2.32]1.85-0.23 [-1.52, 0.68]2.38 [1.37, 2.97]28% [11%, 65%]64%84%+0.21+0.222026-08-06 → 2026-08-30
tmin12Znearest365d30 1 QC2.10 [1.36, 2.79]2.03-0.70 [-2.16, 0.50]2.98 [1.76, 4.18]30% [14%, 57%]63%83%+0.44+0.502026-01-06 → 2026-08-30
tmin12Znearestall30 1 QC2.10 [1.36, 2.80]2.03-0.70 [-2.37, 0.50]2.98 [1.74, 4.37]30% [14%, 56%]63%83%+0.44+0.502026-01-06 → 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 KATL, lead day 5, 00Z bilinear tmax+20+100-10-202026-08-06: -7 °F2026-08-07: -5.81 °F2026-08-08: -6.86 °F2026-08-09: -5.45 °F2026-08-10: -5.19 °F2026-08-11: -5.11 °F2026-08-12: -6.32 °F2026-08-13: -8.2 °F2026-08-14: -6.42 °F2026-08-15: -3.39 °F2026-08-16: -3.65 °F2026-08-17: -4.23 °F2026-08-18: +0.22 °F2026-08-19: -3.27 °F2026-08-20: -2.53 °F2026-08-21: -0.39 °F2026-08-22: -0.54 °F2026-08-23: -2.91 °F2026-08-24: -4.05 °F2026-08-25: -6.47 °F2026-08-26: -3.54 °F2026-08-27: -12.85 °F2026-08-28: -11.24 °F2026-08-29: -3.91 °F2026-08-30: -6.34 °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
2026-08-30-6.34
2026-08-29-3.91
2026-08-28-11.24
2026-08-27-12.85
2026-08-26-3.54
2026-08-25-6.47
2026-08-24-4.05
2026-08-23-2.91
2026-08-22-0.54
2026-08-21-0.39
2026-08-20-2.53
2026-08-19-3.27
2026-08-18+0.22
2026-08-17-4.23
2026-08-16-3.65
2026-08-15-3.39
2026-08-14-6.42
2026-08-13-8.20
2026-08-12-6.32
2026-08-11-5.11
2026-08-10-5.19
2026-08-09-5.45
2026-08-08-6.86
2026-08-07-5.81
2026-08-06-7.00

Error distribution

Distribution of the signed daily error in °F for GraphCast (IFS init) at KATL, lead day 5-20 to -18 °F: 0 days-18 to -16 °F: 0 days-16 to -14 °F: 0 days-14 to -12 °F: 1 day-12 to -10 °F: 1 day-10 to -8 °F: 3 days-8 to -6 °F: 8 days-6 to -4 °F: 7 days-4 to -2 °F: 8 days-2 to 0 °F: 2 days0 to 2 °F: 1 day2 to 4 °F: 0 days4 to 6 °F: 0 days6 to 8 °F: 0 days8 to 10 °F: 0 days10 to 12 °F: 0 days12 to 14 °F: 0 days14 to 16 °F: 0 days16 to 18 °F: 0 days18 to 20 °F: 0 daysdashed: P50 |e| 5.19 · P90 |e| 8.32 °F-200+20 °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
-20.00-18.000
-18.00-16.000
-16.00-14.000
-14.00-12.001
-12.00-10.001
-10.00-8.003
-8.00-6.008
-6.00-4.007
-4.00-2.008
-2.00+0.002
+0.00+2.001
+2.00+4.000
+4.00+6.000
+6.00+8.000
+8.00+10.000
+10.00+12.000
+12.00+14.000
+14.00+16.000
+16.00+18.000
+18.00+20.000

Month by month

00Z bilinear daily maximum, newest first.

Monthn MAE °FBias °F
2026-08255.04 -5.02
2026-0166.43 -6.43

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 +0.27 [-0.75, 1.70] no
NCEP GFS 25 +1.02 [-1.31, 3.60] no
ECMWF IFS HRES 18 +1.47 [0.63, 3.69] yes
Persistence (baseline) 25 +1.80 [0.42, 3.71] yes

Cite this page

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

LongCastCheck (2026). Station-level verification of raw weather-model 2 m temperature forecasts: KATL (Atlanta Hartsfield), 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/KATL/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