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 9

KATL · 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, 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
tmax00Zbilinear30d21 1 QC6.02 [4.26, 8.01]2.99-5.78 [-8.01, -3.68]6.84 [5.10, 8.47]5% [0%, 12%]14%29%-0.39+0.302026-08-10 → 2026-08-30
tmax00Zbilinear90d21 1 QC6.02 [4.24, 8.36]2.99-5.78 [-8.36, -3.60]6.84 [4.98, 8.83]5% [0%, 13%]14%29%-0.39+0.302026-08-10 → 2026-08-30
tmax00Zbilinear365d27 1 QC6.92 [4.83, 9.48]4.18-5.54 [-8.23, -2.82]7.78 [5.73, 9.87]4% [0%, 11%]11%22%-0.26+0.232026-01-10 → 2026-08-30
tmax00Zbilinearall27 1 QC6.92 [4.86, 9.78]4.18-5.54 [-8.21, -2.60]7.78 [5.77, 10.17]4% [0%, 11%]11%22%-0.26+0.232026-01-10 → 2026-08-30
tmax00Znearest30d21 1 QC6.04 [4.30, 7.99]2.94-5.80 [-7.99, -3.71]6.83 [5.07, 8.43]5% [0%, 12%]10%19%-0.39+0.322026-08-10 → 2026-08-30
tmax00Znearest90d21 1 QC6.04 [4.24, 8.29]2.94-5.80 [-8.29, -3.58]6.83 [4.97, 8.66]5% [0%, 13%]10%19%-0.39+0.322026-08-10 → 2026-08-30
tmax00Znearest365d27 1 QC6.97 [4.86, 9.57]4.15-5.61 [-8.28, -2.93]7.83 [5.75, 9.94]4% [0%, 11%]7%15%-0.27+0.232026-01-10 → 2026-08-30
tmax00Znearestall27 1 QC6.97 [4.90, 9.85]4.15-5.61 [-8.25, -2.70]7.83 [5.77, 10.27]4% [0%, 11%]7%15%-0.27+0.232026-01-10 → 2026-08-30
tmax12Zbilinear30d21 1 QC5.69 [3.80, 8.24]3.36-4.92 [-8.12, -2.01]6.56 [4.44, 8.74]10% [0%, 22%]14%19%-0.31+0.222026-08-10 → 2026-08-30
tmax12Zbilinear90d21 1 QC5.69 [3.71, 8.78]3.36-4.92 [-8.78, -1.89]6.56 [4.38, 9.25]10% [0%, 24%]14%19%-0.31+0.222026-08-10 → 2026-08-30
tmax12Zbilinear365d26 1 QC6.82 [4.27, 10.33]4.21-5.59 [-9.06, -2.52]7.84 [5.04, 10.69]8% [0%, 19%]12%15%-0.41+0.132026-01-10 → 2026-08-30
tmax12Zbilinearall26 1 QC6.82 [4.24, 10.85]4.21-5.59 [-8.92, -2.50]7.84 [4.96, 11.17]8% [0%, 20%]12%15%-0.41+0.132026-01-10 → 2026-08-30
tmax12Znearest30d21 1 QC5.72 [3.77, 8.28]3.31-4.94 [-8.19, -2.00]6.60 [4.44, 8.83]10% [0%, 22%]14%19%-0.32+0.232026-08-10 → 2026-08-30
tmax12Znearest90d21 1 QC5.72 [3.72, 8.87]3.31-4.94 [-8.87, -1.89]6.60 [4.39, 9.31]10% [0%, 24%]14%19%-0.32+0.232026-08-10 → 2026-08-30
tmax12Znearest365d26 1 QC6.88 [4.29, 10.49]4.22-5.65 [-9.05, -2.53]7.93 [5.05, 10.81]8% [0%, 19%]12%15%-0.42+0.132026-01-10 → 2026-08-30
tmax12Znearestall26 1 QC6.88 [4.26, 11.03]4.22-5.65 [-9.13, -2.53]7.93 [4.96, 11.38]8% [0%, 20%]12%15%-0.42+0.132026-01-10 → 2026-08-30
tmin00Zbilinear30d21 1 QC2.62 [2.13, 2.92]2.61+0.16 [-1.29, 1.67]2.99 [2.40, 3.35]24% [5%, 47%]33%62%+0.05+0.062026-08-10 → 2026-08-30
tmin00Zbilinear90d21 1 QC2.62 [2.00, 2.91]2.61+0.16 [-1.48, 1.81]2.99 [2.25, 3.32]24% [0%, 50%]33%62%+0.05+0.062026-08-10 → 2026-08-30
tmin00Zbilinear365d27 1 QC3.11 [2.27, 4.64]3.14+0.94 [-0.91, 3.32]4.62 [2.61, 7.54]30% [11%, 50%]41%63%+0.43+0.412026-01-10 → 2026-08-30
tmin00Zbilinearall27 1 QC3.11 [2.29, 4.65]3.14+0.94 [-0.89, 3.33]4.62 [2.70, 7.63]30% [10%, 50%]41%63%+0.43+0.412026-01-10 → 2026-08-30
tmin00Znearest30d21 1 QC2.67 [2.17, 2.97]2.67+0.04 [-1.41, 1.53]3.04 [2.48, 3.37]24% [5%, 47%]38%62%+0.03+0.042026-08-10 → 2026-08-30
tmin00Znearest90d21 1 QC2.67 [2.05, 2.96]2.67+0.04 [-1.59, 1.64]3.04 [2.34, 3.35]24% [0%, 50%]38%62%+0.03+0.042026-08-10 → 2026-08-30
tmin00Znearest365d27 1 QC3.15 [2.32, 4.65]3.18+0.80 [-0.97, 3.14]4.59 [2.69, 7.39]26% [9%, 45%]44%63%+0.42+0.402026-01-10 → 2026-08-30
tmin00Znearestall27 1 QC3.15 [2.34, 4.68]3.18+0.80 [-0.98, 3.15]4.59 [2.78, 7.49]26% [9%, 44%]44%63%+0.42+0.402026-01-10 → 2026-08-30
tmin12Zbilinear30d21 1 QC2.71 [1.92, 3.59]2.82+0.78 [-0.96, 2.56]3.20 [2.19, 4.10]14% [4%, 24%]38%62%+0.02-0.022026-08-10 → 2026-08-30
tmin12Zbilinear90d21 1 QC2.71 [1.84, 3.72]2.82+0.78 [-1.32, 2.87]3.20 [2.12, 4.22]14% [0%, 25%]38%62%+0.02-0.022026-08-10 → 2026-08-30
tmin12Zbilinear365d26 1 QC2.88 [2.10, 3.76]2.95+0.64 [-0.88, 2.29]3.43 [2.39, 4.28]15% [5%, 25%]35%58%+0.40+0.372026-01-10 → 2026-08-30
tmin12Zbilinearall26 1 QC2.88 [2.16, 3.74]2.95+0.64 [-1.01, 2.21]3.43 [2.43, 4.27]15% [6%, 24%]35%58%+0.40+0.372026-01-10 → 2026-08-30
tmin12Znearest30d21 1 QC2.74 [1.96, 3.56]2.83+0.65 [-1.11, 2.43]3.18 [2.23, 4.00]10% [0%, 18%]38%62%+0.01-0.022026-08-10 → 2026-08-30
tmin12Znearest90d21 1 QC2.74 [1.88, 3.67]2.83+0.65 [-1.46, 2.74]3.18 [2.11, 4.13]10% [0%, 20%]38%62%+0.01-0.022026-08-10 → 2026-08-30
tmin12Znearest365d26 1 QC2.91 [2.16, 3.73]2.96+0.51 [-1.02, 2.15]3.43 [2.45, 4.31]12% [0%, 20%]38%58%+0.40+0.372026-01-10 → 2026-08-30
tmin12Znearestall26 1 QC2.91 [2.22, 3.71]2.96+0.51 [-1.17, 2.10]3.43 [2.51, 4.25]12% [0%, 21%]38%58%+0.40+0.372026-01-10 → 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 KATL, lead day 9, 00Z bilinear tmax+20+100-10-202026-08-10: -8.16 °F2026-08-11: -7.66 °F2026-08-12: -5.32 °F2026-08-13: -11.23 °F2026-08-14: -10.6 °F2026-08-15: -4.36 °F2026-08-16: -6.78 °F2026-08-17: -6.37 °F2026-08-18: -1.88 °F2026-08-19: -2.92 °F2026-08-20: -8.26 °F2026-08-21: +2.49 °F2026-08-22: -0.95 °F2026-08-23: -3.3 °F2026-08-24: -5.82 °F2026-08-25: -6.02 °F2026-08-26: -1.5 °F2026-08-27: -2.88 °F2026-08-28: -12.19 °F2026-08-29: -8.79 °F2026-08-30: -8.97 °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
2026-08-30-8.97
2026-08-29-8.79
2026-08-28-12.19
2026-08-27-2.88
2026-08-26-1.50
2026-08-25-6.02
2026-08-24-5.82
2026-08-23-3.30
2026-08-22-0.95
2026-08-21+2.49
2026-08-20-8.26
2026-08-19-2.92
2026-08-18-1.88
2026-08-17-6.37
2026-08-16-6.78
2026-08-15-4.36
2026-08-14-10.60
2026-08-13-11.23
2026-08-12-5.32
2026-08-11-7.66
2026-08-10-8.16

Error distribution

Distribution of the signed daily error in °F for GraphCast (IFS init) at KATL, lead day 9-20 to -18 °F: 0 days-18 to -16 °F: 0 days-16 to -14 °F: 0 days-14 to -12 °F: 2 days-12 to -10 °F: 4 days-10 to -8 °F: 5 days-8 to -6 °F: 4 days-6 to -4 °F: 3 days-4 to -2 °F: 3 days-2 to 0 °F: 3 days0 to 2 °F: 0 days2 to 4 °F: 1 day4 to 6 °F: 1 day6 to 8 °F: 0 days8 to 10 °F: 0 days10 to 12 °F: 1 day12 to 14 °F: 0 days14 to 16 °F: 0 days16 to 18 °F: 0 days18 to 20 °F: 0 daysdashed: P50 |e| 6.78 · P90 |e| 11.59 °F-200+20 °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
-20.00-18.000
-18.00-16.000
-16.00-14.000
-14.00-12.002
-12.00-10.004
-10.00-8.005
-8.00-6.004
-6.00-4.003
-4.00-2.003
-2.00+0.003
+0.00+2.000
+2.00+4.001
+4.00+6.001
+6.00+8.000
+8.00+10.000
+10.00+12.001
+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
2026-08216.02 -5.78
2026-01610.06 -4.67

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 +0.23 [-1.18, 2.44] no
NCEP GFS 21 +1.05 [-0.52, 3.44] no
ECMWF IFS HRES 16 +1.92 [0.70, 4.81] yes
Persistence (baseline) 21 +1.69 [0.85, 3.16] yes

Cite this page

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

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