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

Philadelphia Intl · GraphCast (IFS init) · lead day 1

KPHL · graphcast_ifs · lead 1

Target day = initialization date + 1 day. 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
tmax00Zbilinear30d27 1 QC5.16 [4.47, 6.12]1.45-5.16 [-6.12, -4.47]5.45 [4.76, 6.36]0% [0%, 0%]4%7%-0.62+0.542026-08-02 → 2026-08-30
tmax00Zbilinear90d27 1 QC5.16 [4.42, 6.17]1.45-5.16 [-6.17, -4.42]5.45 [4.72, 6.47]0% [0%, 0%]4%7%-0.62+0.542026-08-02 → 2026-08-30
tmax00Zbilinear365d33 1 QC4.65 [3.49, 5.72]1.69-4.60 [-5.70, -3.36]5.04 [3.89, 6.03]3% [0%, 11%]6%18%-0.33+0.522026-01-02 → 2026-08-30
tmax00Zbilinearall33 1 QC4.65 [3.63, 5.83]1.69-4.60 [-5.83, -3.50]5.04 [4.05, 6.14]3% [0%, 11%]6%18%-0.33+0.522026-01-02 → 2026-08-30
tmax00Znearest30d27 1 QC4.85 [4.16, 5.78]1.42-4.85 [-5.78, -4.16]5.16 [4.47, 6.03]0% [0%, 0%]7%15%-0.52+0.552026-08-02 → 2026-08-30
tmax00Znearest90d27 1 QC4.85 [4.12, 5.87]1.42-4.85 [-5.87, -4.12]5.16 [4.41, 6.15]0% [0%, 0%]7%15%-0.52+0.552026-08-02 → 2026-08-30
tmax00Znearest365d33 1 QC4.34 [3.18, 5.38]1.72-4.26 [-5.37, -2.91]4.75 [3.56, 5.74]0% [0%, 0%]18%30%-0.24+0.512026-01-02 → 2026-08-30
tmax00Znearestall33 1 QC4.34 [3.28, 5.50]1.72-4.26 [-5.50, -3.09]4.75 [3.74, 5.84]0% [0%, 0%]18%30%-0.24+0.512026-01-02 → 2026-08-30
tmax12Zbilinear30d27 1 QC4.92 [4.29, 5.80]1.24-4.92 [-5.80, -4.29]5.17 [4.53, 5.98]0% [0%, 0%]4%7%-0.55+0.612026-08-02 → 2026-08-30
tmax12Zbilinear90d27 1 QC4.92 [4.26, 5.86]1.24-4.92 [-5.86, -4.26]5.17 [4.51, 5.99]0% [0%, 0%]4%7%-0.55+0.612026-08-02 → 2026-08-30
tmax12Zbilinear365d32 1 QC4.50 [3.46, 5.36]1.47-4.50 [-5.36, -3.46]4.84 [3.85, 5.64]3% [0%, 12%]9%19%-0.40+0.552026-01-02 → 2026-08-30
tmax12Zbilinearall32 1 QC4.50 [3.56, 5.51]1.47-4.50 [-5.51, -3.56]4.84 [3.98, 5.81]3% [0%, 12%]9%19%-0.40+0.552026-01-02 → 2026-08-30
tmax12Znearest30d27 1 QC4.65 [4.03, 5.49]1.24-4.65 [-5.49, -4.03]4.90 [4.30, 5.67]4% [0%, 11%]4%11%-0.46+0.612026-08-02 → 2026-08-30
tmax12Znearest90d27 1 QC4.65 [3.98, 5.52]1.24-4.65 [-5.52, -3.98]4.90 [4.27, 5.70]4% [0%, 11%]4%11%-0.46+0.612026-08-02 → 2026-08-30
tmax12Znearest365d32 1 QC4.20 [3.08, 5.06]1.49-4.19 [-5.05, -3.06]4.57 [3.55, 5.34]6% [0%, 15%]9%22%-0.30+0.542026-01-02 → 2026-08-30
tmax12Znearestall32 1 QC4.20 [3.21, 5.24]1.49-4.19 [-5.24, -3.20]4.57 [3.68, 5.52]6% [0%, 15%]9%22%-0.30+0.542026-01-02 → 2026-08-30
tmin00Zbilinear30d27 1 QC2.05 [1.60, 2.37]1.98-0.91 [-1.44, -0.34]2.44 [2.03, 2.73]30% [18%, 46%]48%78%+0.26+0.282026-08-02 → 2026-08-30
tmin00Zbilinear90d27 1 QC2.05 [1.52, 2.35]1.98-0.91 [-1.44, -0.35]2.44 [1.91, 2.71]30% [18%, 50%]48%78%+0.26+0.282026-08-02 → 2026-08-30
tmin00Zbilinear365d33 1 QC1.89 [1.28, 2.24]1.78-0.91 [-1.39, -0.39]2.29 [1.66, 2.61]36% [23%, 60%]52%82%+0.36+0.392026-01-02 → 2026-08-30
tmin00Zbilinearall33 1 QC1.89 [1.33, 2.27]1.78-0.91 [-1.36, -0.45]2.29 [1.66, 2.63]36% [22%, 60%]52%82%+0.36+0.392026-01-02 → 2026-08-30
tmin00Znearest30d27 1 QC2.00 [1.57, 2.29]1.98-0.52 [-1.08, 0.06]2.32 [1.94, 2.57]22% [10%, 42%]56%78%+0.28+0.282026-08-02 → 2026-08-30
tmin00Znearest90d27 1 QC2.00 [1.48, 2.29]1.98-0.52 [-1.07, 0.08]2.32 [1.85, 2.56]22% [9%, 47%]56%78%+0.28+0.282026-08-02 → 2026-08-30
tmin00Znearest365d33 1 QC1.82 [1.19, 2.19]1.78-0.54 [-1.02, -0.01]2.16 [1.54, 2.47]27% [12%, 52%]61%82%+0.38+0.402026-01-02 → 2026-08-30
tmin00Znearestall33 1 QC1.82 [1.26, 2.19]1.78-0.54 [-0.99, -0.08]2.16 [1.58, 2.48]27% [13%, 50%]61%82%+0.38+0.402026-01-02 → 2026-08-30
tmin12Zbilinear30d27 1 QC1.89 [1.59, 2.16]1.62-0.86 [-1.34, -0.41]2.13 [1.79, 2.38]26% [14%, 41%]52%81%+0.32+0.412026-08-02 → 2026-08-30
tmin12Zbilinear90d27 1 QC1.89 [1.59, 2.15]1.62-0.86 [-1.40, -0.39]2.13 [1.79, 2.39]26% [12%, 44%]52%81%+0.32+0.412026-08-02 → 2026-08-30
tmin12Zbilinear365d32 1 QC1.78 [1.38, 2.06]1.45-0.90 [-1.37, -0.47]2.02 [1.55, 2.30]28% [15%, 44%]59%84%+0.37+0.492026-01-02 → 2026-08-30
tmin12Zbilinearall32 1 QC1.78 [1.41, 2.09]1.45-0.90 [-1.35, -0.51]2.02 [1.60, 2.31]28% [14%, 43%]59%84%+0.37+0.492026-01-02 → 2026-08-30
tmin12Znearest30d27 1 QC1.75 [1.47, 2.00]1.61-0.47 [-0.93, -0.06]1.97 [1.70, 2.18]26% [13%, 43%]67%85%+0.37+0.422026-08-02 → 2026-08-30
tmin12Znearest90d27 1 QC1.75 [1.42, 2.00]1.61-0.47 [-0.98, -0.05]1.97 [1.66, 2.19]26% [14%, 45%]67%85%+0.37+0.422026-08-02 → 2026-08-30
tmin12Znearest365d32 1 QC1.63 [1.17, 1.90]1.43-0.54 [-1.01, -0.13]1.86 [1.39, 2.12]31% [17%, 55%]72%88%+0.42+0.492026-01-02 → 2026-08-30
tmin12Znearestall32 1 QC1.63 [1.24, 1.93]1.43-0.54 [-0.99, -0.17]1.86 [1.46, 2.13]31% [16%, 52%]72%88%+0.42+0.492026-01-02 → 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 KPHL, lead day 1, 00Z bilinear tmax+10+50-5-102026-08-02: -8.72 °F2026-08-03: -7.16 °F2026-08-04: -8.46 °F2026-08-05: -5.28 °F2026-08-06: -5.66 °F2026-08-07: -5.47 °F2026-08-08: -3.11 °F2026-08-09: -3.63 °F2026-08-10: -1.66 °F2026-08-11: -2.08 °F2026-08-12: -7.13 °F2026-08-13: -3.93 °F2026-08-14: -6.23 °F2026-08-15: -7.61 °F2026-08-16: -4.53 °F2026-08-17: -4.64 °F2026-08-18: -3.79 °F2026-08-19: -5.95 °F2026-08-20: -5.88 °F2026-08-21: -5.9 °F2026-08-22: -3.06 °F2026-08-23: -3.74 °F2026-08-24: -5.43 °F2026-08-25: -3.87 °F2026-08-26: -6.01 °F2026-08-29: -4.21 °F2026-08-30: -6.09 °F2026-08-022026-08-30°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 (29 days, newest first) · download CSV
Climatological dayError °F
2026-08-30-6.09
2026-08-29-4.21
2026-08-28—
2026-08-27—
2026-08-26-6.01
2026-08-25-3.87
2026-08-24-5.43
2026-08-23-3.74
2026-08-22-3.06
2026-08-21-5.90
2026-08-20-5.88
2026-08-19-5.95
2026-08-18-3.79
2026-08-17-4.64
2026-08-16-4.53
2026-08-15-7.61
2026-08-14-6.23
2026-08-13-3.93
2026-08-12-7.13
2026-08-11-2.08
2026-08-10-1.66
2026-08-09-3.63
2026-08-08-3.11
2026-08-07-5.47
2026-08-06-5.66
2026-08-05-5.28
2026-08-04-8.46
2026-08-03-7.16
2026-08-02-8.72

Error distribution

Distribution of the signed daily error in °F for GraphCast (IFS init) at KPHL, lead day 1-10 to -9 °F: 0 days-9 to -8 °F: 2 days-8 to -7 °F: 3 days-7 to -6 °F: 3 days-6 to -5 °F: 7 days-5 to -4 °F: 3 days-4 to -3 °F: 9 days-3 to -2 °F: 4 days-2 to -1 °F: 1 day-1 to 0 °F: 0 days0 to 1 °F: 1 day1 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| 4.53 · P90 |e| 7.15 °F-100+10 °Fn = 33 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
-9.00-8.002
-8.00-7.003
-7.00-6.003
-6.00-5.007
-5.00-4.003
-4.00-3.009
-3.00-2.004
-2.00-1.001
-1.00+0.000
+0.00+1.001
+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
2026-08275.16 -5.16
2026-0162.37 -2.12

Availability

ModelScored period DaysCoverage
GraphCast (IFS init)2026-01-02 → 2026-08-30 33
GraphCast (IFS init)2026-01-02 → 2026-08-30 33
GraphCast (IFS init)2026-01-02 → 2026-08-30 32
GraphCast (IFS init)2026-01-02 → 2026-08-30 32

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 27 +2.76 [2.42, 3.15] yes
Aurora (GFS init) 1 -0.42 [-0.42, -0.42] yes
Aurora (IFS init) 2 +0.96 [0.67, 1.25] yes
FourCastNet v2 (GFS init) 1 +1.37 [1.37, 1.37] yes
FourCastNet v2 (IFS init) 2 +2.44 [0.84, 4.04] yes
NCEP GFS 27 +2.10 [1.33, 3.26] yes
GraphCast (GFS init) 1 -0.26 [-0.26, -0.26] yes
ECMWF IFS HRES 21 +1.18 [-0.04, 2.70] no
Pangu-Weather (GFS init) 1 +3.56 [3.56, 3.56] yes
Pangu-Weather (IFS init) 2 +3.49 [2.78, 4.20] yes
Persistence (baseline) 27 +1.97 [0.87, 3.39] yes

Cite this page

ShortCastCheck, KPHL graphcast_ifs lead 1, methodology v0.2, accessed 2026-08-31, https://castcheck.zifanzhang.com/station/KPHL/model/graphcast_ifs/lead/1/

LongCastCheck (2026). Station-level verification of raw weather-model 2 m temperature forecasts: KPHL (Philadelphia Intl), model graphcast_ifs, lead day 1. Methodology version 0.2, schema version 0.1, data through 2026-08-30. Accessed 2026-08-31. https://castcheck.zifanzhang.com/station/KPHL/model/graphcast_ifs/lead/1/

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