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

New York Central Park · GraphCast (IFS init) · lead day 0

KNYC · graphcast_ifs · lead 0

Target day = initialization date + 0 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
tmax00Zbilinear30d28 1 QC1.80 [1.38, 2.22]1.49-1.31 [-2.01, -0.58]2.27 [1.76, 2.74]43% [21%, 61%]61%79%——2026-08-01 → 2026-08-30
tmax00Zbilinear90d28 1 QC1.80 [1.38, 2.28]1.49-1.31 [-2.04, -0.42]2.27 [1.72, 2.77]43% [18%, 62%]61%79%——2026-08-01 → 2026-08-30
tmax00Zbilinear365d34 1 QC1.93 [1.50, 2.40]1.54-1.53 [-2.26, -0.65]2.44 [1.91, 2.94]38% [17%, 59%]59%76%——2026-01-01 → 2026-08-30
tmax00Zbilinearall34 1 QC1.93 [1.48, 2.38]1.54-1.53 [-2.24, -0.70]2.44 [1.88, 2.92]38% [17%, 57%]59%76%——2026-01-01 → 2026-08-30
tmax00Znearest30d28 1 QC1.78 [1.35, 2.19]1.52-1.22 [-1.93, -0.48]2.26 [1.75, 2.71]46% [29%, 62%]57%79%——2026-08-01 → 2026-08-30
tmax00Znearest90d28 1 QC1.78 [1.35, 2.24]1.52-1.22 [-1.97, -0.31]2.26 [1.71, 2.74]46% [26%, 63%]57%79%——2026-08-01 → 2026-08-30
tmax00Znearest365d34 1 QC1.88 [1.46, 2.32]1.56-1.42 [-2.17, -0.55]2.40 [1.90, 2.87]44% [28%, 61%]56%76%——2026-01-01 → 2026-08-30
tmax00Znearestall34 1 QC1.88 [1.45, 2.31]1.56-1.42 [-2.12, -0.59]2.40 [1.87, 2.85]44% [26%, 60%]56%76%——2026-01-01 → 2026-08-30
tmin00Zbilinear30d28 1 QC1.82 [1.41, 2.28]1.50+1.34 [0.65, 1.91]2.40 [1.81, 2.98]36% [25%, 47%]68%82%——2026-08-01 → 2026-08-30
tmin00Zbilinear90d28 1 QC1.82 [1.41, 2.36]1.50+1.34 [0.50, 1.85]2.40 [1.74, 3.04]36% [24%, 45%]68%82%——2026-08-01 → 2026-08-30
tmin00Zbilinear365d34 1 QC2.12 [1.52, 2.85]2.03+0.49 [-1.61, 1.69]2.80 [1.93, 3.63]32% [20%, 43%]62%76%——2026-01-01 → 2026-08-30
tmin00Zbilinearall34 1 QC2.12 [1.50, 2.85]2.03+0.49 [-1.45, 1.72]2.80 [1.94, 3.65]32% [21%, 43%]62%76%——2026-01-01 → 2026-08-30
tmin00Znearest30d28 1 QC1.95 [1.54, 2.42]1.51+1.55 [0.87, 2.12]2.51 [1.94, 3.09]29% [19%, 39%]61%79%——2026-08-01 → 2026-08-30
tmin00Znearest90d28 1 QC1.95 [1.50, 2.45]1.51+1.55 [0.76, 2.06]2.51 [1.82, 3.14]29% [19%, 38%]61%79%——2026-08-01 → 2026-08-30
tmin00Znearest365d34 1 QC2.19 [1.62, 2.85]2.04+0.70 [-1.36, 1.89]2.85 [2.02, 3.60]29% [20%, 39%]59%74%——2026-01-01 → 2026-08-30
tmin00Znearestall34 1 QC2.19 [1.60, 2.84]2.04+0.70 [-1.21, 1.90]2.85 [2.03, 3.65]29% [20%, 38%]59%74%——2026-01-01 → 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 KNYC, lead day 0, 00Z bilinear tmax+10+50-5-102026-08-01: -3.37 °F2026-08-02: -0.99 °F2026-08-03: -0.25 °F2026-08-04: -2.99 °F2026-08-05: -3.91 °F2026-08-06: -0.62 °F2026-08-07: -2.22 °F2026-08-08: +0.85 °F2026-08-09: +0.85 °F2026-08-10: +2.5 °F2026-08-11: -0.29 °F2026-08-12: -0.78 °F2026-08-13: -0.5 °F2026-08-14: +0.19 °F2026-08-15: -2.99 °F2026-08-16: -2.12 °F2026-08-17: +0.57 °F2026-08-18: -5.74 °F2026-08-19: -0.84 °F2026-08-20: -3.46 °F2026-08-21: -3.27 °F2026-08-22: -1.1 °F2026-08-23: -1.19 °F2026-08-24: -1.56 °F2026-08-25: -1.19 °F2026-08-28: -3.61 °F2026-08-29: -0.62 °F2026-08-30: +1.93 °F2026-08-012026-08-30°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 (30 days, newest first) · download CSV
Climatological dayError °F
2026-08-30+1.93
2026-08-29-0.62
2026-08-28-3.61
2026-08-27—
2026-08-26—
2026-08-25-1.19
2026-08-24-1.56
2026-08-23-1.19
2026-08-22-1.10
2026-08-21-3.27
2026-08-20-3.46
2026-08-19-0.84
2026-08-18-5.74
2026-08-17+0.57
2026-08-16-2.12
2026-08-15-2.99
2026-08-14+0.19
2026-08-13-0.50
2026-08-12-0.78
2026-08-11-0.29
2026-08-10+2.50
2026-08-09+0.85
2026-08-08+0.85
2026-08-07-2.22
2026-08-06-0.62
2026-08-05-3.91
2026-08-04-2.99
2026-08-03-0.25
2026-08-02-0.99
2026-08-01-3.37

Error distribution

Distribution of the signed daily error in °F for GraphCast (IFS init) at KNYC, lead day 0-10 to -9 °F: 0 days-9 to -8 °F: 0 days-8 to -7 °F: 0 days-7 to -6 °F: 0 days-6 to -5 °F: 2 days-5 to -4 °F: 1 day-4 to -3 °F: 5 days-3 to -2 °F: 5 days-2 to -1 °F: 6 days-1 to 0 °F: 9 days0 to 1 °F: 4 days1 to 2 °F: 1 day2 to 3 °F: 1 day3 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| 1.21 · P90 |e| 3.82 °F-100+10 °Fn = 34 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
-7.00-6.000
-6.00-5.002
-5.00-4.001
-4.00-3.005
-3.00-2.005
-2.00-1.006
-1.00+0.009
+0.00+1.004
+1.00+2.001
+2.00+3.001
+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
2026-08281.80 -1.31
2026-0162.53 -2.53

Availability

ModelScored period DaysCoverage
GraphCast (IFS init)2026-01-01 → 2026-08-30 34
GraphCast (IFS init)2026-01-01 → 2026-08-30 34

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 28 +0.37 [-0.15, 0.80] no
Aurora (GFS init) 2 +0.32 [0.17, 0.47] yes
Aurora (IFS init) 3 -0.13 [-0.85, 0.75] no
FourCastNet v2 (GFS init) 2 -0.42 [-1.60, 0.77] no
FourCastNet v2 (IFS init) 3 -0.84 [-2.62, 1.18] no
NCEP GFS 28 -1.52 [-3.00, -0.13] yes
GraphCast (GFS init) 2 +0.30 [-0.16, 0.77] no
ECMWF IFS HRES 22 -0.69 [-2.45, 0.60] no
Pangu-Weather (GFS init) 2 -2.53 [-2.82, -2.25] yes
Pangu-Weather (IFS init) 3 -1.15 [-3.17, 1.89] no

Cite this page

ShortCastCheck, KNYC graphcast_ifs lead 0, methodology v0.2, accessed 2026-08-31, https://castcheck.zifanzhang.com/station/KNYC/model/graphcast_ifs/lead/0/

LongCastCheck (2026). Station-level verification of raw weather-model 2 m temperature forecasts: KNYC (New York Central Park), model graphcast_ifs, lead day 0. Methodology version 0.2, schema version 0.1, data through 2026-08-30. Accessed 2026-08-31. https://castcheck.zifanzhang.com/station/KNYC/model/graphcast_ifs/lead/0/

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