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

KNYC — New York Central Park

40.7833, -73.9667 · 47 m· standard offset UTC-5 h · truth product CLINYC

These pages are always bilinear interpolation, the headline method. The nearest-node variant, and every other window/initialization/interpolation/variable combination, is published on each permanent link below — follow any number.

Daily maximum temperature — MAE °F

the last 90 days, 12Z, bilinear. MAE with the bias and n underneath; every cell links to its permanent page. The sparkline shows the same model across lead days 1–9 on a scale shared within this table.
Modeld0d1d2d3d4d5d6d7d8d9lead 1–9
ECMWF AIFS Singleaifs_single 1.19-0.03 · n=29 1.64-0.57 · n=28 2.05-0.57 · n=27 1.97-0.24 · n=26 1.82-0.57 · n=24 2.28-0.74 · n=24 1.62-0.74 · n=23 2.12-1.66 · n=22 3.12-2.59 · n=21 ECMWF AIFS Single MAE by lead day 1 to 9, tmax
Aurora (GFS init)aurora_gfs 1.34-1.34 · n=2 0.54-0.54 · n=1 Aurora (GFS init) MAE by lead day 1 to 9, tmax
Aurora (IFS init)aurora_ifs 1.03-1.03 · n=2 0.47+0.47 · n=1 Aurora (IFS init) MAE by lead day 1 to 9, tmax
FourCastNet v2 (GFS init)fourcastnet_gfs 1.54+1.54 · n=2 0.54-0.54 · n=1 FourCastNet v2 (GFS init) MAE by lead day 1 to 9, tmax
FourCastNet v2 (IFS init)fourcastnet_ifs 1.69+0.90 · n=2 0.44+0.44 · n=1 FourCastNet v2 (IFS init) MAE by lead day 1 to 9, tmax
NCEP GFSgfs 3.77+2.99 · n=29 3.77+2.79 · n=28 4.13+2.01 · n=27 3.43+1.96 · n=26 4.55+3.03 · n=25 4.13+2.35 · n=24 5.29+3.61 · n=23 4.30+3.05 · n=22 5.07+2.67 · n=21 NCEP GFS MAE by lead day 1 to 9, tmax
GraphCast (GFS init)graphcast_gfs 1.25-1.25 · n=2 2.96-2.96 · n=1 GraphCast (GFS init) MAE by lead day 1 to 9, tmax
GraphCast (IFS init)graphcast_ifs 1.85-1.47 · n=27 2.38-1.73 · n=26 2.73-2.03 · n=25 2.48-1.27 · n=25 2.96-1.61 · n=25 2.13-1.17 · n=24 2.37-1.42 · n=23 3.85-2.52 · n=22 3.87-2.53 · n=21 GraphCast (IFS init) MAE by lead day 1 to 9, tmax
ECMWF IFS HRESifs_hres 2.23-0.32 · n=24 2.62-0.67 · n=22 3.08-0.32 · n=22 2.83+0.30 · n=21 2.51+1.02 · n=20 2.97+0.64 · n=19 3.65-1.26 · n=17 3.43-0.03 · n=17 3.84-1.34 · n=15 ECMWF IFS HRES MAE by lead day 1 to 9, tmax
Pangu-Weather (GFS init)pangu_gfs 3.94+3.94 · n=2 3.84+3.84 · n=1 Pangu-Weather (GFS init) MAE by lead day 1 to 9, tmax
Pangu-Weather (IFS init)pangu_ifs 3.50+3.50 · n=2 3.85+3.85 · n=1 Pangu-Weather (IFS init) MAE by lead day 1 to 9, tmax
Persistence (baseline)persistence 3.94-0.08 · n=90 5.47-0.09 · n=90 6.43-0.26 · n=90 6.61-0.28 · n=90 6.43-0.34 · n=90 6.44-0.31 · n=90 6.07-0.31 · n=90 6.41-0.39 · n=90 6.76-0.62 · n=90 Persistence (baseline) MAE by lead day 1 to 9, tmax

Up to 1 of the scored days in this table carry a QC flag on the observation (a CLI value that disagreed with the hourly observations, or a fallback source). They are kept in the scores and counted here rather than dropped — see methodology.

+ 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 minimum temperature — MAE °F

the last 90 days, 12Z, bilinear. MAE with the bias and n underneath; every cell links to its permanent page. The sparkline shows the same model across lead days 1–9 on a scale shared within this table.
Modeld0d1d2d3d4d5d6d7d8d9lead 1–9
ECMWF AIFS Singleaifs_single 2.48+2.23 · n=29 2.92+2.70 · n=28 2.89+2.71 · n=27 2.55+2.37 · n=26 2.71+2.32 · n=24 2.77+2.48 · n=24 2.86+2.54 · n=23 2.90+2.61 · n=22 3.19+2.12 · n=21 ECMWF AIFS Single MAE by lead day 1 to 9, tmin
Aurora (GFS init)aurora_gfs 0.77+0.77 · n=2 0.70+0.70 · n=1 Aurora (GFS init) MAE by lead day 1 to 9, tmin
Aurora (IFS init)aurora_ifs 0.77-0.59 · n=2 0.90-0.90 · n=1 Aurora (IFS init) MAE by lead day 1 to 9, tmin
FourCastNet v2 (GFS init)fourcastnet_gfs 1.81+1.18 · n=2 1.50+1.50 · n=1 FourCastNet v2 (GFS init) MAE by lead day 1 to 9, tmin
FourCastNet v2 (IFS init)fourcastnet_ifs 2.04+0.31 · n=2 0.63+0.63 · n=1 FourCastNet v2 (IFS init) MAE by lead day 1 to 9, tmin
NCEP GFSgfs 4.37+4.37 · n=29 4.52+4.52 · n=28 5.06+5.04 · n=27 4.42+4.36 · n=26 4.81+4.44 · n=25 5.61+5.12 · n=24 5.55+4.93 · n=23 5.58+5.01 · n=22 4.52+4.41 · n=21 NCEP GFS MAE by lead day 1 to 9, tmin
GraphCast (GFS init)graphcast_gfs 0.97+0.97 · n=2 0.23+0.23 · n=1 GraphCast (GFS init) MAE by lead day 1 to 9, tmin
GraphCast (IFS init)graphcast_ifs 1.52+0.67 · n=27 1.86+1.12 · n=26 1.82+1.43 · n=25 2.13+1.49 · n=25 2.41+1.20 · n=25 2.83+1.33 · n=24 2.04+1.19 · n=23 2.65+1.13 · n=22 3.57-0.03 · n=21 GraphCast (IFS init) MAE by lead day 1 to 9, tmin
ECMWF IFS HRESifs_hres 1.72+1.46 · n=24 2.14+1.92 · n=22 2.20+2.13 · n=22 3.26+2.69 · n=21 2.24+2.09 · n=20 2.90+2.36 · n=19 2.81+2.58 · n=17 3.45+2.47 · n=17 3.31+1.82 · n=15 ECMWF IFS HRES MAE by lead day 1 to 9, tmin
Pangu-Weather (GFS init)pangu_gfs 2.61-0.53 · n=2 2.16-2.16 · n=1 Pangu-Weather (GFS init) MAE by lead day 1 to 9, tmin
Pangu-Weather (IFS init)pangu_ifs 0.90-0.86 · n=2 1.92-1.92 · n=1 Pangu-Weather (IFS init) MAE by lead day 1 to 9, tmin
Persistence (baseline)persistence 2.69-0.16 · n=90 4.07-0.31 · n=90 4.50-0.50 · n=90 4.51-0.62 · n=90 4.63-0.66 · n=90 4.80-0.60 · n=90 4.93-0.67 · n=90 4.97-0.81 · n=90 5.22-1.00 · n=90 Persistence (baseline) MAE by lead day 1 to 9, tmin

Up to 1 of the scored days in this table carry a QC flag on the observation (a CLI value that disagreed with the hourly observations, or a fallback source). They are kept in the scores and counted here rather than dropped — see methodology.

Month-by-month slices, model availability and the observed-truth table live on the canonical page: they do not depend on the window, initialization or interpolation selected here.

All scores as JSON scores_latest.csv daily_errors.csv.gz data & licences