CastCheck v0.3.1 · data through 2026-08-30

Daily verification · methodology v0.3.1

How far off was each weather model?

Over the last 365 days, across 23 U.S. stations, the most accurate raw instantaneous temperature forecast 1 day ahead is GraphCast (IFS init): mean absolute error 2.95 °F [2.88, 3.01], bias −1.90 °F, on n = 197 scored days.

Scope: raw model output on the native 0.25° grid. The headline variable is the instantaneous 2 m temperature at 00/06/12/18 UTC, scored against the observed 2 m temperature at the same instant (ASOS routine METAR) — the same four instants on both sides, so nothing in the number depends on a model's own diurnal amplitude. Errors are computed in °C and shown in °F (methodology v0.3.1).

Window
the last 365 days
Variable
instantaneous temperature
Init
12Z
Interpolation
nearest
Data through
2026-08-30
Updated
Next update

Leader, lead day 1

2.95°F

[2.88, 3.01] MAE

GraphCast (IFS init)

Persistence baseline

5.08°F

[4.51, 5.63]

yesterday's observation — the bar every model has to clear

Scored days

197

23 stations · 12 systems

in the last 365 days, lead day 1

Systems ranked

2/12

next update

below n = 30: published, greyed, unranked

Mean absolute error with 95 % confidence intervals

Lead day 1 · the last 365 days · 23 stations pooled · 12Z · nearest · instantaneous temperature. Whiskers are moving-block bootstrap intervals; models whose whiskers overlap are not distinguishable at this sample size. Baselines are drawn in grey and never ranked.

MAE by model with 95 % confidence intervals, lead day 1, 365d, 12Z, nearest0246810mean absolute error, °FGraphCast (IFS init)GraphCast (IFS init): 2.95 °F2.95GraphCast (GFS init)GraphCast (GFS init): 2.97 °F2.97ECMWF AIFS SingleECMWF AIFS Single: 2.49 °F2.49FourCastNet v2 (GFS init)FourCastNet v2 (GFS init): 2.55 °F2.55Pangu-Weather (IFS init)Pangu-Weather (IFS init): 2.58 °F2.58FourCastNet v2 (IFS init)FourCastNet v2 (IFS init): 2.6 °F2.6ECMWF IFS HRESECMWF IFS HRES: 2.73 °F2.73Pangu-Weather (GFS init)Pangu-Weather (GFS init): 2.82 °F2.82Aurora (IFS init)Aurora (IFS init): 3.07 °F3.07Aurora (GFS init)Aurora (GFS init): 3.21 °F3.21NCEP GFSNCEP GFS: 3.23 °F3.23Persistence (baseline)Persistence (baseline): 5.08 °F5.08

Source: CastCheck 0.1.0, methodology v0.3.1 · ECMWF open data, NOAA/NCEP GFS, NOAA/CIRA AIWP, NWS ASOS · data through 2026-08-30 · CC BY 4.0. Hover a bar for its value.

Lead day 1target = init + 1 d

permalink

The last 365 days, all 23 stations pooled, 12Z initialization, nearest interpolation, instantaneous temperature against the observed 2 m temperature at the same instant (ASOS routine METAR). Lower MAE is better; bias is positive when the model is too warm. Every number links to its permanent page.

Skill is 1 − MAE ÷ MAE(persistence) computed on the days both have a value — the small print in that column is that denominator and the size of the intersection, so the number can be checked against the persistence row rather than contradicting it. The persistence row's own n is its whole record, marked all days. Skill, debiased (out-of-sample) repeats the score after removing a bias estimated on the 30 scored days before each day and applied forward, never on the day itself. An interval reads — when the bootstrap could not run: fewer than 28 scored days or fewer than 4 blocks. Hit-rate intervals are Wilson score intervals, not bootstraps.

Lead day 1, the last 365 days, 12Z, nearest, instantaneous temperature.
RankModelMAE °FBias °F±3 °F SkillSkill, debiased (out-of-sample) nvs leader
1GraphCast (IFS init)graphcast_ifs2.95[2.88, 3.01]−1.90[−2.00, −1.80]59% +0.42[+0.34, +0.49]vs 5.08 (n=197)+0.66n=182 19722.8 stns · 14 QC★ lowest MAE in this view
2GraphCast (GFS init)graphcast_gfs2.97[2.75, 3.20]−1.61[−1.94, −1.29]59% +0.52[+0.41, +0.59]vs 6.13 (n=36)+0.66n=21 3622.6 stns= not distinguishable from the leader after the Holm correction within this table
—ECMWF AIFS Singleaifs_single2.49[2.35, 2.63]−0.90[−1.01, −0.77]68% +0.19[+0.14, +0.24]vs 3.06 (n=29)+0.41n=14 2922.7 stns · 2 QC
—FourCastNet v2 (GFS init)fourcastnet_gfs2.55—−0.39—69% ——vs 3.51 (n=2)—n=0 219 stns
—Pangu-Weather (IFS init)pangu_ifs2.58—−0.54—66% ——vs 3.51 (n=2)—n=0 219 stns
—FourCastNet v2 (IFS init)fourcastnet_ifs2.60—−0.71—68% ——vs 3.51 (n=2)—n=0 219 stns
—ECMWF IFS HRESifs_hres2.73[2.58, 2.85]−0.54[−0.65, −0.42]66% +0.11[+0.03, +0.18]vs 3.06 (n=29)+0.32n=14 2922.7 stns · 2 QC
—Pangu-Weather (GFS init)pangu_gfs2.82—−0.38—59% ——vs 3.51 (n=2)—n=0 219 stns
—Aurora (IFS init)aurora_ifs3.07—−2.22—51% ——vs 3.51 (n=2)—n=0 219 stns
—Aurora (GFS init)aurora_gfs3.21—−2.21—54% ——vs 3.51 (n=2)—n=0 219 stns
—NCEP GFSgfs3.23[3.10, 3.36]+0.79[0.68, 0.93]55% −0.06[−0.13, +0.01]vs 3.06 (n=29)+0.33n=14 2922.7 stns · 2 QC
—Persistence (baseline)persistence · baseline5.08[4.51, 5.63]−0.15[−0.41, 0.12]47% —— 215all days · 14 QC

+ model too warm − model too cold bias interval includes zero ★ lowest MAE = not distinguishable from the leader ▼ worse than the leader ▲ better than the leader — all three Holm-corrected within this table, over the family of comparisons against the leader; the uncorrected verdict for every pair is published in the pairwise table of each permanent link and in pairwise_latest.csv.gzn < 30 greyed and unranked CI is a 95 % moving-block bootstrap interval on the group's own days, and reads — when it could not be computed skill reads — when fewer than 10 days are common to the model and the baseline: the ratio of two means over a handful of shared days is not a number worth printing

Lead day 3target = init + 3 d

permalink

Same sampling as lead day 1. Skill is 1 − MAE ÷ MAE(persistence) on the days both have a value.

Lead day 3, the last 365 days, 12Z, nearest, instantaneous temperature.
RankModelMAE °FBias °FSkillnvs leader
1GraphCast (IFS init)graphcast_ifs3.19[3.11, 3.29]−1.89[−2.03, −1.76]+0.57[+0.51, +0.62]vs 7.38 (n=195)19522.8 stns · 11 QC★ lowest MAE in this view
2GraphCast (GFS init)graphcast_gfs3.61[3.37, 3.81]−2.13[−2.69, −1.53]+0.60[+0.53, +0.66]vs 9.07 (n=34)3422.7 stns · 2 QC= not distinguishable from the leader after the Holm correction within this table
—ECMWF AIFS Singleaifs_single2.66—−0.87—+0.31—vs 3.85 (n=27)2722.6 stns · 2 QC
—ECMWF IFS HRESifs_hres2.98—−0.21—+0.23—vs 3.85 (n=27)2722.6 stns · 2 QC
—NCEP GFSgfs3.59—+1.03—+0.07—vs 3.85 (n=27)2722.6 stns · 2 QC
—Persistence (baseline)persistence · baseline7.43[6.54, 8.31]−0.42[−1.08, 0.29]—215all days · 14 QC

Lead day 5target = init + 5 d

permalink

Same sampling as lead day 1. Skill is 1 − MAE ÷ MAE(persistence) on the days both have a value.

Lead day 5, the last 365 days, 12Z, nearest, instantaneous temperature.
RankModelMAE °FBias °FSkillnvs leader
1GraphCast (IFS init)graphcast_ifs3.73[3.54, 3.93]−1.82[−2.05, −1.61]+0.52[+0.46, +0.57]vs 7.76 (n=195)19522.8 stns · 11 QC★ lowest MAE in this view
2GraphCast (GFS init)graphcast_gfs4.58[4.33, 4.84]−1.47[−2.30, −0.67]+0.55[+0.50, +0.59]vs 10.17 (n=34)3422.7 stns · 2 QC= not distinguishable from the leader after the Holm correction within this table
—ECMWF AIFS Singleaifs_single2.90—−0.95—+0.28—vs 4.00 (n=25)2522.6 stns · 2 QC
—ECMWF IFS HRESifs_hres3.26—+0.01—+0.19—vs 4.00 (n=25)2522.6 stns · 2 QC
—NCEP GFSgfs3.81—+0.63—+0.05—vs 4.00 (n=25)2522.6 stns · 2 QC
—Persistence (baseline)persistence · baseline7.83[6.92, 8.71]−0.66[−1.66, 0.35]—215all days · 13 QC

Lead day 7target = init + 7 d

permalink

Same sampling as lead day 1. Skill is 1 − MAE ÷ MAE(persistence) on the days both have a value.

Lead day 7, the last 365 days, 12Z, nearest, instantaneous temperature.
RankModelMAE °FBias °FSkillnvs leader
1GraphCast (IFS init)graphcast_ifs4.70[4.30, 5.10]−1.75[−2.15, −1.31]+0.44[+0.39, +0.48]vs 8.38 (n=193)19322.7 stns · 13 QC★ lowest MAE in this view
2GraphCast (GFS init)graphcast_gfs6.13[5.24, 7.19]−0.41[−2.11, 1.39]+0.43[+0.35, +0.50]vs 10.82 (n=34)3422.7 stns · 2 QC= not distinguishable from the leader after the Holm correction within this table
—ECMWF AIFS Singleaifs_single3.14—−1.22—+0.26—vs 4.23 (n=23)2322.6 stns · 2 QC
—ECMWF IFS HRESifs_hres3.44—−0.06—+0.19—vs 4.23 (n=23)2322.6 stns · 2 QC
—NCEP GFSgfs4.11—+0.55—+0.03—vs 4.23 (n=23)2322.6 stns · 2 QC
—Persistence (baseline)persistence · baseline8.41[7.39, 9.34]−0.81[−2.03, 0.36]—215all days · 14 QC

The same four samples as a daily maximum and minimum

Lead day 1, the last 365 days, 12Z, nearest. Here the forecast's max/min of the four samples is scored against the observation's max/min of the same four instants. Like for like: whatever the four-sample definition misses, it misses on both sides. The comparison against the true NWS daily extremes is a different question and lives on the station and permanent-link pages.

Sampled daily maximumtmax_s

RankModelMAE °FBias °FSkillnvs leader
1GraphCast (GFS init)graphcast_gfs2.95[2.69, 3.19]−2.25[−2.46, −2.03]+0.47vs 5.59 (n=36)36★
2GraphCast (IFS init)graphcast_ifs3.01[2.88, 3.12]−2.12[−2.27, −1.97]+0.40vs 5.04 (n=197)197=
—ECMWF AIFS Singleaifs_single2.27[2.23, 2.32]−0.29[−0.42, −0.15]+0.19vs 2.80 (n=29)29
—ECMWF IFS HRESifs_hres2.55[2.41, 2.68]−0.44[−0.61, −0.24]+0.09vs 2.80 (n=29)29
—Pangu-Weather (IFS init)pangu_ifs2.56—−0.50——vs 3.48 (n=2)2
—FourCastNet v2 (GFS init)fourcastnet_gfs2.60—−1.11——vs 3.48 (n=2)2
—FourCastNet v2 (IFS init)fourcastnet_ifs2.62—−1.32——vs 3.48 (n=2)2
—Pangu-Weather (GFS init)pangu_gfs3.06—−0.61——vs 3.48 (n=2)2
—Aurora (IFS init)aurora_ifs3.50—−3.28——vs 3.48 (n=2)2
—NCEP GFSgfs3.58[3.37, 3.76]+1.93[1.60, 2.28]−0.28vs 2.80 (n=29)29
—Aurora (GFS init)aurora_gfs4.30—−4.08——vs 3.48 (n=2)2
—Persistence (baseline)persistence5.04[4.49, 5.58]−0.13[−0.41, 0.18]—215all days

Sampled daily minimumtmin_s

RankModelMAE °FBias °FSkillnvs leader
1GraphCast (IFS init)graphcast_ifs2.57[2.46, 2.68]−1.27[−1.39, −1.14]+0.41vs 4.37 (n=197)197★
2GraphCast (GFS init)graphcast_gfs2.99[2.75, 3.26]−0.51[−0.93, −0.00]+0.51vs 6.10 (n=36)36=
—FourCastNet v2 (GFS init)fourcastnet_gfs2.11—+0.34——vs 2.54 (n=2)2
—FourCastNet v2 (IFS init)fourcastnet_ifs2.11—+0.02——vs 2.54 (n=2)2
—Aurora (GFS init)aurora_gfs2.14—−0.18——vs 2.54 (n=2)2
—Aurora (IFS init)aurora_ifs2.23—−0.87——vs 2.54 (n=2)2
—ECMWF IFS HRESifs_hres2.24[2.10, 2.39]−0.45[−0.68, −0.23]+0.04vs 2.34 (n=29)29
—Pangu-Weather (GFS init)pangu_gfs2.29—+0.27——vs 2.54 (n=2)2
—ECMWF AIFS Singleaifs_single2.34[2.13, 2.52]−1.35[−1.58, −1.14]−0.00vs 2.34 (n=29)29
—Pangu-Weather (IFS init)pangu_ifs2.35—−0.97——vs 2.54 (n=2)2
—NCEP GFSgfs2.52[2.38, 2.69]+0.33[0.07, 0.61]−0.08vs 2.34 (n=29)29
—Persistence (baseline)persistence4.37[3.79, 4.92]−0.17[−0.46, 0.11]—215all days

Every model × every lead day

MAE in °F with the bias and n underneath, the last 365 days, 12Z, nearest, instantaneous temperature. The sparkline is the same model's MAE across lead days 1–9 on a shared vertical scale.

Modeld1d2d3d4d5d6d7d8d9 lead 1–9
ECMWF AIFS Singleaifs_single2.49−0.90 · n=292.57−0.90 · n=282.66−0.87 · n=272.77−0.88 · n=262.90−0.95 · n=253.12−1.03 · n=243.14−1.22 · n=233.42−1.57 · n=223.81−1.97 · n=21ECMWF AIFS Single MAE by lead day 1 to 9
Aurora (GFS init)aurora_gfs3.21−2.21 · n=23.71−2.59 · n=1——————— Aurora (GFS init) MAE by lead day 1 to 9
Aurora (IFS init)aurora_ifs3.07−2.22 · n=23.29−2.15 · n=1——————— Aurora (IFS init) MAE by lead day 1 to 9
FourCastNet v2 (GFS init)fourcastnet_gfs2.55−0.39 · n=22.52−0.34 · n=1——————— FourCastNet v2 (GFS init) MAE by lead day 1 to 9
FourCastNet v2 (IFS init)fourcastnet_ifs2.60−0.71 · n=22.33−0.31 · n=1——————— FourCastNet v2 (IFS init) MAE by lead day 1 to 9
NCEP GFSgfs3.23+0.79 · n=293.36+0.87 · n=283.59+1.03 · n=273.70+0.86 · n=263.81+0.63 · n=254.03+0.44 · n=244.11+0.55 · n=234.43+0.42 · n=224.70+0.26 · n=21NCEP GFS MAE by lead day 1 to 9
GraphCast (GFS init)graphcast_gfs2.97−1.61 · n=363.26−1.95 · n=353.61−2.13 · n=344.05−2.02 · n=344.58−1.47 · n=345.31−1.01 · n=346.13−0.41 · n=346.79+0.12 · n=347.19+0.54 · n=34GraphCast (GFS init) MAE by lead day 1 to 9
GraphCast (IFS init)graphcast_ifs2.95−1.90 · n=1973.08−1.93 · n=1963.19−1.89 · n=1953.39−1.86 · n=1953.73−1.82 · n=1954.15−1.80 · n=1944.70−1.75 · n=1935.20−1.84 · n=1925.80−1.90 · n=191GraphCast (IFS init) MAE by lead day 1 to 9
ECMWF IFS HRESifs_hres2.73−0.54 · n=292.89−0.28 · n=282.98−0.21 · n=273.14−0.06 · n=263.26+0.01 · n=253.39+0.12 · n=243.44−0.06 · n=233.66−0.11 · n=223.96−0.48 · n=21ECMWF IFS HRES MAE by lead day 1 to 9
Pangu-Weather (GFS init)pangu_gfs2.82−0.38 · n=22.81−0.58 · n=1——————— Pangu-Weather (GFS init) MAE by lead day 1 to 9
Pangu-Weather (IFS init)pangu_ifs2.58−0.54 · n=22.53−0.58 · n=1——————— Pangu-Weather (IFS init) MAE by lead day 1 to 9
Persistence (baseline)persistence5.08−0.15 · n=2156.86−0.29 · n=2157.43−0.42 · n=2157.54−0.53 · n=2157.83−0.66 · n=2158.15−0.75 · n=2158.41−0.81 · n=2158.55−0.89 · n=2158.63−0.98 · n=215Persistence (baseline) MAE by lead day 1 to 9

Where the errors are

Both maps show a fixed quantity per station. Neither shows which model wins where: on samples this short the per-station winner is mostly noise, and a map of winners would invite a comparison the data cannot carry.

ECMWF IFS HRES bias

Mean bias of the fixed reference model ECMWF IFS HRES (lead day 1, instantaneous temperature, 365d window, 12Z, nearest). One model everywhere, so the colours compare stations, not models. Dot area grows with the number of scored days; fill is the mean bias (warm = model too warm, cool = too cold, grey = interval includes zero). The frame is a latitude/longitude graticule on an Albers projection.

Mean bias in °F of ECMWF IFS HRES, lead day 1, instantaneous temperature, 365d window, 12Z, nearest125°W115°W105°W95°W85°W75°W65°W25°N30°N35°N40°N45°N50°NKATL Atlanta Hartsfield: bias −1.69 °F, n = 29KATL−KAUS Austin Bergstrom: bias +1.05 °F, n = 29KAUS+KBOS Boston Logan: bias −1.20 °F, n = 26KBOS−KDCA Washington Reagan: bias −1.49 °F, n = 29KDCA−KDEN Denver Intl: bias +1.60 °F, n = 28KDEN+KDFW Dallas-Fort Worth: bias +1.35 °F, n = 29KDFW+KEWR Newark Liberty: bias −2.68 °F, n = 29KEWR−KIAH Houston Bush: bias −1.03 °F, n = 29KIAH−KLAS Las Vegas Harry Reid: bias −3.96 °F, n = 28KLAS−KLAX Los Angeles Intl: bias −2.06 °F, n = 28KLAX−KMIA Miami Intl: bias −1.62 °F, n = 29KMIA−KMSP Minneapolis-St Paul: bias −0.14 °F, n = 29KMSP−KMSY New Orleans Intl: bias +0.32 °F, n = 29KMSY+KNYC New York Central Park: bias −0.23 °F, n = 29KNYC−KOKC Oklahoma City: bias +2.37 °F, n = 29KOKC+KORD Chicago O'Hare: bias −1.31 °F, n = 29KORD−KPHL Philadelphia Intl: bias −2.49 °F, n = 29KPHL−KPHX Phoenix Sky Harbor: bias +0.05 °F, n = 28KPHX+KSAN San Diego Lindbergh: bias −3.76 °F, n = 28KSAN−KSAT San Antonio Intl: bias +0.97 °F, n = 29KSAT+KSEA Seattle-Tacoma: bias +1.16 °F, n = 28KSEA+KSFO San Francisco Intl: bias +1.80 °F, n = 28KSFO+KTTN Trenton Mercer: bias +0.44 °F, n = 29KTTN+

Source: CastCheck · station coordinates NWS/NOAA · Albers conic projection · stations.csv. Hover a dot for its numbers.

model too cold interval includes zero model too warm dot area ∝ scored days

ECMWF IFS HRES bias as a table
StationModelMAE °FBias °Fn
KATL Atlanta HartsfieldECMWF IFS HRES2.78−1.6929
KAUS Austin BergstromECMWF IFS HRES2.30+1.0529
KBOS Boston LoganECMWF IFS HRES2.74−1.2026
KDCA Washington ReaganECMWF IFS HRES2.80−1.4929
KDEN Denver IntlECMWF IFS HRES3.72+1.6028
KDFW Dallas-Fort WorthECMWF IFS HRES1.89+1.3529
KEWR Newark LibertyECMWF IFS HRES3.40−2.6829
KIAH Houston BushECMWF IFS HRES2.48−1.0329
KLAS Las Vegas Harry ReidECMWF IFS HRES4.41−3.9628
KLAX Los Angeles IntlECMWF IFS HRES2.36−2.0628
KMIA Miami IntlECMWF IFS HRES2.72−1.6229
KMSP Minneapolis-St PaulECMWF IFS HRES2.37−0.1429
KMSY New Orleans IntlECMWF IFS HRES2.08+0.3229
KNYC New York Central ParkECMWF IFS HRES2.11−0.2329
KOKC Oklahoma CityECMWF IFS HRES3.81+2.3729
KORD Chicago O'HareECMWF IFS HRES2.28−1.3129
KPHL Philadelphia IntlECMWF IFS HRES3.28−2.4929
KPHX Phoenix Sky HarborECMWF IFS HRES2.63+0.0528
KSAN San Diego LindberghECMWF IFS HRES3.79−3.7628
KSAT San Antonio IntlECMWF IFS HRES1.64+0.9729
KSEA Seattle-TacomaECMWF IFS HRES2.45+1.1628
KSFO San Francisco IntlECMWF IFS HRES2.74+1.8028
KTTN Trenton MercerECMWF IFS HRES2.05+0.4429

All-model mean bias

Mean bias averaged over every scored model at each station (lead day 1, instantaneous temperature, 365d window, 12Z, nearest). A station that is cold for all of them is a station property — elevation, coastline, a grid cell that is partly sea — not a model ranking. Dot area grows with the number of scored days; fill is the mean bias (warm = model too warm, cool = too cold). The frame is a latitude/longitude graticule on an Albers projection.

All-model mean bias in °F, lead day 1, instantaneous temperature, 365d window, 12Z, nearest125°W115°W105°W95°W85°W75°W65°W25°N30°N35°N40°N45°N50°NKATL Atlanta Hartsfield: bias −1.90 °F, n = 196KATL−KAUS Austin Bergstrom: bias +0.68 °F, n = 196KAUS+KBOS Boston Logan: bias −1.17 °F, n = 184KBOS−KDCA Washington Reagan: bias −1.36 °F, n = 196KDCA−KDEN Denver Intl: bias +0.49 °F, n = 195KDEN+KDFW Dallas-Fort Worth: bias −1.29 °F, n = 193KDFW−KEWR Newark Liberty: bias −1.76 °F, n = 195KEWR−KIAH Houston Bush: bias −0.36 °F, n = 197KIAH−KLAS Las Vegas Harry Reid: bias −5.88 °F, n = 195KLAS−KLAX Los Angeles Intl: bias −0.09 °F, n = 196KLAX−KMIA Miami Intl: bias −0.77 °F, n = 196KMIA−KMSP Minneapolis-St Paul: bias −1.99 °F, n = 197KMSP−KMSY New Orleans Intl: bias −1.74 °F, n = 197KMSY−KNYC New York Central Park: bias +0.20 °F, n = 196KNYC+KOKC Oklahoma City: bias +0.09 °F, n = 196KOKC+KORD Chicago O'Hare: bias −0.79 °F, n = 197KORD−KPHL Philadelphia Intl: bias −1.35 °F, n = 197KPHL−KPHX Phoenix Sky Harbor: bias −3.63 °F, n = 196KPHX−KSAN San Diego Lindbergh: bias −2.82 °F, n = 195KSAN−KSAT San Antonio Intl: bias −0.30 °F, n = 197KSAT−KSEA Seattle-Tacoma: bias −0.48 °F, n = 196KSEA−KSFO San Francisco Intl: bias −0.54 °F, n = 196KSFO−KTTN Trenton Mercer: bias +2.00 °F, n = 193KTTN+

Source: CastCheck · station coordinates NWS/NOAA · Albers conic projection · stations.csv. Hover a dot for its numbers.

model too cold interval includes zero model too warm dot area ∝ scored days

All-model mean bias as a table
StationModelsMAE °FBias °Fn
KATL Atlanta Hartsfield11 models3.03−1.90196
KAUS Austin Bergstrom11 models3.14+0.68196
KBOS Boston Logan11 models2.45−1.17184
KDCA Washington Reagan11 models2.66−1.36196
KDEN Denver Intl11 models3.22+0.49195
KDFW Dallas-Fort Worth11 models2.62−1.29193
KEWR Newark Liberty11 models2.75−1.76195
KIAH Houston Bush11 models2.44−0.36197
KLAS Las Vegas Harry Reid11 models6.14−5.88195
KLAX Los Angeles Intl11 models2.24−0.09196
KMIA Miami Intl11 models2.46−0.77196
KMSP Minneapolis-St Paul11 models3.08−1.99197
KMSY New Orleans Intl11 models2.51−1.74197
KNYC New York Central Park11 models2.35+0.20196
KOKC Oklahoma City11 models3.17+0.09196
KORD Chicago O'Hare11 models2.25−0.79197
KPHL Philadelphia Intl11 models2.52−1.35197
KPHX Phoenix Sky Harbor11 models4.39−3.63196
KSAN San Diego Lindbergh11 models3.60−2.82195
KSAT San Antonio Intl11 models2.30−0.30197
KSEA Seattle-Tacoma11 models2.26−0.48196
KSFO San Francisco Intl11 models2.25−0.54196
KTTN Trenton Mercer11 models2.98+2.00193

Data availability

Each model is scored only over its own available period (2026-01-01 → 2026-08-30); the windows above are therefore not identical across models. Pairwise comparisons on the permanent-link pages use common days only.

ModelPeriodScored daysCoverage
ECMWF AIFS Single2026-08-02 → 2026-08-3029
2026-01-012026-08-30
Aurora (GFS init)2026-08-29 → 2026-08-302
2026-01-012026-08-30
Aurora (IFS init)2026-08-29 → 2026-08-302
2026-01-012026-08-30
FourCastNet v2 (GFS init)2026-08-29 → 2026-08-302
2026-01-012026-08-30
FourCastNet v2 (IFS init)2026-08-29 → 2026-08-302
2026-01-012026-08-30
NCEP GFS2026-08-02 → 2026-08-3029
2026-01-012026-08-30
GraphCast (GFS init)2026-01-02 → 2026-08-3036
2026-01-012026-08-30
GraphCast (IFS init)2026-01-02 → 2026-08-30197
2026-01-012026-08-30
ECMWF IFS HRES2026-08-02 → 2026-08-3029
2026-01-012026-08-30
Pangu-Weather (GFS init)2026-08-29 → 2026-08-302
2026-01-012026-08-30
Pangu-Weather (IFS init)2026-08-29 → 2026-08-302
2026-01-012026-08-30
Persistence (baseline)2026-01-01 → 2026-08-30215
2026-01-012026-08-30

Stations23

full station table
  • KNYC — New York Central Park
  • KEWR — Newark Liberty
  • KPHL — Philadelphia Intl
  • KTTN — Trenton Mercer
  • KBOS — Boston Logan
  • KDCA — Washington Reagan
  • KATL — Atlanta Hartsfield
  • KMIA — Miami Intl
  • KORD — Chicago O'Hare
  • KMSP — Minneapolis-St Paul
  • KDFW — Dallas-Fort Worth
  • KIAH — Houston Bush
  • KAUS — Austin Bergstrom
  • KSAT — San Antonio Intl
  • KMSY — New Orleans Intl
  • KOKC — Oklahoma City
  • KDEN — Denver Intl
  • KPHX — Phoenix Sky Harbor
  • KLAS — Las Vegas Harry Reid
  • KLAX — Los Angeles Intl
  • KSAN — San Diego Lindbergh
  • KSFO — San Francisco Intl
  • KSEA — Seattle-Tacoma