Verified blend accuracy
Does blending weather models beat the best one?
Measured against observed snowfall, SnowSure's blended forecast has 5.3% lower error than NAM — the single most accurate weather model — across 48,120 verified forecasts.
Ground truth: measured (SNOTEL + resort-reported), snow days only. Machine-readable: /api/v1/blend-accuracy.
Forecast Trust — Verified, not guessed — corpus scale, blend vs best model, and the full record.
Blend vs best single model, by horizon
| Horizon | Blend MAE | Best model | Best-model MAE | Improvement | Samples |
|---|---|---|---|---|---|
| Overall | 5.6 cm | NAM | 5.49 cm | -2% | n = 129,266 · as of 2026-10-25 |
| 1–3 days out | 5.2 cm | NAM | 5.49 cm | +5.3% | n = 48,120 · as of 2026-10-25 |
| 4–7 days out | 5.68 cm | NBM | 5.23 cm | -8.6% | n = 63,983 · as of 2026-10-25 |
| 8–14 days out | 6.38 cm | NBM | 3.01 cm | -112% | n = 17,195 · as of 2026-10-25 |
Lower MAE = closer to observed snowfall. Improvement = how much lower the blend's error is than the best single model. Updated 2026-10-30.
Every model, ranked by error
| Model | MAE (cm) | Samples |
|---|---|---|
| METNO | 4.1 | 12 |
| NAM | 5.49 | 2,101 |
| ECMWF | 5.66 | 8,475 |
| UKMO | 5.72 | 27,032 |
| JMA | 5.77 | 47,520 |
| Météo-France | 5.86 | 47,519 |
| ICON | 5.88 | 47,520 |
| HRRR | 5.97 | 66 |
| GEM | 6.06 | 47,519 |
| NBM | 6.2 | 22,402 |
| GFS | 6.27 | 47,522 |
| METEOSWISS | 7.22 | 6,040 |
| ICON2I | 7.33 | 3,811 |
See also: most accurate model by region · Forecast Trust.
