Verified blend accuracy
Does blending weather models beat the best one?
Measured against observed snowfall, SnowSure's blended forecast has -233.6% lower error than METNO — the single most accurate weather model — across 47,552 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 | 5.47 cm | METNO | 1.5 cm | -264.6% | n = 126,732 · as of 2026-09-13 |
| 1–3 days out | 5 cm | METNO | 1.5 cm | -233.6% | n = 47,552 · as of 2026-09-13 |
| 4–7 days out | 5.6 cm | JMA | 5.92 cm | +5.5% | n = 63,338 · as of 2026-09-13 |
| 8–14 days out | 6.35 cm | ECMWF | 6.59 cm | +3.6% | n = 15,849 · as of 2026-09-13 |
Lower MAE = closer to observed snowfall. Improvement = how much lower the blend's error is than the best single model. Updated 2026-09-14.
Every model, ranked by error
| Model | MAE (cm) | Samples |
|---|---|---|
| METNO | 1.5 | 3 |
| ECMWF | 5.41 | 7,821 |
| JMA | 5.51 | 47,550 |
| ICON | 5.65 | 47,550 |
| Météo-France | 5.7 | 47,549 |
| GEM | 5.91 | 47,549 |
| GFS | 6.04 | 47,552 |
See also: most accurate model by region · Forecast Trust.
