Verified blend accuracy
Does blending weather models beat the best one?
Measured against observed snowfall, SnowSure's blended forecast has 7.1% lower error than JMA — the single most accurate weather model — across 42,525 verified forecasts.
Ground truth: measured (SNOTEL + resort-reported), snow days only. Machine-readable: /api/v1/blend-accuracy.
Blend vs best single model, by horizon
| Horizon | Blend MAE | Best model | Best-model MAE | Improvement |
|---|---|---|---|---|
| Overall | 5.29 cm | JMA | 5.52 cm | +4.3% |
| 1–3 days out | 4.89 cm | JMA | 5.26 cm | +7.1% |
| 4–7 days out | 5.55 cm | JMA | 5.67 cm | +2.1% |
| 8–14 days out | 5.47 cm | ECMWF | 5.68 cm | +3.8% |
Lower MAE = closer to observed snowfall. Improvement = how much lower the blend's error is than the best single model. Updated 2026-07-04.
Every model, ranked by error
| Model | MAE (cm) | Samples |
|---|---|---|
| JMA | 5.26 | 42,524 |
| ICON | 5.4 | 42,524 |
| Météo-France | 5.53 | 42,522 |
| GEM | 5.62 | 42,523 |
| GFS | 5.82 | 42,525 |
| ECMWF | 6.91 | 42,524 |
See also: most accurate model by region.
