Skip to main content

Most Accurate Forecast Model This Season

Every major weather model, ranked by how close it came to the snow that actually fell — measured against SNOTEL and resort-reported ground truth, across 108,561 verifications. Lower error is sharper. Updated June 25, 2026.

🏆 SnowSure’s 7-model blend wins
Beats the best single model (JMA) by 4.3% this season.
5.29 cm error
Full season1–3 day4–7 day8–14 day
  1. 1
    JMA
    5.52 cm mean error
  2. 2
    GEM
    5.73 cm mean error
  3. 3
    Météo-France
    5.91 cm mean error
  4. 4
    ICON
    5.93 cm mean error
  5. 5
    GFS
    6.11 cm mean error
  6. 6
    ECMWF
    6.85 cm mean error

Method: mean absolute error between each model’s daily snowfall forecast and measured snowfall (SNOTEL + resort-reported), snow days only. The blend combines all reporting models.

Put this on your site or feed

Free to embed. Updates itself and always credits SnowSure.

Embed code
<iframe src="https://www.snowsure.ai/embed/leaderboard/most-accurate-model" title="Most Accurate Forecast Model — SnowSure" width="100%" height="440" frameborder="0" scrolling="no" style="border:0;max-width:420px" loading="lazy"></iframe>

Paste anywhere HTML is allowed. The widget updates automatically and links back to SnowSure.

Social image cards

Landscape (1200×630)Square (1080×1080)Story (1080×1920)

Developers: JSON at /api/v1/leaderboards/most-accurate-model.