Loading…
Loading…
These scores come from winters the model never trained on. For each winter we grade, we rebuild the model without that winter, then have it forecast that winter from the GFS forecasts that were actually issued one, two and three days ahead. We compare the result with the snow NOAA's observer at Alta measured. A winter is November through April. This is a look back: we did not publish these forecasts at the time.
Live forecasts may score a little worse than this. In the test, the model always has that day's run of WeatherNext, the second weather model it uses. Live, the evening forecast goes out before that run exists, so it uses the previous day's and has none for day 3. The overnight forecast waits up to two hours for the fresh run. The GFS forecast behind a graded day is one made that many days earlier, plus up to five hours.
GFS is the weather model we correct, so this shows how much the correction does. It is an easy target: raw GFS badly under-calls Alta snowfall, and no forecaster publishes it as is. It says little about how we compare with forecasts people actually use.
| Lead | Avg miss (ML / GFS) | Correlation | Powder-day ETS (6″+) | P(6″+) Brier (ML / climo) | Range hit rate |
|---|---|---|---|---|---|
| Day 1 | 0.88″ / 1.74″✓ | 0.89 / 0.64✓ | 0.55 / 0.02✓ | 0.047 / 0.119✓ | 88% (target 80%) |
| Day 2 | 0.95″ / 1.76″✓ | 0.87 / 0.65✓ | 0.66 / 0.02✓ | 0.050 / 0.119✓ | 86% (target 80%) |
| Day 3 | 1.16″ / 1.78″✓ | 0.81 / 0.62✓ | 0.47 / 0.05✓ | 0.065 / 0.119✓ | 82% (target 80%) |
The numbers above pool every test winter, and one odd winter can hide in a pool. Each winter here is graded on its own, days 1–3 together. A dry winter makes the average miss look small because little snow fell, so compare a winter's miss with its snow, not across winters.
| Winter | Snow that winter | Avg snow miss | On 6″+ days | Season total | Water miss vs raw GFS |
|---|---|---|---|---|---|
| 2023–24 from Jan 22 | 499″ 6th highest of 33 | 1.36″ | 3.9″ 18 days | 91% of observed | 0.132″ vs 0.150″ ✓ |
| 2024–25 | 456″ 13th highest of 33 | 1.15″ | 4.3″ 31 days | 80% of observed | 0.107″ vs 0.137″ ✓ |
| 2025–26 | 239″ 2nd lowest of 33 | 0.65″ | 4.1″ 15 days | 101% of observed | 0.077″ vs 0.088″ ✓ |
Snow that winter = the NOAA observer's November–April total (median 418″). Season total = forecast total over observed, on the graded days. Days 1–3 can only be graded from January 2024, when the archive of forecasts issued days ahead begins.
Everything above is a test on past winters. This is what the site actually said, every forecast since the archive began (Oct 18 to Nov 8, 70 GFS runs), checked against the NOAA observer. It's a record, not a test: one season is a small sample, and it's never used to tune the model.
| Day | Forecasts | Avg snow miss | Bias | Range hit rate | Total vs observed | Miss, 00Z run | Miss, 06Z run | Miss, 12Z run | Miss, 18Z run |
|---|---|---|---|---|---|---|---|---|---|
| Today | 70 19 days | 0.11″ | +0.05″ | 99% | 114% | 0.09″ | 0.10″ | 0.16″ | 0.09″ |
| Day 1 | 66 18 days | 0.20″ | +0.05″ | 99% | 112% | 0.20″ | 0.12″ | 0.23″ | 0.24″ |
| Day 2 | 62 17 days | 0.33″ | +0.17″ | 95% | 142% | 0.36″ | 0.33″ | 0.34″ | 0.29″ |
| Day 3 | 57 16 days | 0.59″ | +0.55″ | 90% | 227% | 0.29″ | 0.54″ |
Six ways to turn a weather forecast into inches of snow, graded at day 1. “ML” is machine learning: models trained on past winters at Alta. One ML route predicts snowfall directly. Another predicts melted water, then multiplies it by a snow-to-water ratio. Published on the site: Average of the two ML routes, the most consistent across days 1–3. It also has the lowest average miss of the ML routes here.
| Method | Avg miss | Season bias | Powder-day ETS | Correlation |
|---|---|---|---|---|
Average of the two ML routespublished half “snowfall directly”, half “water × learned ratio”: the next two rows | 0.88″ | −30″ | 0.55 | 0.89 |
ML, snowfall directly one model predicts inches of snow itself | 0.90″ | −48″ | 0.58 | 0.88 |
ML, water × learned ratio one model predicts melted water; a second, trained at Alta, how many inches of snow each inch of water makes | 0.92″ | −13″ | 0.58 | 0.86 |
ML, water × Kuchera ratio the same water model × a generic published snow-to-water formula | 1.08″ | −13″ | 0.56 | 0.82 |
GFS, scaled up GFS snowfall × how much more Alta usually gets than GFS says | 3.70″ | +320″ |
| Lead | Avg miss (ML / GFS) | Correlation | Storm ETS (0.5″+) | P(measurable) Brier (ML / climo) | Range hit rate |
|---|---|---|---|---|---|
| Day 1 | 0.091″ / 0.109″✓ | 0.88 / 0.83✓ | 0.57 / 0.49✓ | 0.051 / 0.217✓ | 89% (target 80%) |
| Day 2 | 0.095″ / 0.117″✓ | 0.87 / 0.83✓ | 0.60 / 0.45✓ | 0.072 / 0.217✓ | 87% (target 80%) |
| Day 3 | 0.116″ / 0.136″✓ | 0.81 / 0.76✓ | 0.43 / 0.36 |
Alta reports snowfall from its Collins study plot, higher on the mountain than the NOAA station the models train on. Using the plot's hourly sensor record (cleaned and calibrated to Alta's published season totals), here is how the forecasts did on days the sensor was working, with both sides scaled to Alta's official numbers.
| Lead | Avg miss (ML / GFS) | Correlation (ML / GFS) | Season-scale bias | Days |
|---|---|---|---|---|
| Day 1 | 1.58″ / 2.33″✓ | 0.79 / 0.58✓ | −3″ | 418 |
| Day 2 | 1.57″ / 2.30″✓ | 0.79 / 0.59✓ | −15″ | 418 |
| Day 3 | 1.86″ / 2.38″✓ | 0.70 / 0.53✓ | −26″ | 418 |
Collins averages 1.31× the NOAA station by season (±0.14). Individual storms vary (windier and warmer ones favor Collins), but not in a way we could forecast, so one factor is used. The sensor record ends 2026-03-19.
Google DeepMind's WeatherNext 2 is an AI ensemble with its own archive of forecasts as issued, back to 2022. The same boosted trees, given WeatherNext alongside GFS, scored on the same held-out winters as the GFS-only version:
| Lead | Avg snow miss (with / without) | Stormy days | Snow total vs observed | Days |
|---|---|---|---|---|
| Day 1 | 0.88″ / 1.01″✓ | 3.06″ / 3.57″✓ | 878″ / 848″ of 971″ | 462 |
| Day 2 | 0.95″ / 1.07″✓ | 3.42″ / 3.90″✓ | 852″ / 837″ of 971″ | 462 |
| Day 3 | 1.16″ / 1.32″✓ | 3.90″ / 4.48″✓ | 836″ / 859″ of 971″ | 462 |
Forecast days with WeatherNext available use this model; the rest use GFS alone. Scores above this section are for the GFS + WeatherNext model. Graded with WeatherNext 2 historical forecasts © Google LLC, CC BY 4.0.
Past day 3 the corrected snow totals no longer clearly beat raw GFS, so the site gives only the chance of a 6″+ day, in three tiers. Graded on held-out winters (2023–24 to 2025–26), November to April, against the usual odds for the time of year. A day is shown only while it beats them.
| Lead | Skill vs usual odds | 6″+ followed: quiet / possible / storm signal | 6″+ days |
|---|---|---|---|
| Day 4 | 0.31 | 2% / 22% / 55% | 64 of 461 |
| Day 5 | 0.26 | 4% / 22% / 51% | 64 of 461 |
| Day 6 | 0.21 | 5% / 21% / 48% | 64 of 460 |
| Day 7 | 0.11 | 5% / 24% / 37% | 64 of 459 |
| Skill by winter | Day 4 | Day 5 | Day 6 | Day 7 |
|---|
| 0.45″ |
| 1.05″ |
Each GFS run issues its own forecast for each day, so a day is graded once per run that covered it. Runs are UTC; the 18Z run comes out in the evening and usually has no WeatherNext input on day 3. Updated weekly.
| 0.31 |
| 0.64 |
GFS, raw what most weather apps show | 1.74″ | −247″ | 0.02 | 0.64 |
Season bias = average daily bias × 151 winter days.
| 0.088 / 0.217✓ |
| 83% (target 80%) |
✓ marks where the ML beats the comparison. ETS (equitable threat score) grades yes/no calls of a big day: 1 is perfect, 0 is no better than chance. Brier grades a probability forecast against always forecasting the historical rate (lower is better).
Season total: observed 239″, ML 245″, raw GFS 109″ (days with an observation only).
| Date | Observed snow | ML | p10–p90 | Raw GFS | Observed water | ML water |
|---|---|---|---|---|---|---|
| 2025-11-01 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-11-02 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-11-03 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-11-04 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-11-05 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-11-06 | 0.3 | 0.3 | 0.0–0.0 | 0.3 | 0.08 | 0.07 |
| 2025-11-07 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-11-08 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-11-09 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-11-10 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-11-11 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-11-12 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-11-13 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-11-14 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.02 |
| 2025-11-15 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-11-16 | 0.3 | 0.2 | 0.0–0.0 | 1.4 | 0.16 | 0.22 |
| 2025-11-17 | 1.8 | 6.1 | 2.5–11.1 | 2.6 | 0.20 | 0.78 |
| 2025-11-18 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-11-19 | 0.0 | 0.2 | 0.0–0.0 | 0.1 | 0.01 | 0.09 |
| 2025-11-20 | 0.1 | 1.7 | 0.0–10.4 | 1.9 | 0.04 | 0.32 |
| 2025-11-21 | 0.0 | 0.0 | 0.0–0.0 | 0.1 | 0.00 | 0.01 |
| 2025-11-22 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-11-23 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-11-24 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-11-25 | 0.0 | 0.1 | 0.0–0.0 | 0.0 | 0.00 | 0.01 |
| 2025-11-26 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-11-27 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-11-28 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.01 |
| 2025-11-29 | 0.0 | 1.4 | 0.0–8.6 | 0.5 | 0.00 | 0.11 |
| 2025-11-30 | 7.8 | 5.4 | 2.2–9.7 | 2.0 | 0.79 | 0.39 |
| 2025-12-01 | 2.0 | 0.9 | 0.0–5.4 | 0.1 | 0.09 | 0.05 |
| 2025-12-02 | 1.4 | 0.9 | 0.0–5.5 | 0.2 | 0.05 | 0.08 |
| 2025-12-03 | 1.7 | 2.5 | 0.7–6.2 | 0.6 | 0.14 | 0.17 |
| 2025-12-04 | 0.0 | 0.1 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-12-05 | 6.5 | 6.9 | 2.8–12.4 | 3.7 | 1.20 | 0.83 |
| 2025-12-06 | 11.9 | 10.7 | 7.5–18.7 | 5.6 | 2.00 | 2.05 |
| 2025-12-07 | 0.0 | 0.1 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-12-08 | 0.0 | 0.2 | 0.0–0.0 | 0.4 | 0.00 | 0.01 |
| 2025-12-09 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-12-10 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.01 | 0.00 |
| 2025-12-11 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-12-12 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-12-13 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-12-14 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-12-15 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-12-16 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-12-17 | 3.0 | 1.6 | 0.0–10.2 | 0.7 | 0.98 | 0.23 |
| 2025-12-18 | 0.0 | 0.3 | 0.0–0.0 | 0.2 | 0.00 | 0.02 |
| 2025-12-19 | 0.0 | 0.1 | 0.0–0.0 | 0.1 | 0.00 | 0.02 |
| 2025-12-20 | 2.4 | 2.9 | 0.8–7.4 | 2.1 | 0.46 | 0.48 |
| 2025-12-21 | 0.0 | 0.3 | 0.0–0.0 | 0.0 | 0.00 | 0.08 |
| 2025-12-22 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-12-23 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-12-24 | 0.0 | 0.1 | 0.0–0.0 | 0.0 | 0.06 | 0.03 |
| 2025-12-25 | 0.0 | 1.7 | 0.0–10.7 | 0.7 | 0.66 | 1.06 |
| 2025-12-26 | 1.8 | 1.3 | 0.0–8.2 | 0.6 | 0.28 | 0.27 |
| 2025-12-27 | 7.7 | 7.7 | 3.2–14.0 | 2.0 | 0.58 | 0.84 |
| 2025-12-28 | 1.4 | 2.2 | 0.6–5.6 | 1.0 | 0.12 | 0.13 |
| 2025-12-29 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-12-30 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2025-12-31 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-01-01 | 2.2 | 4.2 | 1.8–7.6 | 3.2 | 0.63 | 0.98 |
| 2026-01-02 | 4.4 | 7.0 | 2.9–12.8 | 4.4 | 0.98 | 1.62 |
| 2026-01-03 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-01-04 | 6.1 | 5.0 | 2.1–9.1 | 2.9 | 0.91 | 0.79 |
| 2026-01-05 | 20.6 | 14.4 | 6.4–21.5 | 5.8 | 2.75 | 2.22 |
| 2026-01-06 | 0.0 | 0.1 | 0.0–0.0 | 0.0 | 0.00 | 0.01 |
| 2026-01-07 | 0.4 | 0.3 | 0.0–0.0 | 0.0 | 0.04 | 0.03 |
| 2026-01-08 | 12.6 | 14.5 | 6.4–21.6 | 2.5 | 0.92 | 0.91 |
| 2026-01-09 | 5.4 | 4.9 | 2.0–8.8 | 1.1 | 0.25 | 0.30 |
| 2026-01-10 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-01-11 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-01-12 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-01-13 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-01-14 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-01-15 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-01-16 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-01-17 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-01-18 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-01-19 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-01-20 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-01-21 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-01-22 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-01-23 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-01-24 | 0.1 | 0.8 | 0.0–4.7 | 0.3 | 0.01 | 0.06 |
| 2026-01-25 | 0.0 | 0.1 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-01-26 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-01-27 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-01-28 | 2.0 | 0.2 | 0.0–0.0 | 0.6 | 0.13 | 0.03 |
| 2026-01-29 | 1.1 | 0.1 | 0.0–0.0 | 0.1 | 0.13 | 0.02 |
| 2026-01-30 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-01-31 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-02-01 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-02-02 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-02-03 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-02-04 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-02-05 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-02-06 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-02-07 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-02-08 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-02-09 | 0.5 | 0.2 | 0.0–0.0 | 0.1 | 0.08 | 0.03 |
| 2026-02-10 | 0.0 | 0.5 | 0.0–0.0 | 0.9 | 0.00 | 0.05 |
| 2026-02-11 | 6.0 | 4.1 | 1.7–7.3 | 2.2 | 0.65 | 0.51 |
| 2026-02-12 | 2.0 | 4.4 | 1.8–7.9 | 1.8 | 0.27 | 0.61 |
| 2026-02-13 | 0.0 | 0.4 | 0.0–0.0 | 0.0 | 0.00 | 0.08 |
| 2026-02-14 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-02-15 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-02-16 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-02-17 | 7.1 | 23.3 | 10.3–34.8 | 4.5 | 1.08 | 2.60 |
| 2026-02-18 | 31.2 | 24.1 | 10.7–36.1 | 7.2 | 2.20 | 1.93 |
| 2026-02-19 | 2.1 | 3.9 | 1.1–9.9 | 0.8 | 0.12 | 0.21 |
| 2026-02-20 | 5.0 | 3.6 | 1.0–9.1 | 0.6 | 0.29 | 0.18 |
| 2026-02-21 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-02-22 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-02-23 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-02-24 | 0.0 | 0.7 | 0.0–4.3 | 0.4 | 0.52 | 0.11 |
| 2026-02-25 | 2.3 | 4.1 | 1.7–7.4 | 0.4 | 1.51 | 0.81 |
| 2026-02-26 | 0.0 | 1.1 | 0.0–7.1 | 0.3 | 0.00 | 0.19 |
| 2026-02-27 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-02-28 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-03-01 | 0.0 | 0.2 | 0.0–0.0 | 0.0 | 0.08 | 0.24 |
| 2026-03-02 | 1.6 | 2.3 | 0.6–5.7 | 0.5 | 0.19 | 0.25 |
| 2026-03-03 | 9.6 | 3.6 | 1.0–9.2 | 1.5 | 1.30 | 0.43 |
| 2026-03-04 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-03-05 | 12.2 | 10.2 | 7.1–17.8 | 2.1 | 1.07 | 0.85 |
| 2026-03-06 | 15.1 | 9.8 | 6.8–17.1 | 3.4 | 0.86 | 0.58 |
| 2026-03-07 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-03-08 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-03-09 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-03-10 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-03-11 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-03-12 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-03-13 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-03-14 | 0.0 | 0.0 | 0.0–0.0 | 0.6 | 0.00 | 0.01 |
| 2026-03-15 | 0.0 | 0.1 | 0.0–0.0 | 0.1 | 0.00 | 0.01 |
| 2026-03-16 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-03-17 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-03-18 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-03-19 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-03-20 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-03-21 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-03-22 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-03-23 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-03-24 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-03-25 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-03-26 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-03-27 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-03-28 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-03-29 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-03-30 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-03-31 | 0.0 | 0.0 | 0.0–0.0 | 0.2 | 0.00 | 0.03 |
| 2026-04-01 | 4.4 | 6.4 | 2.6–11.6 | 7.3 | 0.84 | 1.47 |
| 2026-04-02 | 9.6 | 10.2 | 7.2–17.9 | 4.7 | 1.20 | 1.15 |
| 2026-04-03 | 2.0 | 1.7 | 0.0–10.9 | 0.4 | 0.20 | 0.16 |
| 2026-04-04 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-04-05 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-04-06 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-04-07 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-04-08 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-04-10 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-04-11 | 0.0 | 0.0 | 0.0–0.0 | 0.3 | 0.06 | 0.04 |
| 2026-04-12 | 0.0 | 0.0 | 0.0–0.0 | 0.3 | 0.00 | 0.01 |
| 2026-04-13 | 2.6 | 4.9 | 2.0–8.9 | 3.3 | 0.34 | 0.57 |
| 2026-04-14 | 0.0 | 0.1 | 0.0–0.0 | 0.0 | 0.00 | 0.01 |
| 2026-04-15 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-04-16 | 2.9 | 2.5 | 0.7–6.3 | 1.3 | 0.20 | 0.35 |
| 2026-04-17 | 4.0 | 1.7 | 0.0–10.6 | 1.0 | 0.15 | 0.13 |
| 2026-04-18 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-04-19 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-04-20 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-04-21 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-04-22 | 0.0 | 0.3 | 0.0–0.0 | 1.8 | 0.01 | 0.09 |
| 2026-04-23 | 2.1 | 3.7 | 1.0–9.4 | 1.4 | 0.30 | 0.51 |
| 2026-04-24 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-04-25 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-04-26 | 8.8 | 5.0 | 2.1–9.1 | 7.3 | 0.99 | 0.86 |
| 2026-04-27 | 2.9 | 3.5 | 0.9–8.8 | 3.8 | 0.31 | 0.49 |
| 2026-04-28 | 0.0 | 0.5 | 0.0–3.2 | 0.1 | 0.00 | 0.11 |
| 2026-04-29 | 0.0 | 0.0 | 0.0–0.0 | 0.0 | 0.00 | 0.00 |
| 2026-04-30 | 0.0 | 0.0 | 0.0–0.0 | 0.4 | 0.00 | 0.02 |
| 2023–24 | +0.33 | +0.22 | +0.15 | −0.05 |
| 2024–25 | +0.34 | +0.29 | +0.26 | +0.18 |
| 2025–26 | +0.24 | +0.23 | +0.17 | +0.12 |
Below zero in a winter means the usual odds did better that winter. Skill is 1 minus the ratio of Brier scores (0 = no better than the usual odds, 1 = perfect). The model is one logistic regression on GFS precipitation (that day and the day before), temperature at Alta and the month's usual odds. GFS days 5–7 come from an archive that stored precipitation past 120 hours at a third of NOAA's values for runs from February 2024 to July 2026; it's repaired before training.