History

30 days of forecast vs. what actually happened at the lake. Click a single day to see its details.

Last 30 days

Rule-based forecast (actually used)
ML-based forecast — logistic (experimental)
ML-based forecast — HistGradientBoosting (black-box model, experimental)
Actual measurements (session ≥ 1 h)
Sun Jun 28 Mon Jul 27
session (≥ 1 h ≥ 11 kt) marginal (≥ 1 h ≥ 8 kt) no wind no data thunderstorm (forecast rows: predicted · actual row: observed)

Jul 1, 2026

81106:0021:0007:26 · avg 2.0 kt · gust 4.9 kt07:58 · avg 2.5 kt · gust 4.3 kt08:27 · avg 0.7 kt · gust 2.2 kt09:29 · avg 4.7 kt · gust 6.5 kt09:32 · avg 5.8 kt · gust 7.6 kt09:59 · avg 11.1 kt · gust 16.2 kt10:04 · avg 8.9 kt · gust 17.3 kt11:09 · avg 5.0 kt · gust 8.6 kt11:10 · avg 7.6 kt · gust 8.6 kt14:29 · avg 8.8 kt · gust 10.8 kt14:34 · avg 9.1 kt · gust 10.8 kt15:18 · avg 12.2 kt · gust 14.6 kt15:49 · avg 6.8 kt · gust 10.8 kt18:19 · avg 5.0 kt · gust 10.3 kt18:21 · avg 8.6 kt · gust 13.0 kt19:03 · avg 10.3 kt · gust 11.9 kt19:19 · avg 8.7 kt · gust 13.5 kt19:26 · avg 7.7 kt · gust 11.9 kt
avg wind gust ignition (8 kt) session (11 kt)
For Jul 1, 2026 · as written at the time

NO GO

100% daytime cloud + only 92 W/m² solar — no insolation to drive a thermal

⚡ Storm risk: the thermal often runs well until the front — watch the sky, come in early.

🤖 ML — logistic experimental NO GO GO 10% · MAYBE 14% · NO GO 76%

Strongest inputs — (+) dew-point spread; (−) morning solar, pressure Δ (thermik).

🤖 ML — HGB · black box experimental MAYBE GO 0% · MAYBE 53% · NO GO 47%

Most distinctive inputs: morning solar, daytime cloud, Innsbruck pressure.

Experimental shadow models — they never drive the verdict. How good are they? →

Previous forecast method

How the day was forecast at the time, before the rules were recalibrated — shown for comparison only.