Walchensee wind: forecast and measurements

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 β€” logistic (experimental)
ML-based forecast β€” HistGradientBoosting (black-box model, experimental)
Actual measurements (session β‰₯ 1 h)
Mon Oct 12 Tue Nov 10
session (β‰₯ 1 h β‰₯ 11 kt) marginal (β‰₯ 1 h β‰₯ 8 kt) no wind no data thunderstorm (forecast rows: predicted Β· actual row: observed)

Oct 25, 2026

81106:0021:0008:19 Β· avg 10.5 kt Β· gust 14.6 kt08:29 Β· avg 11.6 kt Β· gust 17.3 kt08:39 Β· avg 9.8 kt Β· gust 13.5 kt08:44 Β· avg 11.4 kt Β· gust 16.2 kt09:14 Β· avg 8.4 kt Β· gust 15.1 kt10:19 Β· avg 6.6 kt Β· gust 14.6 kt10:27 Β· avg 4.7 kt Β· gust 14.6 kt10:39 Β· avg 6.9 kt Β· gust 16.2 kt10:49 Β· avg 8.5 kt Β· gust 14.6 kt10:59 Β· avg 5.5 kt Β· gust 10.3 kt11:09 Β· avg 5.0 kt Β· gust 9.2 kt12:09 Β· avg 4.3 kt Β· gust 4.9 kt12:12 Β· avg 3.1 kt Β· gust 4.9 kt13:21 Β· avg 6.5 kt Β· gust 6.5 kt14:09 Β· avg 1.1 kt Β· gust 3.2 kt14:19 Β· avg 2.2 kt Β· gust 2.2 kt14:29 Β· avg 1.9 kt Β· gust 4.9 kt14:30 Β· avg 0.5 kt Β· gust 0.5 kt14:59 Β· avg 5.5 kt Β· gust 9.2 kt15:04 Β· avg 5.5 kt Β· gust 10.8 kt
avg wind gust ignition (8 kt) session (11 kt)
For Oct 25, 2026 Β· as written at the time

NO GO

97% daytime cloud + only 52 W/mΒ² solar β€” no insolation to drive a thermal

⚑ Storm risk: the thermal often runs well until the front β€” watch the sky, come in early.

πŸ€– ML β€” logistic experimental MAYBE GO 21% Β· MAYBE 44% Β· NO GO 35%

Strongest inputs β€” (+) pressure Ξ” (thermik), FΓΆhn Ξ”; (βˆ’) morning solar.

πŸ€– ML β€” HGB Β· black box experimental NO GO GO 11% Β· MAYBE 0% Β· NO GO 89%

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

Experimental shadow models β€” they never drive the verdict. How good are they? β†’

Previous forecast method

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