WeatherBlend

Multi-model forecast blending for Sennen, Cornwall

Wind models

10 m wind speed — ERA5-truth `wind` + Dunkeswell-truth `wind_speed_lgb`; plus gust and direction (`wind_mvn`). MAE — m/s for speed/gust, ° for direction; lower better. Δ vs best single NWP — negative = blend wins.

10 m wind speed — Sennen Cove

Phase wind · v2026-08-02_154014

10 m wind-speed LightGBM blender, ERA5 WindSpeed10m truth. Per-NWP speed + direction (sin/cos) + ensemble spread + calendar. Trained 2026-08-02. Metric: Test MAE (m/s).

Lead Blend Best single Δ vs best
+24h 0.557 ecmwf_aifs025_single (0.592) -5.9%
+48h 0.706 ecmwf_aifs025_single (0.724) -2.4%
+72h 0.924 ecmwf_aifs025_single (0.898) +2.9%
Verify history (no runs yet)

No verify rows yet for phase wind. Next cycle: Mon/Thu 09:30 UTC, then 5d ERA5 latency.

Phase wind_speed_lgb · v2026-08-02_133021_wind_speed_lgb

10 m wind-speed LightGBM blender trained on Dunkeswell SYNOP obs (real-station truth) instead of ERA5. 29-feature lean+ORO+spread set. The honest real-wind predictor — see the cross-truth bake-off. Trained 2026-08-02. Metric: Test MAE (m/s).

Lead Blend Best single Δ vs best
+24h 1.372 ukmo_seamless (2.434) -43.6%
+48h 1.308 icon_seamless (2.417) -45.9%
+72h 1.423 icon_seamless (2.467) -42.3%
Verify history (no runs yet)

No verify rows yet for phase wind_speed_lgb. Next cycle: Mon/Thu 09:30 UTC, then 5d ERA5 latency.

Wind direction — Sennen Cove

Phase wind_mvn · v2026-08-02_132727_wind_mvn

PyTorch MLP predicting a bivariate normal over the (u, v) wind vector, trained on Dunkeswell SYNOP truth. Direction = atan2(−μ_u, −μ_v); also emits a calibrated direction confidence interval. Trained 2026-08-02. Metric: Test MAE (°).

Lead Blend Best single Δ vs best
+24h 9.517 ecmwf_ifs025 (10.782) -11.7%
+48h 11.952 ecmwf_ifs025 (13.155) -9.1%
+72h 14.972 ecmwf_ifs025 (16.448) -9.0%
Verify history (no runs yet)

No verify rows yet for phase wind_mvn. Next cycle: Mon/Thu 09:30 UTC, then 5d ERA5 latency.

10 m wind gust — Sennen Cove

Phase wind_gust_lgb · v2026-08-02_154134_wind_gust_lgb

10 m wind-gust LightGBM blender, ERA5 WindGusts10m truth. Per-NWP gust + gust/speed ratio + ratio spread over the gust-publishing NWPs. Trained 2026-08-02. Metric: Test MAE (m/s).

Lead Blend Best single Δ vs best
+24h 1.025 gem_seamless (1.445) -29.1%
+48h 1.177 gem_seamless (1.653) -28.8%
+72h 1.404 gem_seamless (1.743) -19.4%
Verify history (no runs yet)

No verify rows yet for phase wind_gust_lgb. Next cycle: Mon/Thu 09:30 UTC, then 5d ERA5 latency.