engleski

Verification of Analog-based Wind Speed Predictions in Complex Terrain

In this work the verification procedure was formulated and used to improve and assess the performance of analog-based wind speed predictions in coastal complex terrain. The analog ensemble (AnEn) predictions were generated by Aire Limitée Adaptation dynamique Développement InterNational model (ALADIN) run. They were tested at several climatologically different locations across Croatia for point-based wind speed predictions at 10 m height. The focus was on the group of stations prone to high wind speeds (e.g., bora wind). The verification procedure included several metrics, considering the wind speed as continuous, categorical and probabilistic predictand. This study shows that deterministic AnEn predictions, compared to model used to generate it, improve correlation between predictions and measurements while reducing bias and root-mean-square error, especially in coastal complex terrain. The combination of analog and Kalman filter approach performs best for the high wind speed categories. Additionally, probabilistic AnEn predictions provide reliable information about their uncertainty.

probabilistic forecasting; complex terrain; analog-based method; wind speed; statistical techniques

nije evidentirano