Forecast skill of regional ensemble system compared to the higher resolution deterministic model

Abstract

⎯ The 11 km regional ensemble ALADIN-LAEF (Aire Limitée Adaption Dynamique Développment InterNational-Limited Area Ensemble Forecasting) is evaluated by comparison with the 5 km deterministic model ALARO (ALADIN and AROME combined model-Application of Research to Operations at Mesoscale), in order to investigate the advantages and disadvantages facing short-range ensemble and high-resolution forecasts. To make rational decisions about the benefits or challenges of both systems, the forecast skill was measured through probabilistic and deterministic approaches over a 2-month period from late spring to summer season of 2013. The verification uses observations from 1219 SYNOP stations and 1 km × 1 km precipitation analysis from INCA (Integrated Nowcasting through Comprehensive Analysis) nowcasting system. The evaluation was carried out for three essential meteorological variables: 2 m temperature, 10 m wind speed, and 6-hour cumulated precipitation. From the probabilistic point of view, the results show that ALADIN-LAEF outperforms ALARO-LAGGED ensemble system, being statistically more reliable. From the deterministic point of view, the high-resolution deterministic system simulates better the precipitation forecast structure with respect to the observations. Compared to the ensemble system, the deterministic system cannot provide guidance concerning the forecast uncertainties or probabilities, making the ensemble products a powerful tool for risk assessment and decision making.