One of the keys to ensemble forecasting lies in the effective generation of initial ensemble perturbations, which aims to represent the probability distribution of analysis errors. Most previous studies have focused on assessing the predictive skills of ensemble forecasts, offering only indirect assessments of the quality of initial ensemble conditions. This study evaluates the performance of ensemble perturbations in sampling errors at both the initial and forecast stages within the operational ensemble forecast system at NCEP. In particular, the ensemble sampling behavior across different spatial scales is assessed quantitatively, with consideration of both the magnitude and direction. The primary results indicate that, at the initial time, smaller scales exhibit a higher spatial correlation between ensemble spread and error magnitude compared to larger scales. This is related to the instability of small-scale perturbations and errors associated with strong localized forcings. However, ensemble perturbations at larger scales demonstrate superior performance in capturing the spatially coherent structures of analysis errors, exhibiting a higher perturbation–error correlation than those at smaller scales. This can be attributed to the substantially lower effective degrees of freedom (DOF) associated with larger-scale perturbations and errors. These differences in sampling performance across scales persist throughout the entire forecast lead times. The contrasting trends in sampling behavior of the magnitude and direction across various scales suggest that the keys for effective ensemble generation vary with spatial scales.

This paper reviews the development of ensemble weather forecast and the primary techniques employed in the main ensemble prediction systems (EPSs) designed by China and other countries. Here, the emphasis is placed on the advancements in the China Meteorological Administration (CMA) global and regional ensemble prediction systems (i.e., CMA-GEPS and CMA-REPS), with particular attention to operational technologies such as initial and model perturbation methods and the applications of ensemble forecast. Through comparative verification with EPSs from other leading international numerical weather prediction (NWP) centers, CMA’s EPSs demonstrate forecast skills comparable to its global counterparts. As EPSs progress to convective scales and coupled systems between sea, land, air, and ice, the paper addresses some key challenges in ensemble forecast technologies across the aspects of operation, science, integration of artificial intelligence (AI), merging of weather and climate models, and challenging user requirements. Finally, a summary of conclusions and future perspectives on ensemble forecast are provided.