集合预报是数值天气预报中发展最快的方向之一，已成为当前预报预测业务准确率和产品丰富性的重要保障。过去30年间，伴随着国际上集合数值预报科学研究的快速发展，业务集合预报技术和系统取得了长足进步。作为集合预报链条中面向下游用户的信息输出端，后处理系统已经成为大规模数据生成和产品制作功能一体化、多层次预报分析方法和技术高度集约化的综合平台，在发挥集合预报优势方面扮演着关键角色。首先回顾了国内外集合预报后处理系统发展历程以及后处理技术发展趋势和走向，归纳并阐述了集合预报后处理系统的主要功能。进一步详细介绍了中国气象局全球/区域集合预报业务系统的集合预报统一后处理系统针对上述各方面主要功能的设计思路、结构框架、关键技术和常规产品等研发进展情况，并重点阐释了如何充分利用该集合预报业务系统实时大样本集合预报数据信息开发的极端预报指数、热带大气季节内振荡、西北太平洋副热带高压和南亚高压等多样化新产品及其在业务中的实际应用。总体来看，集合预报后处理将成为充分发挥集合预报准确度高、信息量大以及社会效益显著等优势的研发新方向。

Ensemble prediction, one of the most rapid developments in numerical weather prediction, has presently become a vital basis for accurate forecasting and assessment of product abundance. In the past three decades, accompanied by the rapid developments in prediction research and techniques, a significant progress has been made in operational technology and systems for ensemble prediction. As the output end of the information facing downstream users in the ensemble prediction chain, the post-processing system has been an integrated platform for the generation of numerous ensemble data, the unification of product-making functions, and the intensification of multilevel forecasting approaches and techniques. In this study, a comprehensive local-to-global review was first conducted for the historical development, current stage, and future direction of the post-processing system and technology for ensemble prediction. Second, the following seven main functions of the post-processing system were summarized: \mathrm{①} Standardized output and distribution of ensemble data; \mathrm{②} Calculation of ensemble mean and spread statistics; \mathrm{③} Analysis of synoptic and climatological diagnostics; \mathrm{④} Generation and issuance of deterministic and stochastic ensemble prediction products; \mathrm{⑤} Extraction and interpretation of big ensemble data and information; \mathrm{⑥} Calibration and improvement of deterministic and stochastic ensemble forecasts; \mathrm{⑦} User-customized product services and visualization. Finally, the unified post-processing system in the China Meteorological Administration-Global Ensemble Prediction System/Regional Ensemble Prediction System (CMA-GEPS/REPS) was discussed in terms of the above main functions. The focus was on finding ways to make better use of the big ensemble data and information from the CMA-GEPS/REPS real-time forecasts to study. Further, the intent was to develop a variety of new ensemble products, particularly including the extreme forecast index, Madden-Julian oscillation, western-Pacific subtropical high, and south-Asian high, as well as learning to apply them to realistic operational forecasting. Overall, the post-processing technique is becoming a predominant research and development direction, building on the advantages of ensemble prediction ranging from high forecast accuracy to actionable insights with significant social, environmental, and economic benefits.