国外气象业务中心核心预报能力的比较和发展
Key Forecasting Skill Scores of Leading Operational Centres: Comparison and Outlook
中文摘要：
基于文献和世界气象组织（WMO）世界气象中心以及部分国家气象中心网站的信息，梳理了世界主要气象业务中心的全球天气预报模式预报指标，分析了当前领先气象业务中心的预报水平，并对反映其核心预报能力的天气和气候预报技巧指标进行了比较。在对过去10多年来预报技巧进步趋势分析的基础上，对领先气象业务中心在2025和2035年可能达到的预报指标进行了推测。尽管未来各主要气象业务中心预报系统的提升速率将明显放缓，但是可预报时效将会大大提高：到2025和2035年，500hPa位势高度的预报时效将分别提高至8.5和10.5d，高分辨率模式的预报时效将分别提升至7.6和8.4 d，而多种模式要素预报的有效性将全面接近并可能进入10d。天气型转变预测和MJO预测是反映气候预测的两个核心指标。针对Ni?o 3.4海温距平的未来3和6个月预测，相关性未来分别可能达到93%和86%（2025年）以及96%和90%（2035年），而气候模式对MJO的预报时效在2025年将可能达到49d。新预报量的设计和业务化、下一代数值预报模式以及资料同化技术的研发等，将成为数值天气预报领域发展的新趋势。
中文关键词：
气象业务中心，预报能力，量化分析
KeyWords：
operational centres, forecast skill, quantitative analysis
Abstract：
This paper compares global weather forecast models of leading operations centres with their forecast skills based on reviews of literature and websites of World Meteorological Centers named by WMO. The skill scores currently applied in assessing core forecasting capabilities of leading operational centres are discussed. The skill scores of 2025 and 2035 are projected based on the progress-curves of forecast skill evolution over the past 10 years or more. Although the advance of forecast skill would obviously slow down, the improvement of skillful forecasts will be evident. Forecast lead of 500 hpa geopotential height would be 8.5 and 10.5 days by 2025 and 2035, respectively. The corresponding values would be 7.6 and 8.4 days for high resolution models and both 10 days for multi-model meteorological element forecasts. Prediction skills of Nino 3.4 SST anomaly for the next 3 months and the next 6 months are likely to reach 93% and 86% in 2025, and 96% and 90% in 2035. The MJO forecast lead might be 49 days in 2025. New predictors, next generation numerical models and improved data assimilation are key players in further advancing forecast skills.