东亚地区数值预报可预报极限及预报不确定性研究

As two major issues in atmospheric sciences, predictability limit and prediction uncertainty for numerical prediction are among the world's hottest and toughest research topics. Prolonging valid time and reducing the degree of numerical prediction uncertainty have a direct bearing on national economic production, military defense as well as people's daily life. Increasing predictability limit of numerical weather prediction has remained a major research objective in US, Europe and some other countries and regions, and huge human and material resources have been poured. With THORPEX, the predictability for high impact weather is also listed as its key scientific issue. To start with, the present research looks into the actual East Asia weather predictability limit and prediction uncertainty resulted from the typical numerical models, and seeks to work out some important reasons for the low predictability in the area. Through analysis of numerical simulation of typical weather processes, it also attempts to probe into the energy cascade of meso scale strong convective weather, with a focus on the reunited mechanisms of up scale processes from micro to meso-scale in convection systems, and evaluate the impacts of numerical model's spin-up on the validity of numerical prediction of strong convective weather.

数值预报有效性中的可预报极限及预报不确定性问题是大气科学面临的两个基本问题，也是当今国际攻关热点与难点。数值预报有效期限的提高以及不确定程度的降低与人们的日常生活，国家的经济生产，国防等现实需求直接相关。欧美等国家地区一直把提高预报期限作为数值天气预报研究领域的主要目标，投入了大量人力物力。THORPEX计划把高影响天气可预报性问题列为首要核心科学问题。本研究项目从分析典型数值模式对东亚地区的可预报极限及预报不确定性入手，力图找出造成东亚地区预报有效性差的若干重要原因；通过典型天气过程数值模拟分析研究中尺度强对流天气能量级串问题，重点分析小尺度系统向中尺度对流系统的组合或合并组织机制，并评估模式spin-up问题对强对流天气数值预报有效性的影响。

数值预报有效性中的可预报极限及预报不确定性问题是大气科学面临的两个基本问题，也是当今国际攻关热点与难点。数值预报有效期限的提高以及不确定程度的降低与人们的日常生活，国家的经济生产，国防等现实需求直接相关。欧美等国家地区一直把提高预报期限作为数值天气预报研究领域的主要目标，投入了大量人力物力。THORPEX 计划把高影响天气可预报性问题列为首要核心科学问题。本研究项目从分析典型数值模式对东亚地区的可预报极限及预报不确定性入手，力图找出造成东亚地区预报有效性差的若干重要原因；通过典型天气过程数值模拟分析研究中尺度强对流天气能量级串问题和spin-up 问题对强对流天气数值预报有效性的影响。试验结果表明保真模式的动能谱与观测得到的实际大气动能谱分布极为相似，能够很好的描述能量级串过程，有着更强的中小尺度系统的描述能力，雷达资料的同化可以使得模拟结果更好的再现能量级串过程，并可以缩短spin up时间，更好的模拟降水过程。