【摘要】 数据同化可以通过同化观测资料为数值天气预报提供高精度的初值数据。但是,在数据同化使用的卫星观测资料中存在与辐射传输模型模拟数据不一致的数据。同化与模拟不一致的观测数据可能导致分析值不平衡,甚至导致系统崩溃。为保证数据同化结果的准确性,在数据同化前应对观测资料进行质量控制。依据国内外的相关研究,尝试对亮温数据的质量控制方案进行系统总结。针对亮温数据,分析了直接同化中亮温数据的误差来源;介绍了亮温数据质量控制方法,包括合理性检验、离群数据剔除、偏差订正和变分质量控制;并以双权重算法为例,进行了亮温数据质量控制的实例分析;最后,针对亮温数据质量控制提出了几点展望。更多还原

【Abstract】 Data assimilation can provide high-quality initial data for numerical weather forecast through assimilation of observational data. However,there may be observations that are inconsistent with the simulation data of radiation transmission model in assimilation system. Assimilation of inconsistent observations can lead to an unbalanced analysis and even a crash of the system. In order to ensure the accuracy of results of data assimilation,the quality of observations should be controlled before data assimilation. Based on the relevant research at home and abroad,this paper attempts to systematically summarize the quality control scheme of the brightness temperature data. In this paper,the source of error of brightness temperature data in data assimilation is analyzed. The methods of quality control of brightness temperature data are introduced,including domain truth validation,outlier data elimination,bias correction and variational quality control.And taking the double weight algorithm as an example,an experiment of quality control of brightness temperature data is done. Finally,some prospects are put forward for the brightness temperature data quality control.更多还原