集合平方根滤波同化AMSU-A辐射率的观测系统模拟试验  

An observation system simulation experiment of assimilating AMSU-A radiance with the ensemble square root filter

作　　者：毛璐[1] 张述文[1] 王恬[1] 刘彦华[1] 

机构地区：[1]兰州大学大气科学学院,甘肃省干旱气候变化与减灾重点实验室,兰州730000

出　　处：《兰州大学学报（自然科学版）》2015年第3期351-357,共7页Journal of Lanzhou University(Natural Sciences)

基　　金：国家重点基础研究发展计划(973计划)项目(2013CB430102;2012CB956200);国家自然科学基金项目(41075074)

摘　　要：借助观测系统模拟试验,针对晴天和有降水两种天气状况,探究用集合平方根滤波同化卫星辐射率资料来改进中尺度天气预报模式初始场的可行性.试验结果表明:在晴天条件下,同化后的分析场较同化前背景场更加接近于真实场,尤其是对所选AMSU-A第5通道辐射率观测值贡献较大的那些层,如500,700,800 hPa3层位温的分析场相对背景场均方根误差分别减小了26%,15%和42%,同化效果显著.在有降水天气状况下,试验结果在特定的高度层上也表现出一定的正贡献,但总体而言对各气象要素场的改善均不及晴天时明显,说明该种天气类型下的卫星资料同化还有待进一步的研究.Using an observation system simulation experiment,the possibility of improving the mesoscale model initial state was investigated under both the sunny and rainy weather conditions by assimilating satellite data into the weather research and forecast model with the ensemble square root filter.Under the sunny condition,the radiance data assimilation experiment showed that the mesoscale analysis including the temperature,humidity and wind fields could be largely improved,compared with the initial background state,especially at the levels with the peak energy contributing to the observed radiance for the channel 5 of AMSU-A.For example,the root mean square error of the potential temperature analysis decreased by 26%,15%and 42%relative to the background field respectively at the levels of 500,700 and 800 hPa.Under the rainy weather condition,it was shown that assimilating satellite radiance data still had some positive contribution to the analysis,but the positive effect was very slight compared with the sunny weather condition,indicating that satellite data assimilation under precipitation remains to be further studied.

关 键 词：卫星辐射率 集合平方根滤波 资料同化 中尺度天气研究和预报 

分 类 号：P414.4[天文地球—大气科学及气象学]