doi:  10.3878/j.issn.1006-9895.1908.18262
ETKF初值扰动方法中真实观测及扰动调节因子研究

Study on the Application of Real Observation Data and Rescaling factor in Ensemble Transform Kalman Filter Initial perturbation Method
摘要点击 49  全文点击 19  投稿时间:2018-12-03  修订日期:2019-06-14
查看HTML全文  查看全文  查看/发表评论  下载PDF阅读器
基金:  
中文关键词:  区域集合预报,ETKF,真实观测,扰动调节因子
英文关键词:  Regional Ensemble forecast, Ensemble Transform Kalman Filter, real observation, rescaling factor
  
作者中文名作者英文名单位
张涵斌Zhang Hanbin中国气象局北京城市气象研究所
引用:张涵斌.2020.ETKF初值扰动方法中真实观测及扰动调节因子研究[J].大气科学
Citation:Zhang Hanbin.2020.Study on the Application of Real Observation Data and Rescaling factor in Ensemble Transform Kalman Filter Initial perturbation Method[J].Chinese Journal of Atmospheric Sciences (in Chinese)
中文摘要:
      目前,GRAPES区域集合预报系统中ETKF方法采用的是模拟观测信息,为进一步完善ETKF方法,拟对ETKF初值扰动通过引入真实常规探空观测资料,使扰动场能够代表真实观测的不确定信息,改善区域集合预报技巧。真实观测资料的引入会使得每日的观测数目和分布发生变化,这对ETKF方法而言可能会引起扰动振幅的不稳定,因此在引入真实观测资料的基础上设计了新的扰动振幅调节因子,通过格点空间中离散度和均方根误差关系来对初值扰动振幅进行自适应调整。从初值扰动结构、概率预报技巧以及降水预报效果等方面对比分析了基于模拟观测、真实观测以及真实观测结合新型调节因子的ETKF方案的差异,结果表明:真实探空资料能够有效应用于GRAPES区域集合预报系统中,真实观测资料与模拟观测资料相比较为稀疏,可以获得更大量级的初值扰动振幅;真实观测资料有助于提高区域集合的离散度,但对集合预报准确度以及概率预报结果的提高有限,对于降水预报效果提高也有限;新型的扰动振幅调节因子可以有效获得稳定的初值扰动振幅,并保持ETKF扰动结构,真实观测资料与扰动振幅自适应调节因子相结合,可以有效提高区域集合的概率预报结果,并有效提高降水预报效果。
Abstract:
      At present, the ETKF method in the GRAPES regional ensemble prediction system uses pseudo observation information, and the number and distribution of observations are fixed. In order to further improve the ETKF method, the real observation data is introduced into the ETKF process, the real observational radio sounding data enables the perturbation field to represent the uncertainty information of the observational state. Considering that the number and distribution of real observational data will change every day, this may cause the instability of the perturbation amplitude for the ETKF calculation. Therefore a new self-adjustment amplitude rescaling factor is carried out. The ETKF schemes based on pseudo observation, real observation and real observation plus new rescaling factor are analyzed and compared in terms of perturbation characteristics, ensemble verification and precipitation forecast skill. The results show that the real sounding data can be effectively applied to the GRAPES regional ensemble forecasting system. The real observation data is sparse compared with the pseudo observation data, so large initial perturbation amplitudes can be obtained. Real observation data can help to improve the spread of regional ensemble, but the improvement of ensemble prediction accuracy and probability forecast skill is limited, and the effect of precipitation prediction is also limited. A new perturbation amplitude rescaling factor is designed to adaptively adjust the initial perturbation amplitude through the spread and root mean square error relationship in grid space. The study on the adaptive adjustment rescaling factor of perturbation amplitude shows that the new rescaling factor can effectively obtain the stable initial perturbation amplitude and maintain the ETKF generated perturbation structure. Since the real observation based ETKF scheme exhibit over spread characteristics and shows limited improvement over pseudo observation based ETKF, the real observation data combined with the adaptive rescaling factor can effectively improve the skill of regional ensemble in terms of the probabilistic forecast results and can also effectively improve the precipitation forecasting skill.
主办单位:中国科学院大气物理研究所 单位地址:北京市9804信箱
联系电话: 010-82995051,010-82995052传真:010-82995052 邮编:100029 Email:dqkx@mail.iap.ac.cn
本系统由北京勤云科技发展有限公司设计
京ICP备09060247号