【作者】 胡文东； 沈桐立； 赵光平； 丁建军； 杨有林； 刘建军； 陈晓娟；

【Author】 HU Wen-dong1,2,3 SHEN Tong-li3 ZHAO Guang-ping1,2 DING Jian-jun1 YANG You-lin1 LIU Jian-jun1 CHEN Xiao-juan2 (1Key Laboratory of Meteorological Disaster Preventing and Reducing in Ningxia, 2 Ningxia Meteorological Observatory, Yinchuan, China, 750002 3Nanjing Institute of Meteorology, Nanjing, China, 210044)

【机构】 宁夏气象防灾减害重点实验室； 南京气象学院； 宁夏气象台；

【摘要】 为提高省级气象部门强降水数值预报业务水平,应用气象卫星资料,进行中尺度数值预报同化研究是一条重要途径。在中纬度地区,直接使用线性方法进行气象卫星资料反演,难以取得满意效果;使用复杂的反演模型,受硬件设备和人员结构的影响,则无法满足省级气象部门的业务需求。本文利用地球同步气象卫星红外资料、宁夏各气象站地面及探空资料,根据卫星资料与相对湿度的对应曲线,设计了多种待选拟合函数。通过最优拟合方法,建立了不同层次相对湿度与气象卫星资料多种导出量的非线性关系,通过试验选定模型,避免了过拟合产生的不稳定现象。利用非线性回归,建立了气象卫星资料反演湿度场的数学模型,计算结果表明反演效果优于线性回归。这项研究为省级预报业务的中尺度数值预报同化的实践找到了一项可行方案。更多还原

【Abstract】 In order to improve the operational capability of severe precipitation prediction in provincial meteorological agency, satellites data, surface observation and sonde data were applied, and the method of optimal fitting was used to set up the nonlinear relation between the relative humidity in different layers and the derived element of meteorological satellite data such as temporal variation and spatial gradient of gray level and brightness temperature. After comparison, the adequate transmissions were adopted to avoid overfit. The nonlinear regression has been used to develop the nonlinear retrieve model of the humidity with satellite data, and the method to retrieve humidity with satellite data in this paper is much better than linear regression. After the variation assimilation and quality control, the retrieved initial field was put into MM5V3 to forecast a sudden happened severe convection in Ningxia, and the fine capability of the method used here was proved by the experiment. A proper way to conduct operational assimilation in provincial meteorological service was found.更多还原