中文摘要：

近年来，GPS／LEO（全球定位系统／低地球轨道）卫星无线电掩星技术给出了地球大气探测的新途径．从LEO卫星观测到的掩星数据可以反演的地球大气的气压、水汽、温度等剖面；它们对气象和大气科学研究，是具有潜在价值的数据资源．掩星数据资料的同化技术可以有效地改进这些气象参数的剖面，从而改进目前的数值天气预报模式．在当前采用的一维变分同化反演技术中，可以用掩星观测资料的大气折射率或弯曲角剖面进行同化，来反演大气水汽和温度剖面以及海平面压强．作为独立自主开发的GPS／LEO掩星技术系统的一部分，以欧洲中尺度天气预报分析（ECMWF）资料为背景场，CHAMP掩星观测得到的折射率剖面为观测值，采用Levenberg—Marquardt方法实行GPS掩星资料一维变分同化．在讨论中，用掩星观测点附近相应的探空气球资料来检验CHAMP掩星资料变分同化的结果．

英文摘要：

A new approach to remote sensing of earth＇s atmosphere using the Global Positioning System /Low Earth Orbit （GPS/LEO） radio occultation technique has been approved in the past decade. The vertical profiles of pressure, water vapor and temperature retrieved from GPS/LEO occultation data are potentially valuable data source for the meteorological community and atmospheric science. The assimilation process of GPS/LEO occultation data may effectively improve the quality of these retrieved meteorological profiles, and further the current numerical weather prediction （NWP） model. Through one dimensional variational （1DVAR） retrieval technique, the refractivity or the bending angle profiles from GPS/LEO occultation is possibly selected to be assimilated into NWP model, and then the atmospheric temperature and water vapor profiles, as well as the surface pressures are retrieved. In this paper, after describing technique procedure of the refractivity 1DVAR retrieval technique, we present some profile samples from CHAMP occultation refractivity observations based on Levenberg-Marquardt algorithm associated with the background field from the European Center for Medium Range Weather Forecasting （ECMWF） analyses. As a part of discussion, the retrieved profiles are compared with the radiosonde data nearby the GPS/LEO occultation sounding points.