COSMIC Radio Occultation Data Error Analysis And Quality Control

Based on the need of the application of radio occultation data assimilation, this study is focus on the error distribution and quality control program of space-based COSMIC occultation data retrieved the atmospheric profile.Compared The COSMIC retrieved temperature and relative humidity profiles from selected respective one week in summer and winter in2009with collocated spatially and temporally ECMWF and NCEP model objective analysis fields as well as radiosonde observations in this paper. The primary research results show that the distribution of the bias and the standard deviations from these data sources are similar in summer and winter, independent of seasons. The vertical and horizontal temperature distribution of these data sources are very close. Either temperature or relative humidity profiles from ECMWF are more close to COSMIC retrievals than the NCEP fields. The humidity profiles of model objective analysis are all higher than radiosonde observations in nearly whole troposphere and more wet than COSMIC retrievals over most ocean in middle-upper troposphere. The COSMIC retrieved humidity profiles are higher than radiosonde measurements, having positive bias with maximum30%at300hPa.To make the COSMIC occultation data improve the weather forecast and the accuracy of numerical forecast, removing the outliers is important, this paper used the refraction index of COSMIC occultation data, researched using biweight quality control methods to check occultation data quality. Analyse the statistical characteristic between COSMIC data and NCEP data, Then evaluate that plan, analyse the result.