Quality Assessment And Assimilation Of Round-trip Flat-floating Sounding Data

The return radiosonde is a new detection technology independently developed by China.It can obtain atmospheric profiles with a high vertical resolution during the ascent and descent stages and detect long-distance atmospheric information in the stratosphere during the flat-floating stages.We have obtained many valuable atmospheric sounding data from some field experiments.In this paper,Global Navigation Satellite System Radio Occultation(GNSS RO)data and ERA-Interim reanalysis data are used to evaluate the quality of the data.The assimilation test is also carried out by using the CMA?MESO?3km high-resolution model3DVar assimilation system.We discussed the strengths and weaknesses of this data and the impact of the data assimilation on the prediction results of numerical models.The research in this paper plays an essential role in promoting the operational application of this data.It has significant reference value for the instrument's improvement and research and development.By comparing the GNSS occultation data,it is found that the temperature in the ascending(descending)segment of the return radiosonde is in good agreement with the RO data.The mean value of T_RS-T_ROranges from-1 to 1°C in the ascending section.The average range of T_RS-T_RObelow 13 km and above 13 km in the descending section are-1?1°C and 1?3°C,respectively.The big differences between T_RS-T_ROin the flat-floating stage can reach10?20?,and the average range is 7?10?,while that decreases to-4?0?at night.Two possible reasons for the temperature differences in the flat drift section:First,the temperature sensing element is affected by the latent heat heating of attached water droplets,resulting in the temperature containing a false 0?isothermal layer;second,the strong influence of solar radiation on the instrument and sounder housing during the daytime hours in the observation of the upper atmosphere,the higher temperature deviation in the descending section of the higher altitude layer may be due to the influence of radiation errors in the flat drift section has not been eliminated.In this paper,we conducted an assimilation test on this data using the CMA?MESO3DVar assimilation system and studied the sparsification technique for assimilating this data.It is found that selecting observations close to the model level and significant leve are an effective sparsification method,and assimilating the sparse return radiosonde data can better characterize the structure of small-and medium-scale convective systems.The assimilated analysis field enhances the meridional,high and low-level temperature difference,making the atmospheric stratification more unstable.The low-level wind field convergence is strengthened and overlaps with the water-vapor flux convergence area.The 700-850h Pa storm area is more humid than the wet field,and the small-scale cyclonic circulation and warm shear flow field are strengthened to favor precipitation.The amount of data observed in the descending section is relatively limited.Its impact on the analysis and forecast of the situation field at the moment of warm initiation is small.However,the ETS scores of accumulated precipitation of different magnitudes from 12 to 24 h are improved,with an average increase of 0.7%.The ETS scores of precipitation above 50.0 mm are improved by 2.2%.