往返式探空观测资料的质量控制及不确定性分析

为了推动新型探测资料在数值预报模式中的应用，本文进行了往返式探空资料同化应用前重要的基础性研究工作。基于国内首次往返式探空观测资料，首先建立了面向业务化应用的往返式探空资料质量控制方案，通过对比和分析质量控制前后观测样本的统计特征，论证了质量控制方案的合理性，质量控制后探测要素抽样分布更为合理，要素间一致性得到提高。进而以数值天气预报高时间分辨率的模式预报场和同站址业务常规探空观测资料为参考，分析质量控制后资料的不确定性，结果表明往返探空探测精度达到了世界气象组织WMO（World Meteorological Center）规定的突破目标，部分探测要素甚至实现了理想目标，探测资料具有可用性。最后结合数值模式背景场探讨往返探空资料的可同化性，研究表明往返探空的风场观测和夜间温度观测满足变分同化系统的高斯、无偏假定，可直接同化；气压、湿度和日间温度观测在资料同化前需要开展偏差订正工作，从而更有效的发挥资料价值。本文的研究工作为今后往返探空资料在模式中的同化应用奠定了基础。

Quality Control and Uncertainty Analysis of Return Radiosonde Data

Aiming at promoting the application of new types of sounding data in NWP (numerical weather prediction) models, this paper presents a basic research work of return radiosonde data. Based on archived return radiosonde observation datasets in China, a quality control scheme for future operational implementation purposes is established. By comparing and analyzing the statistical characteristics of observation samples before and after the quality control, the rationality of the quality control method is demonstrated. After the quality control procedure, the sampling distribution of the detection variables is more reasonable, and the inner-consistency of variables is also improved. An uncertainty analysis of return radiosonde data is then carried out by referring to the high-resolution NWP model forecast field and the conventional sounding observation data of the same site. The results show that the precision of return radiosonde reaches the breakthrough target defined by the WMO (world meteorological organization). Some detection variables even achieve the ideal target. Finally, the assimilability of the return radiosonde data is discussed based on the background field of the NWP model. The results show that wind field observations at all times and night temperature observations satisfy the Gaussian and unbiased assumptions of the variational assimilation system and can be assimilated directly. To play a more effective role in the data assimilation system, air pressure, humidity observations, and daily temperature need to be corrected before data assimilation. This work lays a foundation for the future assimilation application of return radiosonde.