Abstract: Based on the characteristics of atmospheric field observation, the influence of man-made operating error and instrument error is compared, and conclusion is drawn that operating error predominates in this process. Then frequently-used and available methods for processing data of double-theodolite anemometry are presented, and their applicability is analyzed. However, none of those methods can be suitable for use unconditionally while limited maximum error is given. Consequently, a joint method based on the advantages of all available methods is introduced, which can be applied to most comprehensive and extensive conditions. Finally, compared with base-line anemometry by using the observation data from Duhu at 10:00 on 16 August, 2007, and the advantage of the joint method is displayed.

Key words: double-treodolite anemometry, atmospheric observation, measurement error, precision

摘要： 从大气探空实验的特征出发，说明了使用双经纬仪测量数据反演探空气球空间坐标的过程中，人为操作误差远比仪器测量误差造成的影响严重。在此基础上总结了反演双经纬仪测量数据各种方法的优缺点，并分析了各反演方法在最大可信误差要求下的适用范围。分析得出各方法均只能适用于部分空间角度范围，因而提出了一种综合各基础方法优点而得到的联合反演方法，该方法的反演结果可在最大的空间角度范围内能满足最大可信误差的要求。最后通过都斛地区2007年8月16日10时的探空数据反演分析，说明了联合反演方法在反演精度和反演误差估算方面相对于经典的基线测风法的优越性。

关键词: 双经纬仪, 大气探测实验, 测量误差, 反演精度