基于理论分析提出地倾斜视方位角及视偏角概念，以1987~2021年北京延庆西拨子和甘肃兰州十里店水管倾斜仪5个测向分量观测数据为例详细介绍了视偏角计算过程，并对2020年全国地倾斜台网715个测向分量观测数据的视偏角进行计算。2个典型台站结果显示，西拨子NS向视偏角P_NS=－0.1°，EW向视偏角P_EW=9.6°，NE向视偏角P_NE=12.4°；十里店NS向视偏角P_NS=0.3°，EW向视偏角P_EW=－22.5°；2个台站视偏角的最大值为23.2°，最小值为－36.0°；从5个分量的统计均值来看，视偏角最大值为12.4°，最小值为－22.5°，极差为34.9°；2个台站均显示出NS向视偏角偏小、EW向和NE向视偏角偏大的特征。全国地倾斜台网观测数据计算结果表明，视偏角P_S＜10°的分量数为266个，占37.20%，P_S＜22.5°的分量数为445个，占62.24%，P_S＞45°的分量数为111个，占15.52%；视偏角离散误差M_PS＜10°的分量数为247个，占34.55%，M_PS＜22.5°的分量数为446个，占62.38%，M_PS＞45°的分量数为43个，占6.01%。基于本文研究结果，建议对视偏角P_S＞20°的测向分量的起始方位角进行重点核查，对于视偏角离散误差M_PS＞20°的测向分量的观测数据需认真筛查。

Based on theoretical analysis, we propose the concepts of apparent azimuth and apparent declination of ground tilt, and introduce the procedure of calculating the apparent declination. We use the observed data of five measurement components of the water pipe tiltmeter from 1987 to 2021 in Beijing Yanqing Xibozi and Gansu Lanzhou Shilidian stations as examples. In turn, we calculate the results for the apparent declination of the 715 measured component observations of the national ground tilt network in 2020. Results from two typical stations show that: the Xibozi station NS direction angle of apparent declination P_NS=－0.1°; EW direction angle of apparent declination P_EW=9.6°; NE direction angle of apparent declination P_NE=12.4°; the Shilidian station NS direction angle of apparent declination P_NS=0.3°; EW direction angle of apparent declination P_EW=－22.5°; the maximum apparent deflection angle of 2 stations is 23.2°, the minimum value is －36.0°. From the statistical mean of the 5 components, the maximum value of the apparent deviation angle is 12.4°, the minimum value is －22.5°, and the range is 34.9°. Both stations show that the number of apparent deflections in the NS direction is relatively small, while the number of apparent deflections in the EW direction and NE directions is relatively large. The calculation results of the observed data from the national earth tilt network indicate that there are 266 components with an apparent deviation angle P_S<10°, accounting for 37.20%, the number of components with an apparent deviation angle of P_S<22.5° is 445, accounting for 62.24%, and the number of components with P_S>45° is 111, accounting for 15.52%; the number of components with apparent deviation angle dispersion error M_PS<10° is 247, accounting for 34.55%, the number of M_PS<22.5° is 446, accounting for 62.38%, and the number of components with M_PS>45° is 43, accounting for 6.01%. Based on the results of this study, we recommend that the initial azimuth angle of the measurement component with an apparent deviation number greater than 20° should be emphatically checked; also we should carefully screen the observation data of the measurement component with an apparent deviation angle dispersion error M_PS>20°.