We investigated the suitability of the astronomical 15 GHz Very Long Baseline Array (VLBA) observing program MOJAVE-5 for estimation of geodetic parameters, such as station coordinates and Earth orientation parameters. We processed a concurrent dedicated VLBA geodesy program observed at 2.3 GHz and 8.6 GHz starting on September 2016 through July 2020 as reference dataset. We showed that the baseline length repeatability from MOJAVE-5 experiments is only a factor of 1.5 greater than from the dedicated geodetic dataset and still below 1 ppb. The wrms of the difference of estimated Earth orientation parameters with respect to the reference IERS C04 time series are a factor of 1.3 to 1.8 worse. We isolated three major differences between the datasets in terms of their possible impact on the geodetic results, i.e. the scheduling approach, treatment of the ionospheric delay, and selection of target radio sources. We showed that the major factor causing discrepancies in the estimated geodetic parameters is the different scheduling approach of the datasets. We conclude that systematic errors in MOJAVE-5 dataset are low enough for these data to be used as an excellent testbed for further investigations on the radio source structure effects in geodesy and astrometry.

我们调查了天文 15 GHz 甚长基线阵列 (VLBA) 观测程序 MOJAVE-5 用于估计大地测量参数（例如台站坐标和地球方向参数）的适用性。我们处理了从 2016 年 9 月到 2020 年 7 月在 2.3 GHz 和 8.6 GHz 观测到的并发专用 VLBA 大地测量程序作为参考数据集。我们表明 MOJAVE-5 实验的基线长度重复性仅比专用大地测量数据集大 1.5 倍，并且仍低于 1 ppb。估计地球方向参数相对于参考 IERS C04 时间序列的差异的 wrms 差 1.3 到 1.8 倍。我们在数据集对大地测量结果的可能影响方面隔离了三个主要差异，即调度方法，电离层延迟的处理和目标无线电源的选择。我们表明，造成估计大地参数差异的主要因素是数据集的不同调度方法。我们得出结论，MOJAVE-5 数据集中的系统误差足够低，这些数据可用作进一步研究大地测量学和天体测量学中射电源结构效应的极好测试平台。