Evaluating data quality and reference instrument robustness: insights from 12 years DI magnetometer comparisons in the Geomagnetic Network of China

Abstract. A statistical analysis was conducted on 12 years of geomagnetic instrument comparison data from the Chinese Geomagnetic Network (GNC). The study reveals that when applying a 90 % threshold, the corresponding instrument deviation thresholds are 0.21 ′ (D component) and 0.11′ (I component), which can serve as evaluation criteria at the network level. By integrating multi-source uncertainty decomposition (including instrument errors, operator-dependent errors, and pillar correction errors) and weighted ensemble analysis, systematic deviations between reference fluxgate theodolites and test instruments were quantified. Results demonstrate that reference instruments exhibit high stability and reliability, with mean deviations of -0.004 ′ (D) and 0.022 ′ (I), both within the 95 % confidence interval, and no long-term drift was observed. Operator-dependent errors were successfully isolated, with 0.13 ′ (D) and 0.06 ′ (I), consistent with observed experimental findings, confirming that operator-dependent errors constitute the primary contributor to instrument deviations. Notably, operator-dependent errors in D are significantly higher than I due to the complexity of azimuth alignment. These findings highlight the critical role of instrument comparisons in effectively monitoring equipment performance and assessing observational quality across stations, while validating the feasibility of standardized instrument evaluation methods and the long-term stability of reference instruments. Future efforts should integrate sensors and automation technologies to minimize human errors, thereby providing a higher-quality data foundation for geophysical studies.