Comparison of Geomagnetic Models with Different Spatio-temporal Resolutions and Measured Data from Geomagnetic Stations

This study focuses on four INTERMAGNET geomagnetic stations located in the Arctic region and the South Atlantic Anomaly regions, where rapid geomagnetic changes have been observed in recent years. Utilizing two global geomagnetic field models with different spatial and temporal resolutions, the International Geomagnetic Reference Field (IGRF) model and the Swarm model, a multi-field source model that offers the highest spatiotemporal resolution currently available in China. The influence of model spatial resolution and update cycles on observatory measurements is quantitatively assessed. The findings are as follows. In regions without significant lithospheric magnetic anomalies, the IGRF and Swarm models yield relatively consistent results. However, for regions with more prominent lithospheric magnetic anomalies, the Swarm model with higher spatial resolution provides better agreement with the measurements. While the IGRF model showed significant temporal drift (often exceeding 100 nT) against station measurements during its five-year non-update period (from 2016 to 2020), the Swarm model exhibited no such drift. The magnitude of the IGRF drift varied regionally, reflecting the heterogeneity of global geomagnetic field variations. These results highlight the need to shorten the update interval for core field models during periods of rapid geomagnetic change. Statistical analysis from 120 global INTERMAGNET (International Real-time Magnetic Observatory Network) stations further confirms that, within the IGRF’s five-year update cycle, the Swarm model exhibits smaller mean and median absolute deviation compared to the IGRF model. The higher spatiotemporal resolution of the Swarm model yields results that are more consistent with observations and thus offers a more accurate representation of the geomagnetic field. This study provides a valuable basis and reference for the application of global geomagnetic models.