WHUEMM: global core field modelling from multiple satellites and comparative evaluation of inversion strategy

Read the full article See related articles

Listed in

This article is not in any list yet, why not save it to one of your lists.
Log in to save this article

Abstract

Measurements from low Earth orbit satellites play an important role in modern geomagnetic field modelling. In this study, we present two geomagnetic field models, WHUEMM-S, derived by sequential inversion, and WHUEMM-C, derived by comprehensive inversion. Both models are constructed calibrated Swarm A/B, GRACE-FO 1, and CryoSat-2 observations collected between January 2019 and July 2024. Both models represent the core field with degree 15 spherical harmonics and sixth order B-splines. This study assesses the impact of these inversion strategies and evaluates the value of non-dedicated satellites in geomagnetic field modelling. Power spectral analysis shows that both models produce a temporally stable main field (MF) energy and secular variation (SV) energy, with differences from CHAOS-7.18 of about 1 nT² and 1 (nT/year)² for spherical harmonic degrees below 6. Stronger regularization damping in WHUEMM causes a sharp decrease in secular acceleration (SA) at degrees above 7. WHUEMM-C departs from CHAOS-7.18 mainly in the axial dipole and a few low-order sectoral terms, whereas the high-degree misfits in WHUEMM-S are probably driven by spectral truncation and residual external signals. Global MF maps confirm that both models reproduce mid- and low-latitude features well; however, at high latitudes WHUEMM-S deviates more from CHAOS-7.18 than WHUEMM-C does. SV derived from observatory records confirm that each model maintains smooth temporal end points and reliably captures long-term trends. This demonstrates that carefully calibrated, non-dedicated data from GRACE-FO 1 and CryoSat-2 be used to build global geomagnetic models without compromising robustness. Finally, using WHUEMM-S as the parent model, we produced and submitted three IGRF-14 candidate models.

Article activity feed