Paleosecular variation analysis is a primary tool for characterizing ancient geomagnetic behavior and its evolution through time. This study presents a new high-quality directional dataset, paleosecular variation of the Paleogene (PSVP), with and without correction for serial correlation (SC), compiled from 1,667 sites from 45 different localities from the Paleogene and late Cretaceous (84 – 23 Ma). The dataset is used to study the variability, structure, and latitude dependence of the geomagnetic field during that period by varying selection criteria and PSV models. Modeled values for the equatorial virtual geomagnetic pole (VGP) dispersion have over-lapping uncertainty intervals within their uncertainty bounds between 8.3° and 18.6° 30 for the past 250 Ma. We investigate the suitability of two descriptive models of PSV, Model G-style quadratic fits and covariant GGP models, and find that both styles of model fail to satisfactorily reproduce the latitude dependent morphology of PSV,  but suggest that estimates of the equatorial VGP dispersion may still robustly characterize aspects of Earth’s long-term field morphology. During this time where the PSV behavior has not changed substantially, the reversal frequency has varied widely. The lack of a clear relationship between PSV behavior and reversal frequency is not trivially explained in the context of published findings regarding numerical geodynamo simulations.This manuscript is the version accepted for publication in G-Cubed on 16th May, 2024.

Numerical geodynamo simulations capture several features of the spatial and temporal geomagnetic field variability on historical and Holocene timescales. However, a recent analysis questioned the ability of these numerical models to comply with long-term palaeomagnetic field behaviour. Analysing a suite of 50 geodynamo models, we present here the first numerical simulations known to reproduce the salient aspects of the palaeosecular variation and time-averaged field behaviour since 10 Ma. We find that the simulated field characteristics covary with the relative dipole field strength at the core-mantle boundary (dipolarity). Only models dominantly driven by compositional convection, with an Ekman number (ratio of viscous to Coriolis forces) lower than $10^{-3}$ and a dipolarity in the range $0.34-0.56$ can capture the observed palaeomagnetic field behaviour. This dipolarity range agrees well with state-of-the-art statistical field models and represent a testable prediction for next generation global palaeomagnetic field model reconstructions.

Statistical studies of paleosecular variation (PSV) are used to infer the structure and behavior of the geomagnetic field. This study presents a new database, PSVM, of high-quality directional data from the Miocene era (5.3 – 23 Ma), compiled from 1,454 sites from 44 different localities. This database is used to model the latitude dependence of paleosecular variation with varying selection criteria using a quadratic form after Model G. Our fitted model parameter for latitude-invariant PSV (Model G a) is 15.7° and the latitude dependent PSV term (Model G b) is 0.23. The latitude invariant term is substantially higher than previously observed for the past 10 Myrs or any other studied era. We also present a new stochastic model of the time-average field, BB-M22, using a covariant giant Gaussian process (GGP) which is constrained using data from PSVM and Earth-like geodynamo numerical simulations. BB-M22 improves the fit to PSVM data relative to prior GGP models, as it reproduces the higher VGP dispersion observed during the Miocene. Our findings suggest a more variable magnetic field and more active geodynamo in the Miocene era than the past 10 Myrs, perhaps linked to stronger driving by elevated core-mantle heat flow. Although our results support that the average axial dipole dominance of the time-instantaneous field was lower than in more recent times, we note that based on inclination anomaly estimates cannot rule out that the Miocene time averaged field resembles a geocentric axial dipole.

The Mesozoic Dipole Low (MDL) is a period, covering at least ~80 million years, of low dipole moment that ended at the start of the Cretaceous Normal Superchron. Recent studies of Devonian age Siberian localities identified similarly low field values a few tens of million years prior to the Permo-Carboniferous Reverse Superchron (PCRS). To constrain the length and timing of this potential new dipole low, this study presents new paleointensity estimates from Strathmore (~411-416 Ma) and Kinghorn (~332 Ma) lava flows, UK. Both localities have been studied for paleomagnetic poles (Q values of 6-7) and the sites were assessed for their suitability for paleointensity from paleodirections, rock magnetic analysis, and microscopy. Thermal- and microwave-IZZI protocol experiments were used to determine site mean paleointensity estimates of ~3-51 μT (6-98 ZAm²) and 4-11 μT (9-27 ZAm²) from the Strathmore and Kinghorn localities, respectively. These, and all of the sites from 200-500 Ma from the (updated) PINT15 database, were assessed using the Qualitative Paleointensity criteria (Qᴘɪ). The procurement of reliable (Qᴘɪ ≥5), weak paleointensity estimates from this and other studies indicates a period of low dipole moment (median field strength of 17 ZAm²) for ~80 Myrs, from 332-416 Ma. This “Mid-Paleozoic Dipole Low (MPDL)” bears a number of similarities to the MDL, including the substantial increase in field strength near the onset of the PCRS. The MPDL also adds support to inverse relationship between reversal frequency and field strength and a possible ~200 million-year cycle in paleomagnetic behavior relating to mantle convection.

Understanding the evolution of Earth’s magnetic field can provide insights into core processes and can constrain plate tectonics and atmospheric shielding. The absolute paleointensity database PINT provides a curated repository of site mean, (i.e., cooling unit), estimates of the strength of the magnetic field. We present a minor update to the PINT database to version 8.1.0 by adding 248 records from 31 studies. The PINT database is used to define a continuous model of the dipole field, using an approach combining non-parametric and Monte Carlo resampling termed MCADAM. Three dipole field strength models spanning 50 ka to 3.7-4.2 Ga (MCADAM.1a-c) are presented, reflecting three tiers of increasingly more stringent data selection. The MCADAM models allow for the estimation of the magnetic standoff distance, constraining the shielding of Earth’s atmosphere against solar wind erosion provided by the geodynamo.