Search for: All records

It is well known that the axial dipole part of Earth’s magnetic field reverses polarity, so that the magnetic North Pole becomes the South Pole and vice versa. The timing of reversals is well documented for the past 160 Myr, but the conditions that lead to a reversal are still not well understood. It is not known if there are reliable ‘precursors’ of reversals (events that indicate that a reversal is upcoming) or what they might be. We investigate if machine learning (ML) techniques can reliably identify precursors of reversals based on time-series of the axial magnetic dipole field. The basic idea is to train a classifier using segments of time-series of the axial magnetic dipole. This training step requires modification of standard ML techniques to account for the fact that we are interested in rare events—a reversal is unusual, while a non-reversing field is the norm. Without our tweak, the ML classifiers lead to useless predictions. Perhaps even more importantly, the usable observational record is limited to 0–2 Ma and contains only five reversals, necessitating that we determine if the data are even sufficient to reliably train and validate an ML algorithm. To answer these questions we use several ML classifiers (linear/non-linear support vector machines and long short-term memory networks), invoke a hierarchy of numerical models (from simplified models to 3-D geodynamo simulations), and two palaeomagnetic reconstructions (PADM2M and Sint-2000). The performance of the ML classifiers varies across the models and the observational record and we provide evidence that this is not an artefact of the numerics, but rather reflects how ‘predictable’ a model or observational record is. Studying models of Earth’s magnetic field via ML classifiers thus can help with identifying shortcomings or advantages of the various models. For Earth’s magnetic field, we conclude that the ability of ML to identify precursors of reversals is limited, largely due to the small amount and low frequency resolution of data, which makes training and subsequent validation nearly impossible. Put simply: the ML techniques we tried are not currently capable of reliably identifying an axial dipole moment (ADM) precursor for geomagnetic reversals. This does not necessarily imply that such a precursor does not exist, and improvements in temporal resolution and length of ADM records may well offer better prospects in the future.

Paleomagnetic records from sediments, archeological artifacts, and lava flows provide the foundation for studying geomagnetic field changes over 0–100 ka. Late Quaternary time‐varying spherical harmonic models for 0–100 ka produce a global view used to evaluate new data records, study the paleomagnetic secular variation on centennial to multimillennial timescales, and investigate extreme regional or global events such as the Laschamp geomagnetic excursion. Recent modeling results (GGF100k and LSMOD.2) are compared to previous studies based on regional or global stacks and averages of relative geomagnetic paleointensity variations. Time‐averaged field structure is similar on Holocene, 100 ky, and million‐year timescales. Paleosecular variation activity varies greatly over 0–100 ka, with large changes in field strength and significant morphological changes that are especially evident when field strength is low. GGF100k exhibits a factor of 4 variation in geomagnetic axial dipole moment, and higher‐resolution models suggest that much larger changes are likely during global excursions. There is some suggestion of recurrent field states resembling the present‐day South Atlantic Anomaly, but these are not linked to initiation or evolution of excursions. Several properties used to characterize numerical dynamo simulations as “Earth‐like” are evaluated and, in future, improved models may yet reveal systematic changes linked to the onset of geomagnetic excursions. Modeling results are useful in applications ranging from ground truth and data assimilation in geodynamo simulations to providing geochronological constraints and modeling the influence of geomagnetic variations on cosmogenic isotope production rates.

Study of the late Quaternary geomagnetic field contributes significantly to understanding the origin of millennial‐scale paleomagnetic secular variations, the structure of geomagnetic excursions, and the long‐term shielding by the geomagnetic field. A compilation of paleomagnetic sediment records and archeomagnetic and lava flow data covering the past 100 ka enables reconstruction of the global geomagnetic field on such long‐term scales. We use regularized inversion to build the first global, time‐dependent, geomagnetic field model spanning the past 100 ka, namedGGF100k(GlobalGeomagneticField over the past100 ka). Spatial parametrization of the model is in spherical harmonics and time variations with cubic splines. The model is heavily constrained by more than 100 continuous sediment records covering extended periods of time, which strongly prevail over the limited number of discrete snapshots provided by archeomagnetic and volcanic data. Following an assessment of temporal resolution in each sediment's magnetic record, we have introduced smoothing kernels into the forward modeling when assessing data misfit. This accommodates the smoothing inherent in the remanence acquisition in individual sediment paleomagnetic records, facilitating a closer fit to both high‐ and low‐resolution records in regions where some sediments have variable temporal resolutions. The model has similar spatial resolution but less temporal complexity than current Holocene geomagnetic field models. Using the new reconstruction, we discuss dipole moment variations, the time‐averaged field, and paleomagnetic secular variation activity. The new GGF100k model fills the gap in the geomagnetic power spectrum in the frequency range 100–1,000 Ma⁻¹.