We investigated the (paleo)latitudinal dependence of angular dispersion sets of virtual geomagnetic poles (VGPs) for the Permian‐Carboniferous Reversed Superchron (PCRS: 262–318 Ma). In order to analyze the paleosecular variation during this period, we prepared different paleomagnetic data sets from scientific databases and recent literature, based on selection criteria which provided high degree of refinement. Model G of McFadden et al. (1988) was fitted to the VGP dispersion data, providing the shape parameters a and b, which were further compared with similar results for the Cretaceous Normal Superchron (CNS) and other periods of different geomagnetic reversal rates throughout the last 3 Ga. Our results indicate high similarity between the angular VGP dispersion from the PCRS and CNS (a low VGP dispersion at low paleolatitudes, and strong paleolatitudinal dependence), in contrasting difference for periods of higher reversal rates (the last 5 Ma). Despite the geodynamic differences related to both Phanerozoic superchrons, such evidence could point out similar heat flux conditions in the CMB, which may favor compatible stability conditions throughout these magnetozones. Notably, two additional observations which arose from the shape parameters relations for both Phanerozoic superchrons and Precambrian data sets demand further investigation: (i) a pattern of increase for the b/a ratio throughout the 0.5–1.6 Ga interval was observed, which could be partially explained by the Maya Superchron, recently proposed by Gallet et al. (2012) and (ii) the b × a relationship for both Phanerozoic superchrons and the 2.45–2.82 Ga interval pointed to the presence of one or more long‐term magnetozones throughout the Paleoproterozoic.

Spatial and temporal geomagnetic field variations have been observed over different geological timescales. Ancient field measurements, mainly obtained from geological materials (sedimentary and igneous rocks), allow investigations of directional and intensity variability of the paleomagnetic field that result from processes operating in Earth's fluid core (see, e.g., Hulot et al., 2010). Particularly, information about paleosecular variation (PSV), long-term variations of the order of 10 5 -10 6 years (e.g., Johnson & McFadden, 2015), is essential to better understand temporal geomagnetic field evolution and to constrain numerical geodynamo models (

We provide an evaluation of the paleolatitudinal dependence of the paleosecular variation throughout the Paleozoic-Mesozoic transition – linked to the high geomagnetic reversal frequency interval Illawarra Hyperzone of Mixed Polarity (IHMP; ~266.7–228.7 Myr). Our findings were compared with those for intervals of distinctive geomagnetic reversal frequencies within the Phanerozoic. Our results for the IHMP were conducted through estimates of angular dispersion (SB) of virtual geomagnetic pole (VGP) data groups, taken from a high quality paleomagnetic database. Model G was fitted to these data, providing its shape parameters a and b (respectively related to the antisymmetric and symmetric harmonic terms for the time-average geomagnetic field). Results for the IHMP exhibited compatible patterns with two well-known intervals of higher reversal frequency – Jurassic and the last 5 Myr. A comparison of b/a ratio results – considered as an efficient indicator for the relative contribution of the axial dipole field – for the last 270 Myr, indicated an inverse correspondence with the relative core-mantle boundary (CMB) heat flux, according to recent discussions, clarifying the physical meaning of the Model G shape parameters a and b.

Summary Paleosecular variation determinations and studies of the geometry of the Earth’s main magnetic field provide important information about the field evolution, and to constrain numerical geodynamo models. Paleomagnetic directional data from lava flows over the last few million years is of particular interest because the regional and global tectonic effects are minimal. However, the distribution of this type of paleomagnetic data is uneven where the Southern Hemisphere is the destitute side. Therefore, the better knowledge of the geomagnetic field behavior depends on the increased availability of high-quality data, especially in the Southern Hemisphere. A paleosecular variation (PSV) and the time-averaged field (TAF) study was then performed in 0-5 Ma lava flows from the Caviahue-Copahue Volcanic Complex located in Northern Patagonia, Argentina (37○0’S, 71○10’W). The magnetic mineralogy of lava flows was investigated through thermomagnetic susceptibility curves, isothermal remanent magnetization (IRM) acquisition curves, hysteresis loops, and first-order reversal curves (FORCs). Samples are essentially comprised of titanomagnetite with different Ti contents and magnetic domain structures typical of vortex state particles. A total of 50 volcanic sites were sampled, which provided 42 reliable paleomagnetic site-mean directions after alternating field and thermal demagnetization. From these 42 sites, 36 are of normal and 6 are of reversed polarity. The mean direction from normal (reversed) sites is D = 356.2○, I = −50.1○, α95 = 4.0○, N = 36 (D = 176.5○, I = 59.5○, α95 = 14.1○, N = 6). Using only site-level data with the precision parameter k ≥ 100, we obtain 26 paleomagnetic sites for PSV and TAF investigations in the study region. The filtered dataset has a mean direction (D = 354.4○, I = −53.2○, α95 = 5.1○) close to the expected direction for a geocentric axial dipole (GAD) field (IGAD = −57.3○). The paleopole (Plat = 84.4○, Plon = 229.1○, A95 = 5.7○) coincides with the Earth’s spin axis within the 95% confidence interval. Virtual geomagnetic pole scatter ($S_{B} = {15.8^{18.9}_{11.8}}^{\circ }$) and the inclination anomaly ($\Delta I = 4.1_{-1.0}^{{9.2}^{\circ }}$) are both consistent at the 95% confidence level with recent PSV and TAF models, respectively. Our results support the presence of small non-dipole field contributions ($< 3\%$) superimposed on the GAD term, as reported by South American studies at mid southern latitudes.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.