The global paleointensity database for 65–200 Ma was analyzed using a modified suite of paleointensity quality criteria (QPI) such that the likely reliability of measurements is assessed objectively and as consistently as possible across the diverse data set. This interval was chosen because of dramatic extremes of geomagnetic polarity reversal frequency ranging from greater than 10 reversals per million years in the Jurassic hyperactivity period (155–171 Ma) to effectively zero during the Cretaceous Normal Superchron (CNS; 84–126 Ma). Various attempts to establish a relationship between the strength of Earth's magnetic field and the reversal frequency have been made by previous studies, but no consensus has yet been reached primarily because of large uncertainties in paleointensity estimates and sensitivity of these estimates to data selection approaches. It is critical to overcome this problem because the evolution of the dipole moment is a first order constraint on the behavior of the geodynamo. Here we show that conventional statistical tests and Bayesian changepoint modeling consistently indicate the strongest median/average virtual dipole moment during the CNS. In addition, the CNS and Jurassic hyperactivity period are characterized by the highest and lowest percentage of virtual dipole moments exceeding the overall median for the 65‐ to 200‐Ma interval, respectively. These observations suggest that the superchron dynamo was able to generate stronger fields than the dynamo operating in the frequently reversing regime. While the precise mechanism remains unclear, our results are compatible with the hypothesis that field strength and reversal rate variation are controlled by changes in core‐mantle boundary thermochemical conditions.
1 March 2019 in Geochemistry, Geophysics, Geosystems
Nearly three decades ago paleomagnetists suggested that there existed a clear link between latitude dependence of geomagnetic paleosecular variation (PSV) and reversal frequency. Here we compare the latitude behavior of PSV for the Cretaceous Normal Superchron (CNS, 84–126 Ma, stable normal polarity) and the preceding Early Cretaceous‐Jurassic interval (pre‐CNS, 126–198 Ma, average reversal rate of ~4.6 Myr−1). We find that the CNS was characterized by a strong increase in the angular dispersion of virtual geomagnetic poles (VGPs) with latitude, which is consistent with the results of earlier studies, whereas the VGP dispersion for the pre‐CNS period was nearly invariant with latitude. However, the PSV behavior for the last 5 or 10 million years (average reversal frequency of ~4.4–4.8 Myr−1) shows that the latitude invariance of VGP scatter cannot be considered as a characteristic feature of a frequently reversing field and that a strong increase in VGP dispersion with latitude was not restricted to the long periods of stable polarity. We discuss models describing the latitude dependence of PSV and show that their parameters are not reliable proxies for reversal frequency and should not be used to make inferences about the geomagnetic field stability. During the pre‐CNS interval, the geodynamo may have operated in a regime characterized by a high degree of equatorial symmetry. In contrast, more asymmetric geodynamos suggested for 0–10 Ma and the CNS were evidently capable of producing a very wide range of reversal frequencies.
We present absolute paleointensity results obtained from a collection of samples from ~250 Ma Kuznetsk Traps (Kuznetsk Depression, Southern Siberia). In addition to similar age these rocks display geochemical signatures similar to those reported for basalts of Siberan Traps and represent the southernmost affinity of the latter.The primary nature of magnetic remanence in studied rocks was established by previous paleomagnetic studies Rock magnetic analysis indicates that the main magnetic mineral is titanomagnetite in a predominantly single-domain state with Curie temperatures between ~275 and 350⁰C. Scanning electron microscopy showed that titanomagnetite grains range in size from 0.5 to 1 μm. Individual grains are separated from each other and “sealed” within silicate matrix, which largely predetermined perfect preservation of primary mineral textures.Paleointensity estimates were obtained using the Coe-version of Thellier-Thellier double-heating protocol with partial TRM checks. 36 samples (5 sites taken along the Tom River) yielded straight Arai-Nagata diagrams within temperature interval between 100 to 275⁰C. The average paleointensity value obtained from these samples was calculated at 12.7 ± 1 µT (with a factor q of about 11) with corresponding VDM=2.1 ± 0.2 × 1022 Am2.Arai-Nagata diagrams for 20 samples from two other sites (collected in quarries on the Karakansky ridge) display more complex behavior. Straight linear segment of Arai plots between ~100 and 300⁰C yielded an average paleointensity value of 44 ± 1 μT, which corresponds to a VDM=7.0 ± 0.1 × 1022 Am2. However on higher temperatures, NRM vs. TRM data have a trend of flattening that increase in artificial TRM is accompanied by no loss of NRM. We interpret this observation as a result of laboratory-induced thermochemical alteration, namely, unmixing of homogeneous Ti-magnetites into Ti-rich and Fe-rich phases, with the latter phase responsible for such NRM-lost vs. TRM-gained behavior. Thermomagnetic analyses on these samples indicated mineralogical changes that set approximately at 300⁰C, supporting our interpretation. However, reversible thermomagnetic curves and p-TRM checks within 10% from initial pTRM, below ~300⁰C suggest that paleointensities determined between ~100 and 300 ⁰C of Arai plots are trustworthy.We will discuss our results and reasons for such radical differences between paleointensity estimates obtained from the same suite of rocks sampled at different locations.The study was supported by the Russian Foundation for Basic Research, grant No. 18-05-00234 and the Russian Science Foundation, grant No. 19-17-00091.