Defense
Student: Gelson Ferreira de Souza Junior
Program: Geophysics
Title: "Micropaleomagnetism"
Advisor: Prof. Dr. Leonardo Uieda
Judging Comitee:
- Prof. Dr. Leonardo Uieda - Presidente e Orientador - IAG/USP
- Prof. Dr. Greig Paterson - University of Liverpool (por videoconferência)
- Prof. Dr. Wilbor Poletti - UFVJM (por videoconferência)
- Prof. Dr. Vanderlei C. Oliveira Jr. - Observatório Nacional (por videoconferência)
- Prof. Dr. Wyn Williams - University of Edinburgh (por videoconferência)
Other Members:
- Prof. Dr. Ricardo Ivan Ferreira da Trindade - IAG/USP
- Prof. Dr. Carlos Aberto Mendonça - IAG/USP
- Prof. Dr. Lesleis Nagy - University of Liverpool
- Prof. Dr. Paul Yves Jean Antonio - University of Montpellier
- Prof. Dr. Gelvan Hartmann - Unicamp
Abstract:
The geomagnetic field plays a fundamental role in the maintenance of life on our planet, because it shields Earth's atmosphere against solar wind erosion. Studying its ancient direction and intensity remains one of the few quantitative windows into the long-term evolution of the geodynamo and paleogeographic reconstructions. Conventional paleomagnetic methods rely on bulk measurements of centimeter-scale specimens, which average the response of magnetic carriers with very different grain sizes and stabilities and offer no direct way to identify and isolate the reliable recorders of this information. New magnetic microscopy techniques, such as Quantum Diamond Microscopy (QDM), provide the means to resolve the magnetic signal of the rocks at the grain scale rather than the sample as a whole, opening a new avenue of investigation, which we call here Micropaleomagnetism. However, the use of QDM as a quantitative paleomagnetic tool has been limited by two open problems: (i) recovering the position and magnetic moment of each source from a field map is an inherently non-unique inverse problem, typically resolved only with the aid of additional measurements; and (ii) classical single-domain theory suggested that millions of grains would be required to average out the stochastic nature of thermoremanent magnetization (TRM), a population far beyond what magnetic microscopy can image. This thesis addresses the first problem using Euler deconvolution (ED), a technique adapted from potential-field geophysics, as an efficient source of prior information to avoid the ambiguity of the magnetic inversion. A semi-automated algorithm is first developed to estimate the position and dipole moment of individual magnetic sources directly from magnetic microscopy data. Then it is extended with a non-linear refinement and iterative signal-removal scheme to tackle the mutual interference between closely spaced sources that limited the original approach's accuracy. This workflow is further generalized to a multipolar formulation capable of accommodating the high-order contributions when the dipolar approximation fails, by using spherical harmonics with ED results as the center of expansion. The second problem is addressed by applying the progressively refined pipeline to real data to test whether the long-standing statistical objection to grain-scale paleomagnetism still holds. By using high-resolution QDM data from TRM-bearing samples, both paleodirections and absolute paleointensities are recovered from populations of only hundreds to thousands of magnetic anomalies, in agreement with bulk measurements. Together, these results enable the field of Micropaleomagnetism, in which magnetic microscopy can be used as a paleomagnetic tool and establish a practical pathway for reconstructing the ancient field from small, rare, or otherwise irreplaceable materials, including extraterrestrial samples, for which conventional bulk measurements are not an option.
Keywords: Paleomagnetism, Magnetic Microscopy, Magnetic Data Inversion, Quantum Diamond Microscope, Paleointensity