"BiCEP and TROUT: Objective methods for reanalysis of MagIC measurement data" by Brendan Cych, University of Liverpool
Talk presented at the 2023 Magnetics Information Consortium (MagIC) Workshop: Magnetism and Earth History: Field Evolution, Environmental Change and Paleogeography, Feb 28th-Mar 2nd, 2023. Convened at the Scripps Institute of Oceanography, UCSD in La Jolla, California and sponsored by the National Science Foundation (NSF).
Workshop website: [ Ссылка ]
Talk Abstract:
Paleomagnetic data can be used to make inferences about the Earth core dynamics and plate tectonics over geological timescales. However, many natural samples contain non-ideal magnetic recorders which do not record a consistent magnetization over these timescales. These rocks may produce data which are difficult to interpret. The paleomagnetic community’s understanding of non-ideal recorders has improved over time, potentially changing the interpretation of legacy data. However, these data are not always publicly available, leading researchers to include/exclude data based on qualitative criteria such as the Q criteria (doi:10.1016/0040-1951(90)90116-P) for paleomagnetic poles or the QPI criteria (doi:10.3389/feart.2014.00024) for paleointensities. The MagIC database enables researchers to make their data publicly available for reanalysis. Here we present two statistical methods which can be used to reanalyze hard to interpret legacy paleomagnetic data.
The Thermal Resolution of Unblocking Temperatures method (TROUT, Cych et al. in prep) is an approach for analyzing multi-component demagnetization data where the unblocking temperature or coercivity distributions of the magnetic components overlap. TROUT can automatically determine temperature steps to use to isolate a magnetization acquired in a single field for paleodirection and intensity experiments. Additionally it can be used to automatically find the temperature to which a rock was reheated or rotated, for use in studies of pyroclastic emplacement temperatures and baked contact tests.
Bias Corrected Estimation of Paleointensity (BiCEP, doi:10.1029/2021GC009755) is a new approach for estimating paleointensities from Thellier-type experiments. Rather than excluding specimens based on whether they pass or fail a set of subjectively chosen “Selection Criteria”, BiCEP assumes an intrinsic relationship between a the “Arai plot curvature” criterion and bias, and uses this to make an unbiased estimate of the paleointensity. BiCEP produces statistically robust error bars which depend on the number of specimens used for the paleointensity estimate and the quality of the data, allowing for consistent analyses of legacy paleointensities.
