Plain language summary:
Rocks can preserve a primary record of Earth’s magnetic field from the time of their formation. However, this record is often obscured or replaced by secondary processes such as thermal and chemical overprinting. If primary and secondary magnetisation components are carried by minerals with similar magnetic properties, it can be difficult to determine which (if any) component is primary and whether a true magnetic field direction was preserved. We developed a method for mathematically unmixing components in paleomagnetic data through inverse modelling. Our method can quantify contributions from multiple magnetisation components and provide insights into how these components combine. We show that mixtures of primary and secondary components can generate spurious directional data and lead to false interpretations when analysed with conventional best-fitting line procedures. We demonstrate the efficacy of the method using real data from hematite-bearing sandstones in South China.
Paper:
Tonti-Filippini, J.A.D., & Gilder, S.A. (2023). Vector unmixing of multicomponent palaeomagnetic data. Geophysical Journal International, 233(3), 1632-1654.
https://doi.org/10.1093/gji/ggac505