[S-EM15] Geomagnetism and paleomagnetism

[SEM15-P01] Difference of relative paleointensity recording efficiency in magnetic mineral constituents in a sediment core off Chile

*Toshitsugu Yamazaki¹, Jiaxi Li ¹, Takaya Shimono², Toshiya Kanamatsu³ (1.Atmosphere and Ocean Research Institute, The University of Tokyo, 2.KAIYO ENGINEERING CO. LTD., 3.Japan Agancy for Marine-Earth Science and Technology)

Keywords:paleomagnetism, relative paleointensity, magnetofossil

Progress of relative paleointensity (RPI) estimations using marine sediments have contributed to better understanding of the behavior of the geomagnetic field in the past. To enhance further the reliability of RPI estimations, we must overcome the problem that magnetic-property variations of magnetic-mineral assemblages in sediments may contaminate RPI records (e.g., Yamazaki et al., 2013). Two major constituents of magnetic-mineral assemblages in marine sediments are magnetofossils and terrigenous magnetic minerals, and the latter consists of silicate-hosted magnetic inclusions and unprotected magnetic minerals. It is considered that different RPI recording efficiencies among those components are responsible for the contamination. We have been conducting research for understanding these differences. Using sediments in the western equatorial Pacific, we revealed that RPI recording efficiency of magnetofossil is lower than the terrigenous counterpart, and that increasing proportion of magnetofossil causes underestimation of RPI (Inoue et al., 2021; Li et al., 2022). As the ARM/SIRM ratio can be a proxy of the proportion of magnetofossil, inverse correlation between RPI and ARM/SIRM arises. This conclusion needs to be tested using deep-sea sediments in various oceanographic and sedimentological environments.

We have further studied this issue using a sediment core (MR0806-PC3) taken from the southeast Pacific Ocean off Chile. The 19-m long core preserves a paleomagnetic record since ~1.4 Ma. Rock-magnetic investigations including IRM component analyses and FORC diagrams revealed that the magnetic mineral assemblage of this core is a mixture of low-coercivity magnetofossils and middle-coercivity partially oxidized detrital magnetite, and the coercivity distribution of the two components little overlap. Thus we could successfully separate RPI signals carried by the two components by calculating RPI from the gradients of 20-40 mT and 70-160 mT segments in NRM-ARM and NRM-IRM demagnetization diagrams. We have confirmed that RPI recording efficiency of magnetofossil is lower than the terrigenous component also in this core. Because the coercivity ranges of the two components little overlap, changes in the proportion of the two components do not influence RPI estimations. This condition is ideal for RPI estimations, and the RPI curve obtained from this core closely coincide with that of the PISO-1500 stack despite changes in ARM/SIRM.