11:30 AM - 11:45 AM
[MTT42-09] Towards quantitative evaluation of the chondritic Earth model: A combined geo-neutrino, -physics, and -chemistry study
Keywords:geoneutrino, chondritic Earth, crustal modeling
It is commonly assumed that the bulk Earth has chondritic relative abundances of refractory elements. Evaluating the validity of this long-standing paradigm is crucial for a wide range of issues in Earth science, such as the timing of Earth’s formation and differentiation and the structure of the mantle. Electron antineutrinos produced by natural radioactivity within the Earth, the so-called geoneutrinos, offer a unique opportunity to quantify the abundances of refractory-lithophile elements U and Th in the entire mantle, which in turn allows us to evaluate the chondritic Earth model. The geoneutrinos have been measured with a liquid scintillator detector at KamLAND in Japan since 2005. The uncertainty of the geoneutrino flux measurement is now reduced to a level small enough to set useful constraints on the mantle U and Th abundances, provided that the crustal contribution to the measured total geoneutrino flux is well established by determining U and Th distributions around the detector. Yet, the U and Th distributions in the Japan Arc crust, in particular deep crust, remains to be constrained quantitatively.
Towards quantitative evaluation of the chondritic Earth model, we have conducted a combined geo-neutrino, -physics, and -chemistry study. We first developed a methodological framework to construct a 3-D crustal composition model by combining geological, seismic, and geochemical data (Takeuchi et al., 2019 PEPI). In this method, we utilized the Bayesian inference technique to model the 3-D lithology map by combining seismic data as “observation” with a prior model based on a geologic map of local deep crustal exposure. Besides, we constructed U and Th abundance distribution models for different rock types based on published geochemical data for rock samples. We adopted the gamma distribution for the elemental abundance modeling, whereas previous studies adopted normal and log-normal distributions. Importantly, the gamma distribution modeling allows to estimate the unbiased mean from rock sample data that could be distorted due to detection limit and outliers. This point is crucial but often overlooked in mass balance calculations of geochemistry and geophysics, as detailed in a companion presentation by Enomoto et al. By convolving the 3-D lithology model with the rock U and Th distribution models, we constructed 3-D probability density function (PDF) models of U and Th abundances in the Japan Arc crust, and then calculated crustal geoneutrino flux. This new methodology enables constructing a fully probabilistic flux model, but the uncertainties are large at ~70%. This is due to the difficulty in handling of correlations among all the PDFs at different locations in the flux integration: We conservatively assumed maximum correlations, leading to the large uncertainties. An investigation into the correlations is in progress. In this presentation, we report the current status and prospects of our interdisciplinary study.
Towards quantitative evaluation of the chondritic Earth model, we have conducted a combined geo-neutrino, -physics, and -chemistry study. We first developed a methodological framework to construct a 3-D crustal composition model by combining geological, seismic, and geochemical data (Takeuchi et al., 2019 PEPI). In this method, we utilized the Bayesian inference technique to model the 3-D lithology map by combining seismic data as “observation” with a prior model based on a geologic map of local deep crustal exposure. Besides, we constructed U and Th abundance distribution models for different rock types based on published geochemical data for rock samples. We adopted the gamma distribution for the elemental abundance modeling, whereas previous studies adopted normal and log-normal distributions. Importantly, the gamma distribution modeling allows to estimate the unbiased mean from rock sample data that could be distorted due to detection limit and outliers. This point is crucial but often overlooked in mass balance calculations of geochemistry and geophysics, as detailed in a companion presentation by Enomoto et al. By convolving the 3-D lithology model with the rock U and Th distribution models, we constructed 3-D probability density function (PDF) models of U and Th abundances in the Japan Arc crust, and then calculated crustal geoneutrino flux. This new methodology enables constructing a fully probabilistic flux model, but the uncertainties are large at ~70%. This is due to the difficulty in handling of correlations among all the PDFs at different locations in the flux integration: We conservatively assumed maximum correlations, leading to the large uncertainties. An investigation into the correlations is in progress. In this presentation, we report the current status and prospects of our interdisciplinary study.