5:45 PM - 6:00 PM
★ [BPT23-27] Statistical characterization of deep-sea sediments within the Minamitorishima EEZ by Independent Component Analysis
The extremely REY-rich mud has been probably produced by a combination of some processes that functioned in the abyssal ocean, and signatures of them could be preserved as characteristic geochemical compositions of the sediments. Therefore, as a first step to elucidate the REY-enrichment processes, it is quite important to decode geochemical signals derived from the sediments. To this end, multivariate statistical analyses are very useful because they can treat multi-elemental information concurrently and comprehensively.
Here we applied Independent Component Analysis (ICA) to the data set composed of elemental contents of the deep-sea sediments collected from the Minamitorishima EEZ. ICA is a relatively new computational statistical technique established in the past quarter century, which can extract original independent source signals or factors from observed signals on the basis of a fundamental assumption that the observed data consist of mutually independent source signals but do not constitute a joint Gaussian distribution (Hyvärinen et al., 2001). We report the results and interpretation of our new analysis, and statistically characterize the deep-sea sediments within the Minamitorishima EEZ.
-- References --
Fujinaga, K. et al. (2013) JpGU Meeting 2013.
Hyvärinen, A. et al. (2001) Independent Component Analysis. John Wiley and Sons.
Kato, Y. et al. (2013) JpGU Meeting 2013.
Suzuki, K. et al. (2013) JpGU Meeting 2013.