11:45 AM - 12:00 PM
[PPS08-11] Unsupervised Classification and Geographical Correlation of Near-Lunar Electron Energy Spectra Observed by Kaguya (SELENE)
Keywords:Moon, Kaguya Satellite, Plasma Environment, Automated Classification, Electron Energy Spectra
From the 1960s through the 1970s, extensive observations of low-energy charged particles around the Moon were conducted by satellites in lunar orbit and plasma instrumentation installed on the lunar surface, leading to numerous new discoveries. The plasma environment of the Moon and its surroundings involves complex interactions, giving rise to various phenomena such as changes in surface potential, the influence of magnetic fields, and plasma dynamics. Japan’s lunar explorer “Kaguya (SELENE),” which operated at low altitude, provided crucial data for understanding these phenomena. Among its onboard instruments, the Electron Spectrum Analyzer (ESA) significantly contributed to clarifying physical parameters, including magnetic field strength and surface potential, by observing electron dynamics near the lunar surface and analyzing electron energy flux spectra.
In this study, we performed a multifaceted analysis of electron behavior in the lunar plasma environment using electron energy spectrum data acquired by Kaguya (SELENE). Specifically, we applied an unsupervised classification approach—combining Principal Component Analysis (PCA) with k-means clustering—to systematically categorize electron energy flux spectra. As a result, multiple clusters, each characterized by factors such as the prominence of the loss cone, local magnetic field strength, and day-night conditions, were extracted, and each cluster exhibited a distinct spatial distribution on the lunar surface coordinate system.
Furthermore, by leveraging multi-directional energy spectra obtained by the onboard ESA, we evaluated the pitch-angle distribution of electrons and the degree of reflection and scattering from the lunar surface. Our findings include a marked loss cone structure at higher energies in regions with strong crustal magnetic fields, while clusters showing a drastic reduction in flux from the lunar side appear during nighttime conditions, when solar wind collisions and photoelectron supply are limited. These results provide important insights for comprehensively understanding the distribution of electron energy in relation to geographic conditions and magnetic field strength around the Moon. Looking ahead, the outcomes of this research may serve as a basis for developing new approaches and observation strategies in future lunar exploration missions and large-scale data analysis methods.
In this study, we performed a multifaceted analysis of electron behavior in the lunar plasma environment using electron energy spectrum data acquired by Kaguya (SELENE). Specifically, we applied an unsupervised classification approach—combining Principal Component Analysis (PCA) with k-means clustering—to systematically categorize electron energy flux spectra. As a result, multiple clusters, each characterized by factors such as the prominence of the loss cone, local magnetic field strength, and day-night conditions, were extracted, and each cluster exhibited a distinct spatial distribution on the lunar surface coordinate system.
Furthermore, by leveraging multi-directional energy spectra obtained by the onboard ESA, we evaluated the pitch-angle distribution of electrons and the degree of reflection and scattering from the lunar surface. Our findings include a marked loss cone structure at higher energies in regions with strong crustal magnetic fields, while clusters showing a drastic reduction in flux from the lunar side appear during nighttime conditions, when solar wind collisions and photoelectron supply are limited. These results provide important insights for comprehensively understanding the distribution of electron energy in relation to geographic conditions and magnetic field strength around the Moon. Looking ahead, the outcomes of this research may serve as a basis for developing new approaches and observation strategies in future lunar exploration missions and large-scale data analysis methods.