Ryosuke Nakamura (Leader) and Yoshiaki Ishihara (Researcher), Geoinformatics Research Group, the Information Technology Research Institute (Director: Satoshi Itoh) of the National Institute of Advanced Industrial Science and Technology (AIST; President: Tamotsu Nomakuchi), have discovered the relic of a massive impact on the moon which could be responsible for the differences in the topography on the near and far sides of the moon. They examined more than 20 billion visible and infrared reflectance spectra collected by the lunar explorer “Kaguya” at approximately 70 million sites across the entire surface of the moon by using a data-mining method.

On the moon, there are regions called "maria" (Latin for "seas"), where visible reflectance is low and a small number of craters exist, and regions called "highlands," where visible reflectance is high and many craters exist. "Maria" are prevalent on the near side of the moon (the side that faces the earth) but they are rarely found on the far side. In addition, the far side of the moon has a higher altitude and thicker crust than the near side. This "dichotomy" between the near and far sides of the moon is thought to be due to a massive impact during the early stages of moon formation that removed much of the crust material comprising the "highlands" on the near side of the moon. This study used a data-mining method to analyze visible and infrared reflectance spectra in order to examine the distribution of low-calcium pyroxene, a substance highly prevalent in impact melt ejecta. As a result, the researchers discovered circular distribution of low-calcium pyroxene having a diameter of 3,000 km that corresponded to the Procellarum basin on the near side of the moon. This type of distribution is thought to be due to impact melt ejecta, and this was the first compositional evidence  that could support a massive impact at the early stages of moon formation.

Ever since the moon was formed, it has always been adjacent to the earth. Therefore, an explanation of how the moon was born could lead to understanding the early stages of the earth formation. Also, this analysis procedure can be applied to data collected by earth observing satellites, for mineral resource exploration, environmental monitoring, etc.

Details of the results will be published online in a scientific journal, Nature Geoscience, on October 29, 2012 (JST).

Left: Near side (shown at left) and far side (shown at right) of the moon The dichotomy of the moon is evident from these visible images. Right: The red circles show the distribution of low-calcium pyroxene rich impact melt ejecta The relic of the massive impact, which is difficult to perceive visually, can be identified.

Studying the origin and evolution of the earth is important for resource exploration. However, there is hardly any geological information from 4 billion or more years ago that still remains on the earth. Meanwhile, rocks formed 4 billion or more years ago have been relatively well-preserved on the moon, which has always been next to the earth. Therefore, it is possible to glean information on the initial stages of earth formation by closely studying the history of the moon.

Remote sensing by spacecrafts are an indispensable tool for the global study of the moon and the earth. Recent hyperspectral observations yield an enormous quantity of remote sensing data. There is a growing need for data analysis methods capable of extracting useful information from this massive amount of data.

The Information Technology Research Institute of AIST has been developing data analysis tools for analyzing big data on the ”Cloud".  Meanwhile, the Institute of Geology and Geoinformation of AIST has been conducting research over many years on mineral resource exploration using visible and infrared reflectance spectra obtained by earth observation satellites. Both of these research and development efforts were integrated and applied to data obtained by the lunar explorer Kaguya in order to study the surface composition of the moon.

The most plausible theory as to the origin of the moon is the “giant impact” theory, which states that a massive celestial body collided with the earth and the fragments generated by that collision aggregated to form the moon. According to this theory, the surface of the newly formed moon was covered by a magma ocean. The "highlands" of the moon are comprised of rocks solidified from this magma ocean. Meanwhile, it is believed that the "maria (seas)" were formed by lave eruption from the inside after the formation of the “highlands”. As shown in Fig. 1, the dark "maria" are spread widely over the near side of the moon but are hardly seen on the far side. Also, previous explorations showed that not only does the ratio of "maria" to "highlands" completely differ between the near and far sides of the moon, but so does the thickness of the crust and the distribution of radioactive elements. This asymmetry of the near and far sides of the moon is called a "dichotomy," and its cause is unknown. The following hypothesis to explain this dichotomy was proposed in the past: A massive collision of a celestial body with the near side of the moon blasted away the material of the highlands, forming a massive impact basin (Procellarum basin shown on the left in Fig. 1) with a diameter of 3,000 km and creating the dichotomy of the moon. However, no compositional evidence has been found.

Figure 1 : Near and far sides of the Moon observed by "Kaguya" The slightly dark region on the southern far side, the South Pole-Aitken basin, is the one of the largest impact basins in the solar system. The "Mare Imbrium (Sea of Showers)" on the near side was generated by a massive impact approximately 3.8 billion years ago. There has been a hypothesis that Procellarum basin could be an impact basin.

As shown in Fig. 2, the representative minerals on the surface of the moon have characteristic visible and infrared reflectance spectra. The Japanese lunar explorer Kaguya launched in 2007 is equipped with an observational instrument, “Spectral Profiler”, capable of measuring the reflectance spectra of these minerals in order to survey the composition of the surface of the moon. Between December 2007 and June 2009, the Spectral Profiler obtained more than 20 billion reflectance spectra at approximately 70 million sites on the surface of the moon.

AIST established a decision tree algorithm to classify the massive quantity of hyperspectral data. This study focused on lithologies containing approximately 20% or more low-calcium pyroxene (LCP) and closely examined the global distribution on the moon (Fig. 3). While LCP is a common mineral contained in various rocks, the observed LCP-rich exposures are supposed to have been created by recrystallization of huge impact melt pool. Furthermore, LCP was found to be a dominant component in impact melt ejecta from the "Mare Imbrium (Sea of Showers)" collected by Apollo program. The laboratory spectrum shows a perfect match with those of LCP-rich sites measured by Kaguya.

Figure 2 : Reflectance spectra of representative minerals measured in a laboratory

As shown in Fig. 3, the LCP-rich exposures are concentrated in three regions: the rim of the Mare Imbrium, the inside of the South Pole-Aitken basin, and the rim of the Procellarum basin covering the near side of the moon. The topography of  the Mare Imbrium and the South Pole-Aitken basin indicates that they are impact basins with diameters of 1,000 and 2,500 km, respectively. The basin-forming impact events could have melted not only the crust but also mantle material. LCP was likely produced through the differentiation of the huge impact melt pool. Similarly, the rest of  LCP-rich points around the Procellarum basin could  be originated from a huge impact (Fig. 1, left). In order to form the Procellarum basin, which stretches 3,000 km in diameter, the impactor size should be several hundred kilometers in diameter. Such a huge impact must have excavated the "highlands" crust completely. Moreover, the excavation of the crust would have caused long-lived decompressional melting in the mantle, which leads to later lava eruption and formation of “Maria”. In other words, the dichotomy of the moon could be resulted from the massive impact that created the Procellarum basin.

Figure 3 : Distribution of low-calcium pyroxene-rich exposures over the entire surface of the moon

The uneven spatial distribution of LCP-rich exposures on the moon provided the first compositional evidence for the hypothesis that the dichotomy of the moon resulted from a massive impact. In addition, the data-mining method used in this study was found to be useful for analyzing more than 20 billion remote-sensing data.

The researchers will integrate the hyperspectral data with other data sets, such as topography and elemental composition to examine how the massive impact actually occurred and to better understand the conditions under which the moon formed. AIST is developing GEO Grid to integrate all kinds of earth observation data. In addition, the researchers plans to apply the GEO Grid technology to lunar exploration data in order to build a platform opened to the planetary scientists community.

Also, the researchers will apply the data-mining method used in this study to data from the hyperspectral earth observing satellite sensor HISUI (Hyperspectral Imager Suite) that is being developed by the Ministry of Economy, Trade, and Industry. HISUI will produce the data sets several tens of thousands times larger than the data from the Spectral Profiler of Kaguya. By improving the data-mining method, the huge hyperspectral data from HISUI will be managed to explore the mineral resource and regularly monitor the environmental burden.