2:00 PM - 2:15 PM
[PPS06-12] Magma composition during ancient lunar expansion inferred from linear gravity anomalies and their surrounding spectra
Keywords:Moon, gravity anomaly, geodesy, thermal history, spectrum, impact simulation
In this study, we surveyed spectral and gravity datasets around 160-km-sized Rowland and Roche craters. Because LGAs inside both craters are weaker than those outside, LGA materials could have been excavated by these craters. Using the Multiband Imager (MI) data from the Kaguya mission, we first identified non-mare spots holding pyroxene-like absorptions at 950, 1050, and 1250 nm. The TiO2 and FeO contents are estimated from an empirical relationship between multi-band reflectance and abundance. We then examined if these exposures contain high-calcium pyroxene (HCP) like mare basalt by analyzing the absorption centers of 1000- and 2000-nm bands with the Chandrayaan-1 Moon Mineralogy Mapper (M3) data. Whether the discovered exposures originate from the LGA intrusion is next discussed from gravity signatures. Tracing subsurface deformation and porosity change beneath the craters with iSALE hydrocode, we constrained the density and configuration of LGA sources to satisfy the GRAIL data.
Our spectral analysis shows that Rowland crater has no clear HCP exposures in its periphery. The FeO content surrounding Rowland is mostly as low as typical highland regolith. Although FeO higher than 8 wt. % is found at a few fresh craters, their spectra do not show HCP-like features. This lack of HCP implies that the LGA material was not excavated during the Rowland formation, which is consistent with our gravity simulation. The gravity drop inside Rowland is more significant than that reproducible in our simulation because the remaining root of the LGA intrusion generates a certain gravity anomaly even after the Rowland-forming impact.
On the other hand, more than 30 HCP exposures were found around Roche crater. Their FeO contents mostly exceed 8 wt. % and reach up to 14 wt. %. Their absorption centers lie within the range of HCP reported by previous works and are similar to those of small basalt maria around Roche. We exclude the pond-like locations to avoid misdetection of mare basalts. In addition, the distribution of impact melt or contribution of ejecta from craters on maria cannot be attributed to these HCP discoveries based on empirical cratering models. Thus, these exposures possibly originate from the LGA material ejected by the Roche formation. Our gravity simulation reproduces the LGA signature consistent with the data, supporting this excavation hypothesis. After correcting contamination by highland regolith using the FeO contents, the LGA material is estimated to have a TiO2 content of 0.5–1.5 wt. %. Considering the similarity of low titanium content to nearby young maria with ages of 2-3 Ga, ancient LGA magmatism shares a common magma composition with the latter basalt eruption. This implies that a mantle plume from overturned titanium-rich layer did not provide magma directly and that indirect heat from the plume or another mechanism caused partial melting.
We gratefully acknowledge the developers of iSALE-2D (www.isale-code.de/projects/iSALE), including Kai Wünnemann, Dirk Elbeshausen, Boris Ivanov, and Jay Melosh. We also thank Tom Davison for the development of pySALEPlot.
[1] Andrews-Hanna J. C. et al. (2013), Science, 339, 675–678.
[2] Zhang N. et al. (2013) JGR: Planets, 118, 1789–1804.
[3] Kato S. et al. (2017) Meteoritics & Planet. Sci., 52, 1899–1915.