Quantification of the chemical composition of lunar soil in terms of its reflectance spectra by PCA and SVM

Abstract

In the second phase of the Chang'E Program an unmanned lunar rover will be launched onto the Moon. When ground scientists get a full understanding of the chemical composition of lunar soil around the rover, they can make more detailed survey plans for the rover and various payloads onboard so as to satisfy their scientific objectives. There is an obvious relationship between the reflectance of lunar soil and its chemical characteristics. Both principal component analysis (PCA) and support vector machine (SVM) models were applied to establishing the relationship between the reflectance spectra and chemical compositions of lunar highland and mare soil samples sent back by Apollo missions 11, 12, 14, 15, 16 and 17 and measured by Lunar Soil Characterization Consortium (LSCC). PCA was used to reduce and select the features of the reflectance spectra of lunar soil samples. Then, these features were put into SVM to estimate the abundances of various chemical components in lunar soil. We also compared the results of our measurement with those obtained by the SVM model [partial least squares (PLS)] and the principal component regression (PCR) model reported in literature. Our studies showed that with the exception of TiO 2 , the results of prediction of the abundances of chemical compounds in lunar soil by our model are much more reliable than those reported in literature. The reflectance spectra of lunar soil are closely related to the materials from which it was derived.

Global estimates of lunar iron and titanium contents from the Chang' E‐1 IIM data

“…Generally there are two types of models. One approach is based on the statistics of relationships between spectral reflectance and compositional data such as principal component analysis [e.g., Pieters et al , 2002], partial least square regression [e.g., Li , 2006], or support vector machine [e.g., Zhang et al , 2009]. Though these methods are purely mathematical, they demonstrated the potential of predicting elements with the spectral reflectance.…”

Major elements and Mg# of the Moon: Results from Chang’E-1 Interference Imaging Spectrometer (IIM) data

“…Although many high accuracy prediction models have been established with laboratory reflectance spectra (e.g., Pieters et al, 2002;Li, 2006;Zhang et al, 2009), their regression coefficients could not be directly applied to the remote sensing data because reflectance values measured from the orbital satellite and in the laboratory are significantly different. The appropriate regression method for mapping lunar composition with remote sensing data is to use reflectance values extracted directly from the images.…”

“…Pieters et al (2002) also predicted lunar soil chemistry from laboratory reflectance spectra with this approach, which can be named principal component regression (PCR). Li (2006) and Zhang et al (2009) predicted many major elements with the partial least squares regression (PLSR) and the combination of PCA and support vector machine (SVM), respectively. The approach used by Li (2006) and Zhang et al (2009), i.e., PLSR and SVM, improved the prediction accuracy over the PCR method.…”

“…Li (2006) and Zhang et al (2009) predicted many major elements with the partial least squares regression (PLSR) and the combination of PCA and support vector machine (SVM), respectively. The approach used by Li (2006) and Zhang et al (2009), i.e., PLSR and SVM, improved the prediction accuracy over the PCR method. All the researches by Pieters et al (2002), Li (2006) and Zhang et al (2009) used the spectral and chemical data measured by Lunar Soil Characterization Consortium (LSCC).…”

“…The approach used by Li (2006) and Zhang et al (2009), i.e., PLSR and SVM, improved the prediction accuracy over the PCR method. All the researches by Pieters et al (2002), Li (2006) and Zhang et al (2009) used the spectral and chemical data measured by Lunar Soil Characterization Consortium (LSCC).…”