EVIDENCE THAT UVVIS RATIO IS NOT A SIMPLE LINEAR FUNCTION OF TiO 2 CONTENT FOR LUNAR

Introduction: Lunar basalts exhibit obvious variations in ultraviolet-visible (UVVIS) color and these variations have been attributed to Ti content [1], the major compositional variable in mare basalts. Charette [2] first showed that Ti content in lunar samples is correlated with telescopic measurements of UVVIS ratio in mature (agglutinate rich) basaltic regolith, whereby samples with high TiO2 concentrations exhibit relatively flatter and “bluer” UVVIS slopes (e.g., Mare Tranquillitatis) than basalts with lower concentrations in TiO2, which are considered to be redder (e.g., central Mare Serenitatis). The optical properties of spectrally neutral opaques are suggested to influence the relationship between UVVIS spectral slope and Ti [3, 4]. Ilmenite, the major carrier of Ti and the principal oxide phase in lunar soils, is dark and spectrally neutral relative to the spectrally red mature lunar mare soils. Consequently, the UVVIS color ratio has been one of the most widely used spectral parameters in lunar science to estimate the TiO2 composition of mature mare basalts. The extent to which color variations are dependent upon TiO2 or other compositional parameters, however, is not well established. For example, uncertainties in the Charette relation were revealed with failure to accurately predict the low-Ti contents of the Luna 24 basalts. Consequently we examine the correlation between UVVIS color and TiO2 content by integrating Clementine Spectral Reflectance (CSR) and Lunar Prospector neutron spectrometer (LPNS) data. Data: Clementine collected 11-band multispectral data (415, 750, 900, 950, 1000, 1100, 1250, 1500, 2000, 2620, and 2792 nm) at 100-200 meter per pixel spatial resolution for most of the lunar surface [5]. We used CSR images processed by the USGS with calibrations of [6-8]. From these data color ratio maps are made using the 415 and 750 nm images. ISIS (Integrate Software for Imagers and Spectrometers) was used to reproject the full-resolution CSR data to match the resolution of the LPNS using bilinear interpolation. The data reduction method of [9, 10] is used to map the distribution of TiO2 in the lunar soil using Lunar Prospector data. The LPNS data provide independent measurements of titanium at half-degree resolution (~15 km/pixel), which we use to test the correlation between TiO2 content and CSR data. Although LPNS data cover broad areas, they provide direct, global measurements of TiO2 that serve as ground-truth to the CSR data. So where as before there were only 6 ground-truth points in the Charette relation, there are now over a thousand data points with which to rigorously test the correlation between UVVIS and TiO2.