GEOMETRIC RESTORATION OF MOON MINERALOGY MAPPER DATA AND IMPLICATIONS FOR ANALYSIS OF THE APOLLO 17 LANDING SITE

Data from NASA’s Moon Mineralogy Mapper (M³) are important for mapping and characterization of lunar surface units. We report on the status of geometric restoration of the M³ data, using data from Global and Targeted imaging modes (140 and 70 m/pixel spatial resolution, respectively) in 85 channels at wavelengths between 460 and 3000 nm. We used the Lunar Reconnaissance Orbiter Wide Angle Camera (WAC) topographic model and global mosaic as stable bases to improve positional accuracy of M³ frames. The M³ Level 1B pipeline was used to process the data through ray tracing and geometric modeling, creating a new orthorectified product. For further geometric processing, we used an M³ camera model in ISIS3 (http://isis.astrogeology.usgs.gov/), added tie points between M³ frames and ground (constrained) points with automated and manual procedures, and bundle-adjusted the frames. We then map-projected the images and evaluated the consistency of overlapping images in map coordinates. The resulting M³ control network is based on 870 images, 176,769 points, and 627,052 measurements. Analysis of 175 images from late in the mission (i.e., Optical Period 2C1, during which both spacecraft and instrument position and pointing were poorly understood) showed that the reprocessed M³ data are registered to the WAC mosaic at 1 pixel (~140 m), as compared to 5 to 12 pixels in the archived data.

The Level 2 pipeline was used to compute normalized reflectances, the Lommel-Seeliger single-scattering model was used to develop photometric correction coefficients, and these were then applied to thermally corrected M³ data. While overall results showed little to no offsets and may have been acceptable, data for the equatorial regions showed a mean offset of 60 km and a maximum of ~180 km in phase angles when compared to the opposition surge point. To correct these problems, we developed improved pipeline and bundle-adjustment procedures with more rigorous error analysis that leveraged the latest control points and SPICE kernels. New results show significant improvement in orbital reconstruction, modeling of spacecraft and instrument pointing and position, photometric angles, and final mosaics. We will discuss implications for these improvements on analysis of the M³ data for the Apollo 17 landing site region.