Topography of Candidate Phoenix Landing Sites from Moc Images

Introduction: This abstract continues our series of reports [1-3] on high-resolution topographic mapping of martian terrains based on stereo and photoclinometric analysis of Mars Global Surveyor Mars Orbiter Camera narrow-angle (MGS MOC-NA) images [4]. Our initial efforts [1,2,5] were devoted to quantifying the topographic and slope hazards of candidate landing sites for the Mars Exploration Rovers (MER). Subsequent work has been funded by the NASA Critical Data Products (CDP) Initiative, and initially focused on providing high-resolution topographic data in support of the MARSIS and SHARAD sounding radar investigations [3]. Here, we report on further CDP-funded work to assess the topographic hazards of candidate landing sites for the 2007 Phoenix Lander. Methodology: Our techniques for deriving topographic data by stereoanalysis and photoclinometry (PC) are described in detail elsewhere [5]. We use the USGS in-house digital cartographic software ISIS [6] for mission-specific data ingestion and calibration steps, as well as "2D" processing such as map-projection and image mosaicking, and photoclinometry [7]. Slope analysis is also performed with (unreleased) software that reads ISIS image files. Our commercial digital photogrammetric workstation running SOCET SET ® BAE Systems software [8] is used for "3D" processing steps such as control of the images and automatic extraction and manual editing of DEMs. SOCET SET includes a pushbroom scanner sensor model that is physically realistic but "generic" enough to describe most MOC-NA (and WA) images and allows low-order adjustments to register the images to the globally adjusted MOLA coordinates [9]. Many MOC images are also affected by high-frequency pointing variations ("jitter") that cannot be corrected with the available software for image control. Jitter in the stereobase direction gives rise to topographic artifacts in the form of stripes across the DEM; these can be suppressed by highpass filtering. Severe jitter at right angles to the stereobase interferes with matching; a workaround is to segment the image into regions that can be controlled and DEMs collected separately. MOC stereo DEMs may also contain artifacts in the form of a parabolic height variation across-track, caused by uncorrected optical distortions in the camera. Where present, such artifacts were modeled and removed with a polynomial fit. The two-dimensional PC algorithm of Kirk [7] was used to construct DEMs of selected image regions with single-pixel resolution. Accuracy of these DEMs depends crucially on the validity of photometric assumptions, in particular, on the atmospheric haze contribution to a given image, which is essentially an unknown. …