Presentation #308.05 in the session “Active Satellites”. Cross-listed as presentation #109.06.
Europa is one of the key places in the search for life beyond Earth largely due to its liquid water ocean underneath its icy surface. While orbital data may provide clues to Europa’s habitability, successfully detecting life on Europa may require a lander to conduct experiments in situ. However, the current lack of global high resolution data of Europa makes it difficult to determine a safe lander location with a high degree of confidence. Understanding how image resolution affects the detection of potential hazards is important, both for planning landing site reconnaissance with Europa Clipper and for developing future autonomous landing systems. To that end, this project explores how hazards scale on icy areas of the Earth that share physical processes with surface terrain on Europa. We use images at different resolutions of analogs in Greenland and Antarctica that contain Europa- like features in an attempt to simulate data taken by Europa Clipper. We analyzed images of these analogs in QGIS obtained on the same day from Landsat (30 m/px), Sentinel 2 (10 m/px), and Planet Scope (3 m/px), creating maps of icebergs, crevasses, and mélange. We mapped four main units: icebergs, ambiguous rough terrain, smooth terrain, and cracks/crevasses across four analogues. Data from Helheim Glacier in Greenland, which produces icebergs and melange resembling Europa’s ridged and chaos terrain, is an example of our work. Initial results from Landsat data show 122 icebergs accounting for 16.4% of the total area mapped, while 34% was classified as ambiguous rough terrain, and 49.6% was smooth terrain. At 30m/px, our data suggest that most hazards greater than 3 to 4 square pixels were identifiable, which we set as our “confidence limit.” Using data from Sentinel 2 of the same area, 829 icebergs were mapped which made up 22.4% of the total area at 10 m/px, up from 16.4% in 30m/px images. 48.2% of the total area was mapped as ambiguous rough terrain, leaving 29.4% mapped as smooth terrain. At 10m/px, the data suggest that hazards greater than 4 to 5 square pixels in size were identifiable. This increase in the number of pixels required to identify a hazard at higher resolution arises from ice fragmentation at small scales, resulting in closely packed hazards. Our work demonstrates that there may be important terrain-dependent factors to consider in landing site reconnaissance and landing system design. We will present the results of our study across multiple sites, and provide preliminary suggestions for future mission projects.