Searching for New Surface Features on Mars Sean Marshall Adviser: Dr. Phil Christensen School of Earth and Space Exploration
Background Nighttime infrared (IR) images from the Thermal Emission Imaging System instrument (THEMIS) on the 2001 Mars Odyssey orbiter Nighttime images have more contrast (after processing), and they tell you more about the area’s physical properties Nighttime IR, from THEMIS image I10814003 Daytime IR, from THEMIS image I15949013
Background About 50,000 nighttime IR THEMIS images Need images that were taken at the same location (but at different times) Spatial sampling of 100 m per pixel My expectation: just a few new features I expected 3 pixels to be the threshold of visibility
Procedure Get a list of nighttime IR images: Both of decent quality With at least 50% overlap Within 5° LS of each other, to avoid seasonal differences LS is solar longitude, a way of keeping track of the time of year (and the season) on Mars 0° is the beginning of northern spring, 90° is the beginning of northern summer, and so on. There are about 180 pairs, with more coming. Look up images in the online THEMIS database Open them in GIMP Align the top of the overlapping section
Procedure Flicker back and forth between images Inspired by Clyde Tombaugh Image from The Planetary Society Record any important differences Scroll down to the next section Repeat with the next section, then with the next images
Complications Noisy images Different brightness between images Unprojected images Could project, but that is slow In general, sophisticated corrections to these problems are possible but slow THEMIS nighttime IR images I09478018 (left) and I01152002; image credit NASA/JPL/ASU
Complications Upside-down images The “footprints” (100 meters per pixel) in different THEMIS images don’t generally align Possible false negatives? From MRO HiRISE image PSP_006917_1595; image credit NASA/JPL/UofA; scaled to 2.8 m/pixel
Results I carefully analyzed 66 image pairs, which represent 0.56% of the Martian surface On average, images in a pair were taken 43 months apart No features that were clearly new Dozens of “noise effects”, where something looked like a new feature, but on closer analysis turned out to be a result of noise It can be hard to tell the difference…
Results 67.113 E, 13.197 N: Noise effect? Or new crater? In Syrtis Major About 3 pixels across (which is about 300 m) I01602003 (April 2002) I18177014 (January 2006) 5-image animation
Conclusions A feature would have to be at least 5 pixels wide/long to clearly not be a noise effect Image sampling rate averages to 0.16% of the Martian surface per (Earth) year So, suppose there was one new crater in the surveyed area that I missed, and assume that impacts occur randomly at a fairly constant average rate… That would work out to about 600 new craters (500 m diameter or more) on the entire surface of Mars per (Earth) year But that is an upper limit Actual rate lower; quite possibly much lower
Acknowledgments Phil Christensen Chris Edwards, Sylvain Piqueux Advising Chris Edwards, Sylvain Piqueux Consulting Jonathon Hill Instructing Dale Noss Script generating Everyone at Space Grant!
Image sources Arizona Space Grant Consortium logos from http://spacegrant.arizona.edu/about/azsgc_logos ASU logo from http://www.asu.edu/asuthemes/1.0/images/asu_logo2.gif Slides 1, 2, 3: ASU THEMIS, http://themis.asu.edu/about Slide 4: The Planetary Society, http://www.planetary.org/explore/topics/pluto/plutodiscovery4.html Slides 5, 7, 8: ASU THEMIS, image credit NASA/JPL/ASU, http://themis-data.asu.edu/ Slide 6: University of Arizona HiRISE, http://hirise.lpl.arizona.edu/PSP_006917_1595