Semi-automatic measures of activity in selected south polar regions of Mars using morphological image analysis


The High Resolution Imaging Science Experiment (HiRISE) onboard Mars Reconnaissance Orbiter (MRO) has been used to monitor the seasonal evolution of several regions at high southern latitudes. Of particular interest have been jet-like activities that may result from the process described by Kieffer (2007), involving translucent CO2 ice. These jets are assumed to create fan-shaped ground features, as studied e.g. in Hansen (2010) and Portyankina (2010). In Thomas (2009), a small region of interest (ROI) inside the south polar Inca City region (81° S, 296° E) was defined for which the seasonal change of the number of fans was determined. This ROI was chosen for its strong visual variability in ground features. The mostly manual counting work showed, that the number of apparent fans increases monotonously for a considerable amount of time from the beginning of the spring time observations at Ls of 178° until approx. 230° , following the increase of available solar energy for the aforementioned processes of the Kieffer model. This fact indicates that the number of visual fan features can be used as an activity measure for the seasonal evolution of this area, in addition to commonly used evolution studies of surface reflectance. Motivated by these results, we would like to determine the fan count evolution for more south polar areas like Ithaca, Manhattan, Giza and others. To increase the reproducibility of the results by avoiding potential variability in fan shape recognition by human eye and to increase the production efficiency, efforts are being undertaken to automise the fan counting procedure. The techniques used, cleanly separated in different stages of the procedure, the difficulties for each stage and an overview of the tools used at each step will be presented. After showing a proof of concept in Aye (2010), for a ROI that is comparable to the one previously used for manual counting in Thomas (2009), we now will show results of these semi-automatically determined seasonal fan count evolutions for Inca City, Ithaca and Manhattan ROIs, compare these evolutionary patterns with each other and with surface reflectance evolutions of both HiRISE and CRISM for the same locations. References: Aye, K.-M. et. al. (2010), LPSC 2010, 2707 Hansen, C. et. al (2010) Icarus, 205, Issue 1, p. 283-295 Kieffer, H.H. (2007), JGR 112 Portyankina, G. et. al. (2010), Icarus, 205, Issue 1, p. 311-320 Thomas, N. et. Al. (2009), Vol. 4, EPSC2009-478

Michael Aye
Michael Aye
Research Scientist in Planetary Science

My research interests include remote sensing of surfaces, related machine learning studies and open source software.