Automatic Extraction of Icecap Layers from Radar Sounding Data over Greenland and the South Polar Residual Cap on Mars Siting Xiong, JanPeter Muller Imaging Group, Mullard Space Science Laboratory (MSSL), University College London, Department of Space & Climate Physics, Holmbury St Mary, Dorking, Surrey, RH5 6NT, UK, [email protected], [email protected] Methods Radar depth sounding employs low frequency radar operating at several hundreds of KiloHz to MegaHz frequencies and has been applied to the field of subsurface investigations on both the Earth and Mars. Over Antarctica and Greenland, the Multichannel Coherent Radar Depth Sounder (MCoRDS) onboard the NASA Operation IceBridge missions[1] has collected radar echograms since 2009 showing the subsurface ice layers caused by ice accumulation and interrupted by subsurface ice flow. Over the Martian polar regions, subsurface layers are also detected by low frequency radar systems, i.e. MARSIS (Mars Advanced Radar for Subsurface Ionosphere Sounding on board ESA’s Mars Express) and SHARAD (SHAllow subsurface RADar on board NASA’s Mars Reconnaissance Orbiter) [2]–[5]. Although these subsurface layers are formed by different mechanisms, there is a need for fast and automatic information extraction from these subsurface radar reflectors with the larger and larger coverage acquired nowadays. The detection and automatic extraction of subsurface layers is very important preliminary work to future studies of surface evolution and past climate. This study presents a method based on the Radon Transform (RT) to automatically extract the subsurface layers over Greenland on Earth and South Polar Residual Cap on Mars. Abstract Ice Layers of SPRC on Mars Ice Layers of Greenland Figure 4. Study area and the flight tracks of the IceBridge MCoRDS data used in this study. Velocity Map (m/year) Figure 5. Radar echogram along profile (a) AA’and (b) BB’. Obvious folding structures can be recognised from these echograms. Figure 7. Linear features extracted from two intersected radar echograms shown in 3-D format. [1] J. Li, J. Paden, C. Leuschen, F. RodriguezMorales, R. D. Hale, E. J. Arnold, R. Crowe, D. GomezGarcia, and P. Gogineni, “HighAltitude Radar Measurements of Ice Thickness Over the Antarctic and Greenland Ice Sheets as a Part of Operation IceBridge,” Geosci. Remote Sensing, IEEE Trans. , vol. 51, no. 2, pp. 742–754, Feb. 2013. [2] R. Jordan, G. Picardi, J. Plaut, K. Wheeler, D. Kirchner, A. Safaeinili, W. Johnson, R. Seu, D. Calabrese, E. Zampolini, A. Cicchetti, R. Huff, D. Gurnett, A. Ivanov, W. Kofman, R. Orosei, T. Thompson, P. Edenhofer, and O. Bombaci, “The Mars express MARSIS sounder instrument,” Planet. Space Sci. , vol. 57, no. 14{â} 15, pp. 1975–1986, 2009. [3] R. Orosei, R. L. Jordan, D. D. Morgan, M. Cartacci, A. Cicchetti, F. Duru, D. A. Gurnett, E. Heggy, D. L. Kirchner, R. Noschese, W. Kofman, A. Masdea, J. J. Plaut, R. Seu, T. R. Watters, and G. Picardi, “Mars Advanced Radar for Subsurface and Ionospheric Sounding (MARSIS) after nine years of operation: A summary,” Planet. Space Sci. , vol. 112, pp. 98–114, 2015. [4] F. Fois, R. Mecozzi, M. Iorio, D. Calabrese, O. Bombaci, C. Catallo, A. Croce, R. Croci, M. Guelfi, E. Zampolini, D. Ravasi, M. Molteni, P. Ruggeri, A. Ranieri, and M. Ottavianelli, “Comparison between MARSIS & SHARAD Results,” pp. 2134–2139, 2007. [5] L. Castaldo, A. Séjourné, and R. Orosei, “SHARAD detection model of buried CO2 ice in Mars South Polar Layered Deposit,” in 3rd Planetary Cryosphere Workshop, 2015. References Radar Sounders MCoRDS SHARAD Launch year 2009 2006 Mission Operation IceBridge NASA Mars Reconnaissance Orbiter Technique SAR SAR Frequency 193.9 MHz 20 MHz Bandwidth 10 MHz 10 MHz Antenna /2 bowtie dipole 10 m dipole Transmitted Power 500 W 10 W Pulse Length 1030 microseconds 85 microseconds Pulse repetition frequency 9000 Hz 700 or 350 Hz Platform altitude 255320 km Vertical resolution 15 m (vacuum) Azimuth resolution 0.31 km Acrosstrack resolution 37 km Penetration depth < ~ 1 km Conclusions and Future work Figure 1. An example of RT to detect linear feature. Figure 2. The calculation of correlation coefficient to connect and index peaks (The black dots on the line are detected peaks). Figure 3. Flowchart of slice processing. There are three stages: (i). Slice the radar echogram; (ii). Extraction of curved linear features based on RT, (iii). Smoothing and (iv). 3-D visualisation. (green rectangles indicate the steps, white ones denote the intermediate results). The automated extraction of icecap layers is based on the RadonTransform. A slice processing flow is proposed and utilised in this study. The study site is located south of the North Greenland Eemian Ice Drilling (NEEM) ice core, where the MCoRDS acquired two intersecting radar echograms, as shown in Fig. 4. The lower parts of the core, which contain ice from the Eemian interglacial, have been subject to folding. SHARAD is a radar depth sounder onboard Mars Reconnaissance Orbiter to study the subsurface ice water on Mars. It is planned to obtain a 3D map with this data as the coverage over the Martian poles is dense. Greenland 1.The slice processing based on Radon Transform can effectively extract continuous ice sheet layers. 2.Visualisation of ice sheet layers from intersecting radar echograms helps study the 3D structure of folded ice. The gaps along the extracted ice sheet layers indicate the occurrence of anomalies in the ice sheet. 3.In addition, effective interpolation for each ice layer is needed to fill the gaps and smooth the layers. 4.However, the effectiveness of the ice sheet layer detection is qualitatively analysed at this moment, quantitative analyses are needed in future work. Moreover, the surface layer will be coregistered to surface elevation data in the future. SPRC on Mars 1.Due to the different mechanisms of the icecap sedimentation, the ice layers show different appearance in radar sounder data over the Martian south polar icecap, which makes the slice processing based on the Radon Transform not applicable. 2.In this situation, a nonorthogonal, complex valued, logGabor wavelets can be applied to SHARAD data to suppress the noise, and then thresholded to extract the radar reflection boundaries. 3.In the future it is planned to develop a method to build a 3D block diagram of the different layers. Figure 6. Extraction of ice shet layers along profile AA’and BB’. SHARAD orbit = 032611 Original radar sounder data Denoised image after application of a nonorthogonal,complexvalued,log Gabor wavelets Extraction reflectors of high reflectance The radar sounders which are used to detect subsurface features of icecap over Greenland and Martian polar regions, are nadir looking low frequency radars. The reflections are caused by dielectric contrast in the icecap deposits. Figure 8. Imaging geometry of subsurface radar sounder. This study is jointly sponsored by China Scholarship Council (CSC) and University College London (UCL). The authors would also like to express thanks to European Space Agency (ESA) for providing the travel support under the ESAMOST China Dragon Cooperation (Project ID: 10665). Acknowledgements