 TERRACED CRATERS AND SUBSURFACE ICE IN ARCADIA PLANITIA, MARS.  A. M. Bramson 1, S. Byrne 1, S. Mattson 1 and J. J. Plaut 2, 1 Lunar and Planetary Laboratory, University of Arizona, Tucson, AZ (bramson@lpl.arizona.edu). 2 Jet Propulsion Laboratory, Pasadena, CA.   Introduction:  Understanding the availability of ground ice on Mars is important to future exploration of the planet. Near-surface ice was predicted theoretically for decades before its detection by the GammaRay Spectrometer [1-3] aboard Mars Odyssey. The extent and depth of this ice can also tell us much about recent Martian climates (e.g. [4]).  Obliquity reached peaks of ~35° several times within the past million years and values as high as ~45° 5-6 million years ago [5]. These high-obliquities would have allowed for ice to migrate from the polar regions to mid-latitudes [4]. While not stable at the surface at mid-latitudes today, ice can be stable at these latitudes when covered by a thin layer of debris to protect the ice from sublimation. Dickson et al. [6] mapped over 1500 ice-related features in the mid-latitudes of Mars, demonstrating the widespread extent of buried ice at these latitudes. Excavation by small impact craters and the Phoenix lander indicates that this ice is almost pure in widespread locations [7,8]. Climate simulations with obliquities of 45° [9] indicate thick surface ice accumulations in certain locations such as eastern Hellas.  Examination of lobate debris aprons with the Shallow Radar (SHARAD) instrument onboard the Mars Reconnaissance Orbiter (MRO) in eastern Hellas [10] and Deuteronilus Mensae [11] indicates these lobate features have dielectric constants consistent with almost pure ice.  Figure 1: A terraced crater (with another, unrelated crater in the middle giving the bullseye appearance) located at 194.84°E, 46.58°N. Red dots show where elevation information was taken on the wall-terrace; blue dots are from elevation points for the inner floor. HiRISE image ESP_018522_2270  SHARAD has also detected a layer up to 50-90 m thick that covers much (2000 km by 500 km) of the Arcadia Planitia region [12]. This subsurface reflection indicates a change in the dielectric constant of the material, but the depth of the change and absolute values of the material dielectric constants involved cannot be determined from the radar data alone. Plaut et al. [12] considered a range of dielectric constants from that of pure ice (~3) to basalt (~8). Because of the numerous ice-related features in the region, it is possible this layer is composed of ice. However, many compositions, including volcanic or sedimentary materials, would also be consistent with the SHARAD data. In this work, we utilize independent information from terraced craters to constrain the depth of this interface.  Knowledge of its depth allows the radar data to be interpreted in terms of dielectric properties and constrain composition.  Bullseye Crater: The "bullseye" crater located at 194.84°E, 46.58°N (Figure 1) is an example of a terraced crater (that happens to have another, unrelated crater in the middle giving it the bullseye appearance). Terraces within craters form due to changes in the strength of the target material at certain depths. This change in material properties could be related to a transition from icy material to dry material if a layer of subsurface ice exists in the area. The crater also has a flat floor indicating another possible change in material properties (craters in this size range should form simple bowl shapes) (Figure 3).  Figure 2: SHARAD echo power binned over latitude (left) and the corresponding radargram (right). A second reflection at 0.45 μs below the surface is visible.  100 m The depth that the strength transition takes place in this bullseye crater was measured by extracting elevation points using Socet Set on HiRISE (High Resolution Imaging Science Experiment) stereo image pairs. The elevation data for the outer terrain, inner floor (blue dots in Fig. 1) and wall-terrace (red dots in Fig. 1) of the crater were each fitted to planes. The standard deviations of the residuals of the fits were much lower than the variation in original elevation data, meaning flat planes describe the elevation data well for each of these three locations (surroundings, terrace, and floor). The elevation extracted at the center of the planar fit for the crater floor is 38 meters below the surrounding terrain and the terrace lies 13 meters below the surrounding terrain. A nearby SHARAD track (within ~1km of this crater) was used to look for a reflection that could be from the same material change that led to the terracing of the crater. A subsurface reflection from ~0.45 μs below the surface was found (Figure 2).  This is the same regional reflector documented by [12]. A reflection with this delay time from the terrace level at depth 13 m (Figure 3) would correspond to a permittivity of 28.55 for the upper material. However, if the reflection is from the floor level of the crater, a 0.45 μs delay from an interface at a depth of 38 m (Figure 3) would imply a permittivity of 3.17 for the upper material. This means it is possible the material between the crater floor and surface is icy (ε=3.15 for pure ice).  Figure 3: Diagram showing the expected shape of a 700 m diameter crater and the actual shape of the bullseye crater. Two possibilities are that an icy-to-dry transition occurs at the terrace level (middle diagram) or at the crater floor (bottom diagram).  Mapping Terraced Craters: We search for terraced craters using high-resolution data from MRO's Context Camera (CTX), and obtain elevations and depths of terrace transitions using the same method described for the bullseye crater where stereo images are available. Within ~23 km of bullseye crater, we find 10 other terraced craters (Fig. 4). We will present results from mapping these terraced craters within the region of interest specified by Plaut et al. [12]: 38°50°N latitude and 180°-225°E longitude.    Figure 4: We find 10 additional terraced craters (green) within ~23 km of bullseye crater (red).  Future Work: We continue to look for more terraced craters within the Arcadia Planitia region where SHARAD detects the subsurface boundary. We will also search other areas to identify how common these craters are and to probe subsurface layers elsewhere in the mid-latitudes. We will compare the results from the depth calculations of terraced craters to reflections seen in nearby SHARAD data. Assuming the change in material properties within the terraced craters corresponds to the boundary seen in the SHARAD data, we can use the delay time to constrain the thickness, and thus the dielectric constant, of this layer and how it varies across Arcadia Planitia. Mapping the depth of this reflection across the entire region will allow us to constrain where and how deep the ice is, leading to a better understanding of the past climates on Mars.   References: [1] W.V. Boynton et al. (2002) Science 297, 81. [2] W.C. Feldman et al.   (2002) Science 297 75. [3] I. Mitrofanov et al. (2002)   Science 297, 78. [4] M.A. Chamberlain & W.V. Boynton, (2007), JGR 112, E06009. [5] J. Laskar et al. (2004) Icarus 170, 2, 343. [6] J.L. Dickson et al. (2012) Icarus 219, 723. [7] S. Byrne et al. (2009) Science 325, 1674. [8] P.H. Smith et al. (2009) Science 325, 58. [9] F. Forget et al. (2006) Science 311, 368. [10] J.W. Holt et al. (2008) Science 322, 1235. [11] J.J. Plaut et al. (2009) GRL 36, L02203. [12] J.J. Plaut et al. (2009) LPS XL, #2312. 
