51st Lunar and Planetary Science Conference (2020) 2790.pdf

SUBSURFACE WATER ICE MAPPING (SWIM) ON : SURFACE REFLECTIVITY. G. A. 1 1 2 1 2 2 4 Morgan ,​ N. E. Putzig, B​ . A. Campbell ,​ Z. M. Bain ,​ A. M. Bramson ,​ E. I. Petersen ,​ M. Mastrogiuseppe ,​ M. R. 2 ​ 5​ 1 ​ 6 ​ 1 ​ 1 ​ 1 ​ Perry ,​ D. M. H. Baker ,​ I. B. Smith, R​ . H. Hoover ,​ H. G. Sizemore .​ A. Pathare and the SWIM Team, P​ lanetary ​ ​ ​ 2 ​ 3 ​ ​ Science Institute. ([email protected]) S​ mithsonian Institution, L​ unar and Planetary Laboratory, University of 4 ​ ​ ​ 5 ​ 6 Arizona, S​ apienza University of , N​ ASA Goddard Space Flight Center, S​ outhwest Research Institute. ​ ​ ​

Introduction: The Subsurface Water Ice Mapping and MOLA), and SHARAD surface and subsurface (SWIM) project supports an effort by NASA’s Mars radar echoes. Exploration Program to determine in situ resource Consistency Mapping: For the SWIM 2019 maps, ​ availability [1–2]. We are performing global we used the SWIM equation [2–3] to provide a reconnaissance mapping as well as focused quantitative assessment of how consistent (or multi-dataset mapping to characterize the distribution inconsistent) the various remote sensing datasets are of water ice from 60ºS to 60ºN latitude. In 2019, we with the presence of shallow (<5 m) ice. The SWIM produced ice consistency maps for the northern Equation yields consistency values ranging between hemisphere (0 to 60ºN) from 0–225ºE and 290–360ºE +1 and -1, where +1 means that the data are . In 2020, we are extending our mapping consistent with the presence of ice, 0 means that the into the southern hemisphere (0 to 60ºS) and from data give no indications of the presence or absence of 225–290ºE longitude in the northern hemisphere at ice, and -1 means that the data are inconsistent with elevations <+1km (Fig. 1). Our maps are being made the presence of ice. Here, we focus on our mapping available on the SWIM Project website of ice consistency values from geomorphic mapping. (https://swim.psi.edu), and we intend to complete our For more information on the project and its ​ ​ global mapping by the summer of 2020. Follow us on techniques and datasets, visit our website and Twitter @RedPlanetSWIM for project news and associated presentations at this LPSC: Putzig et al. product release information. (results), Perry et al. (methods), Sizemore et al. The SWIM Datasets: To search for and assess the (thermal/neutron analysis), Baker et al. ​ presence of shallow ice across our study regions, we (geomorphology), Petersen et al. (radar subsurface are integrating multiple datasets to provide a holistic mapping) and Bain et al. (site analyses). view of the upper 10s of m of the subsurface. 2019 Methods: SHARAD was designed to study The individual datasets and methods we employ the subsurface structure of Mars through the include neutron-detected hydrogen maps (MONS), detection of reflections originating from boundaries thermal behavior (TES, THEMIS, and MCS), between underground layers with contrasting multiscale geomorphology (HiRISE, CTX, HRSC dielectric properties. Indeed, Petersen et al [this

Fig. 1. SWIM project study regions and 2019 results (northern hemisphere) for radar surface reflectivity ice consistency mapping. Background MOLA hillshade

51st Lunar and Planetary Science Conference (2020) 2790.pdf

LPSC] provides a full description of how the SWIM returns are associated with the glacial features within project uses SHARAD data for this very purpose in . The strong correlation suggests order to search for buried ice-rich layers. However, high ice concentrations in the upper 5m, either in the echoes returned from the surface of Mars also contain glaciers themselves or the overlying mantle units. a wealth of useful information, including surface As low density material will result in a similar roughness and near-surface Fresnel reflectivity. power return as ice-rich deposits, the reflectivity map Fresnel reflectivity provides a measure of the is also expected to contain “false positive” ice density of the near surface. As ice is a low density detections. For example, the low-power band material, especially in comparison to the regolith and centered at 33ºN within Arcadia (180–210ºE) rock that make up most of the , corresponds to a known region of active dust measuring reflectivity offers a strategy to search for upwelling that has been monitored since the Mars ice rich deposits. Within this context, the ‘surface’ is Global Surveyor Mission [6]. It is therefore likely defined by the SHARAD central wavelength (15 m) that a locally thick dust cover is responsible for the and actually refers to the upper ~5 m of the lower-power SHARAD returns. subsurface. Consequently, the bulk density over this Determining the cause of the lower power range can be constrained. As every SHARAD demonstrates the strength of the SWIM Project measurement includes a value of the surface power approach. By incorporating multiple datasets that also returned, we can generate density estimates across the probe the shallow near-surface of Mars, ice planet and use them to search for regions of low detections of high confidence can be distinguished power that can be indicative of shallow ice. from likely false positives through the SWIM Density variations derived from the range in equation. substrate geology expected to exist on Mars (water 2020 Methods: The next phase of the SHARAD ​ ​ ice to dense basalt) should account for ~6 dB in reflectivity analysis is to further reduce the width of return power variations. However, SHARAD was not the power distribution by explicitly correcting for the calibrated to specifically measure surface power, and influence of MRO roll and solar array positioning [7]. multiple factors external to surface reflectivity Both parameters have a significant effect on the influence the return power measured by the antenna. amount of power transmitted to the martian surface. Hence, individual SHARAD measurements cannot be The new corrections will be applied to the full directly converted into density estimates. SHARAD coverage of SWIM study area (Fig. 1). Nevertheless, it is possible to narrow the power Within the highest priority regions we will also distribution measured by SHARAD and thereby leverage the overlapping, but different probing depths better constrain the influence of density. To achieve of the thermal datasets (<1m) relative to the this, we broadly follow a similar methodology first SHARAD surface return product (5m) to permit two attempted with MARSIS data [4], while taking into layered modelling of the composition of the upper account the higher frequency of SHARAD. 5m. As ice is a low density, yet high thermal inertia First, to limit the ionosphere effects we exclude all material, directly comparing radar reflectivity with daytime tracks. Next, we normalize the power for thermal data provides an avenue to exclude false surface roughness using the SHARAD roughness positive signals associated with each dataset - i.e. for parameter [5]. To further account for topographic thermal, rock/ice are indistinguishable, whereas for effects at longer baselines, we correct for the loss of radar, porous sediments/ice appear near identical to power due to regional slope using the median MOLA one another. slope value over a Fresnel zone (D: 3 km). Finally to Acknowledgments: The SWIM project is ​ ​ account for additional MRO influences ( supported by NASA through JPL Subcontracts role, solar panel configuration), we take the median 1611855 and 1639821. value of corrected SHARAD returns sampled over References: [1] Morgan et al., submitted, Nature ​ ​ ​ Astronomy. [2] Putzig et al. (2019) Ninth Int. Conf. Mars, 1/12° bins. ​ ​ ​ SWIM 2019 Results: Regions of low power no. 6427. [3] Perry et al. (2019) LPSC 50, no. 3083. [4] ​ ​ ​ returns, consistent with the presence of ice are Mouginot, J. et al., (2010) Icarus, 210, 612–625. [5] present across the high-mid latitudes (Fig. 1). From a Campbell B.A. et al. (2013) JGR, 118, 436-450. [6] Fisher, J.A., (2005) JGR, 110. [7] Croci, R et al., 2007, IEEE Int. resource perspective, the most promising low power ​ Geosci. Remote Sensing Sympos. (pp. 5218-5223).