 PHOTOMETRIC PROPERTIES OF SOILS AT THE MARS PHOENIX LANDING SITE: PRELIMINARY ANALYSIS FROM CRISM EPF DATA.  S. C. Cull 1, R. E. Arvidson 1, F. Seelos IV 2, M.J. Wolff 3 1 Earth and Planetary Sciences Department, Washington University in St. Louis, St. Louis, MO 63112 (selby@levee.wustl.edu), 2 Advanced Physics Laboratory, Johns Hopkins University, Baltimore, MD.  3 Space Science Institute, Brookfield WI.   Introduction:  The combination of ground observations from the Mars Phoenix lander and orbital data collected by the Compact Reconnaissance Imaging Spectrometer for Mars (CRISM) provide a unique dataset for assessing scattering properties of Martian soils.  Phoenix acquired information on soil properties from its Surface Stereo Imager (SSI), Optical Microscope (OM), and Robotic Arm Camera (RAC), from solar longitude (Ls) ~80˚ to ~149˚.  CRISM acquired ~90 high-resolution hyperspectral images directly over and close to the landing site before, during, and after Phoenix mission operations (Ls~11˚~177˚).  The combined data sets provide information on how scattering properties change as water ice condenses on the surface. In this work, we attempt to constrain scattering parameters of ice-free Phoenix soils using data from CRISM's Emission Phase Function (EPF) observations.  CRISM EPFs are multiple images of the same area taken from different angles as the spacecraft approaches, flies over, and moves away from the target [Murchie et al. 2007].  By comparing radiance observed at these various viewing geometries to Hapke models [e.g., Hapke 1993] of Mars soil analogs, we can constrain Phoenix soil scattering properties, including soil grain size, single-scattering albedo, and surface phase function.   Preliminary Results:   Soil Properties from Orbit - Hapke modeling of the surface, including explict modeling of atmospheric gases and aerosols (using DISORT, Wiseman et al. 2009) was performed for multiple observations of the Phoenix site while the site was ice-free (Ls~59˚-154˚).  These ice-free observations were best-fit by a surface model with two layers of soil: a fine-grained (~15 μm) layer, ~100 μm thick, overlying a layer of sand-sized (~2 mm) soil particles.  A sensitivity analysis conducted on these models shows that the size of the sand-sized particles is poorly constrained on the upper end; however, the dust grain size is constrained to be between ~10 and 30 μm, with a clear minimum at 15 μm.  Emission Phase Function Results -- CRISM EPFs over the Phoenix landing site are poorly approximated by a Lambertian surface (a surface which produces a scattered radiance that is independent of emergence or phase angles).  They are also poorly approximated by scattering properties similar to the Gray Rock or Red Rock endmembers described by Johnson et al. [2006].  The Phoenix EPFs are closer to the Soil endmembers derived by Johnson et al. [2006] for the Spirit landing site at Gusev Crater: an asymmetry parameter of 0.498, forward-scattering fraction of 0.817, B0 of 1, and h of 0.385.   The Gusev Crater Soil endmember is a widespread plains unit that is photometrically similar to many Martian soils, including dusty surfaces at the Viking 1 landing site [Arvidson et al. 1989] and soils at the Mars Pathfinder landing site [Johnson et al. 1999].  Wolff et al. [2009] have found that Gusev soil scattering properties closely approximate most soils on Mars, when multiplied by a scale factor to adjust for albedo vairations.   Discussion CRISM spectra over the Phoenix landing site are consistently best-fit only by including a sandsized (2 mm) component to the lower layer of soil (adding it to the upper layer darkens the spectrum more than is observed), and a fine-grained (~15 μm) component to the top layer of soil.   This bimodal distribution in grain size contrasts with measurements by Phoenix's Optical Microscope (OM) experiment, which measured a mean grain size of ~90 μm by mass [Pike et al. 2009].  However, the OM experiment is biased toward smaller grain sizes: samples were first delivered to the imaging substrate, then rotated 90 degrees prior to imaging, and, as a result of tilting, larger particles may have fallen off.   Additionally, Phoenix ground observations indicate that the soil was highly cohesive and large aggregates of small particles were commonly observed [Arvidson et al. 2009].  RAC images of soil attached to the "divot" of the Icy Soil Acquisition Device (ISAD; Bonitz et al. 2008) routinely showed aggregates of soil on the order of 5 mm (in for example, RAC images RS 072 EFF 902585678_18230MB M1 and RS 099 FFL 904986760_1B7F0MR M1).   We therefore conclude that the 2-mm "grains" needed for modeling these spectra are in fact aggregates of small particles that behave like larger grains.   Conclusion  Ice-free Phoenix soils closely approximate scattering properties of Gusev Crater soils, and are best modeled as fine-grained soil (~15 μm) with a sand-sized component that may represent aggregates of smaller grains.   Future work will include further constraining ice-free Phoenix soil scattering properties, and measurements of seasonal changes in scattering properties.  References: Arvidson et al. 1989, Arvidson et al. 2009, Bonitz et al. 2008, Johnson et al. 1999, Johnson et al. [2006], Hapke 1993, Murchie et al. 2007, Pike et al. 2009,  Figure 1 (above) - CRISM EPF observations over the Phoenix landing site (black squares) vs. Lambertian surfaces of varying brightnesses (LA) under identical atmospheric conditions.  The Phoenix EPFs are a poor fit for any Lambertian surface.   Figure 2 (right) - DISORT models (gray lines) simulating FRT0000B1D2 (black line) over the Phoenix landing site for various emergence (A) and phase (B) angles.  Each gray line represents a Hapke surface with scattering parameters from one of the Gusev Crater materials described by Johnson et al. [2006].  The surfaces were overlaid with a model atmosphere based on FRT0000B1D2, and the viewing geometry varied to observe the effects on radiance.   The emergence angle gap between -13° and 13° is due to the spacecraft roll angle during acquisition.    
