Atmospheric Mapping Retrievals Using MRO/MARCI Ultraviolet Images

The Mars Color Imager (MARCI) instrument on-board the Mars Reconnaissance Orbiter (MRO) is a wide-angle, 7-color ``push-frame'' imager whose overall components, capabilities, and performance (including calibration methodology) are described in the recent literature (e.g., Malin et al., 2008, Icarus; Bell et al., 2009, Icarus). Of particular note for atmospheric studies is the presence of two ultraviolet (UV) bands (centered near 265 and 330 nm) which allow for the direct retrieval of column-integrated water ice cloud optical depths and ozone abundances. Combining this UV capability with the essentially global daily mapping coverage offers the capability to provide systematic spatial and seasonal characterization of water ice aerosols and ozone for the local times sampled by the MRO orbit and the MARCI cross-track geometry. The core of the two retrieval algorithms is the public-domain multiple-scattering discrete-ordinates code developed by K. Stamnes and collaborators (ie., DISORT). For computationally efficiency, the actual retrieval employs the "look-up" table approach, where each of the relevant input parameters (photometric angles, aerosol optical depths, ozone abundance) is tessellated over the sampled or expected range of values. We employ a multi-dimensional linear interpolation to provide a one-to-one transformation of radiance factor (I/F) on-sensor to the desired atmospheric parameter. The atmospheric state is specified through a combination of Thermal Emission Spectrometer (TES) and Compact Reconnaissance Imaging Spectrometer (CRISM) climatologies. The UV aerosol properties are taken from the recent analyses of the MARCI 2007 global dust event observations (Wolff et al., 2009, Icarus, submitted) and the empirical extension of the TES solarband water ice properties (e.g., Clancy et al., 2003, JGR). Initial results suggest a precision of 0.05 on cloud optical depths and 1-2 micron-atm for ozone. Our presentation will focus on the aphelion cloud belt mapping results from the first four months of the mapping mission (Ls=110-160, Sept 2006-Jan 2007). We also provide a brief overview of our retrieval along with the results of our parameter sensitivity studies.