Sensitivity analysis of ice/dust aerosol and phase function assumptions on Hapke spectral unmixing and band depth parameters of martian water ice

Abstract

Spectral data from the Compact Reconnaissance Imaging Spectrometer for Mars (CRISM), specifically full-resolution targeted (FRT) observations, can provide quantitative information about the martian polar surface at local scales; however, it has been minimally used in such ways. This is partly due to the challenging feat of photometric and atmospheric data corrections needed, especially for icy surfaces at the poles. The scattering behavior of the dirty ice is not well-constrained. The presence of dust and ice aerosols makes it difficult to retrieve the surface's inherent albedos needed for quantitative studies of polar ice. The following work demonstrates how the limited phase angle range of CRISM “central scene” data and the assumptions made about the properties and conditions of the atmosphere (ice and dust aerosol opacity) and surface (surface phase function) going into the processing and correction of the data affect the derived surface properties of the north polar ice cap of Mars (i.e., H2O ice band depth parameter and Hapke-modeled grain size and abundance). Through this work, we discovered the following. 1) Both dirty ice and more pure ice were best modeled as broad, moderate forward scatterers out of the four different behaviors tested. 2) The sensitivity of Hapke-modeled grain size and abundance to ice and dust aerosol opacity and phase function uncertainty is minor (well below Hapke model error). However, all possible measures should be taken to not propagate more error since Hapke modeling already inherently has large error bars (10–30%). 3) H2O index values are affected—up to ~10% deviation—when aerosols are untreated or not treated appropriately. This should only be of concern for temporal studies. 4) Hapke-modeled grain size and abundance estimates were most sensitive to the phase angle range used for the retrieval of SSA.