Europa's surface composition from near-infrared observations: A comparison of results from linear mixture modeling and radiative transfer modeling

Abstract

Quantitative estimates of the abundance of surface materials and of water ice particle grain sizes at five widely separated locations on the surface of Europa have been obtained by two independent methods in order to search for possible discrepancies that may be attributed to differences in the methods employed. Results of radiative transfer (RT) compositional modeling (also known as intimate mixture modeling) from two prior studies are here employed without modification. Areal (or “checkerboard”) mixture modeling, also known as linear mixture (LM) modeling, was performed to allow direct comparisons. The failure to model scattering processes (whose effects may be strongly nonlinear) in the LM approach is recognized as a potential source of errors. RT modeling accounts for nonlinear spectral responses due to scattering but is subject to other uncertainties. By comparing abundance estimates for H2SO4 · nH2O and water ice, obtained through both methods as applied to identical spectra, we may gain some insight into the importance of “volume scattering” effects for investigations of Europa's surface composition. We find that both methods return similar abundances for each location analyzed; linear correlation coefficients of ≥ 0.98 are found between the derived H2SO4 · nH2O and water ice abundances returned by both methods. We thus find no evidence of a significant influence of volume scattering on the compositional solutions obtained by LM modeling for these locations. Some differences in the results obtained for water ice grain sizes are attributed to the limited selection of candidate materials allowed in the RT investigations.