Spatial and Temporal Variation of Mars South Polar Ice Composition from Spectral Endmember Classification of CRISM Mapping Data

Abstract

Multispectral mapping data from the Compact Reconnaissance Imaging Spectrometer for Mars (CRISM) provide a unique opportunity to characterize south polar ice deposits at higher spectral sampling, spatial resolution, or spatiotemporal coverage than previous work. This new perspective can help to constrain the nature and distribution of different mixtures of CO2 ice, H2O ice, and dust that influence the formation, evolution, and preservation of Mars climate records. We processed 1103 CRISM observations spanning southern summer of six Mars Years (MY) through a combination of k-means clustering and random forest classification. Using a set of 12 spectral endmembers directly tied to previous work with high-resolution CRISM targeted data, we made a series of temporally restricted mosaics showing surface spectral variation over time. The mosaics show the effects of the MY 28 dust storm on the removal of the seasonal CO2 ice cap that year and new details of how this process differed from the years that followed. A mosaic showing residual ice surfaces displays broad agreement with previous compositional maps while resolving new details in the distribution of H2O ice-rich material around the periphery of the bright CO2 ice cap. By showing how surface composition varies across a broad swath of the south polar region though time, the endmember set and classified mosaics produced in this work can provide critical context for future studies of the dynamic processes that shape south polar ice deposits.

Characterizing Seasonal and Residual Ices at the South Pole of Mars Using a Universal Set of CRISM Spectral Endmembers

“…With a CRISM image cube as input, the module can be used to produce both a classification map showing the cluster to which each pixel has been assigned and the corresponding mean spectrum of each cluster (i.e., the spectral endmembers that can describe the scene). Testing of this method applied to a set of 14 south polar CRISM observations (Cartwright et al., 2021) demonstrated that k ‐means clustering can identify meaningful spectral endmembers closely associated with different surface morphologies and exposed strata.…”

“…However, due to the illumination geometry at the poles and photometric characteristics of icy terrains, the characteristics of the model underpinning the geometric normalization fit polar data poorly and can introduce, skew, or obscure interpretable spectral structure at longer wavelengths. We, therefore, chose to modify the TER processing workflow to exclude the geometric normalization step with the goal of preserving long‐wavelength spectral information that was found in previous work to help differentiate icy spectra that appear similar at shorter wavelengths (Cartwright et al., 2021). All other TER processing steps were implemented without modification and produced spectral cube products that report reflectance as I/F, the ratio of observed spectral radiance to incident solar radiance.…”

Spatial and Temporal Variation of Mars South Polar Ice Composition From Spectral Endmember Classification of CRISM Mapping Data

“…CRISM Targeted Reduced Data Records (TRDRs, Version 3) used in this study are publicly available via the Geosciences Node of the Planetary Data System (Murchie, 2006;Seelos et al, 2023). The derived CRISM endmember spectral library and classified maps are available via a Zenodo repository (Cartwright, 2023). Code written for this study used the open-source Python scikit-learn and R randomForest libraries (Liaw & Wiener, 2002;Pedregosa et al, 2011).…”