5:15 PM - 7:15 PM
[PPS09-P13] Integration and Optimization of Spectral Data Analysis Tools to Understand the Water Environment of Mars
Keywords:Mars, CRISM, SHARAD
Previous Mars studies have revealed much evidence suggesting the presence of liquid water in the past (e.g., Malin and Edgett, 2000). These findings have been derived mainly from the analysis of the Compact Reconnaissance Imaging Spectrometer for Mars (CRISM) and Shallow Radar (SHARAD). The CRISM instrument obtains VIS-NIR spectral cube data, which contain spectral signatures of hydrous minerals on the surface. The SHARAD is a radar sounder that can detect dielectric discontinuities, which should correspond to water and/or ice in the subsurface of Mars. Hidaka and Ogawa (2023) developed tools to integrate and visualize the CRISM and SHARAD data. However, the tools left some challenges in encompassing complex operational procedures and the difficulty of interpreting the outputs. We aim to optimize the tools for the integrated analysis of the CRISM and SHARAD data. To achieve this goal, we plan to focus on three tasks. First, we will optimize data preprocessing procedures to enhance the efficiency of handling the CRISM data. Second, we will evaluate algorithms for detecting hydrous minerals from spectral data. Third, we will display three-dimensional visualizations to represent the integrated analysis results.
In the CRISM data processing, we implemented the entire preprocessing workflow in Python, including normalization during preprocessing. Following the preprocessing step, we applied two algorithms to identify hydrous minerals, referring to Hidaka and Ogawa (2023) and Veena et al. (2024). Hidaka and Ogawa (2023) algorithm detects absorption bands in specific wavelength ranges. Veena et al. (2024) algorithm is based on the latest supervised classification algorithm.
In the SHARAD data processing, dielectric discontinuities are detected using an algorithm from Hidaka and Ogawa (2023). The depth of detected discontinuities is converted into elevation using data from the Mars Orbiter Laser Altimeter (MOLA). We conduct clustering analysis of discontinuities based on spatial distribution using the DBSCAN algorithm. Grouped discontinuities are then interpolated to construct a spatially continuous gridded layer. We applied interpolation using the Kriging method proposed by Gupta et al. (2020).
The results from the CRISM data analysis and SHARAD data analysis are integrated and visualized in 3D space with topographic features. The locations where hydrous minerals are detected through the CRISM data analysis are intuitively compared with subsurface layers derived from the SHARAD data analysis in 3D space using geographic information.
We conducted test for such series of integrated analyses in the Ceraunius Tholus area on Mars (24.25, 262.75°E). As a result of analyzing the CRISM data (FRT00003621), using the fully Python-based workflow, the absorption bands in specific wavelength ranges indicative of hydrous minerals were detected automatically. The distribution of the detected hydrous minerals in the area has been mapped. We also performed the analysis of the SHARAD data covering the same area (7 observation lines). We found dielectric discontinuities in the subsurface of the area. The clustering of such discontinuities using DBSCAN resulted in 9 distinct clusters. The different subsurface layers in the area were visualized in 3D space after Kriging interpolation, also with the altitude model from the MOLA data. The integrated results from the CRISM and SHARAD data analyses confirmed the presence of discontinuities under some spots where hydrous minerals were detected. The results indicate a potential correlation between the distribution of surface hydrous minerals and subsurface water/ice layer. We have succeeded in developing a tool which enables series of integrated analyses.
In the CRISM data processing, we implemented the entire preprocessing workflow in Python, including normalization during preprocessing. Following the preprocessing step, we applied two algorithms to identify hydrous minerals, referring to Hidaka and Ogawa (2023) and Veena et al. (2024). Hidaka and Ogawa (2023) algorithm detects absorption bands in specific wavelength ranges. Veena et al. (2024) algorithm is based on the latest supervised classification algorithm.
In the SHARAD data processing, dielectric discontinuities are detected using an algorithm from Hidaka and Ogawa (2023). The depth of detected discontinuities is converted into elevation using data from the Mars Orbiter Laser Altimeter (MOLA). We conduct clustering analysis of discontinuities based on spatial distribution using the DBSCAN algorithm. Grouped discontinuities are then interpolated to construct a spatially continuous gridded layer. We applied interpolation using the Kriging method proposed by Gupta et al. (2020).
The results from the CRISM data analysis and SHARAD data analysis are integrated and visualized in 3D space with topographic features. The locations where hydrous minerals are detected through the CRISM data analysis are intuitively compared with subsurface layers derived from the SHARAD data analysis in 3D space using geographic information.
We conducted test for such series of integrated analyses in the Ceraunius Tholus area on Mars (24.25, 262.75°E). As a result of analyzing the CRISM data (FRT00003621), using the fully Python-based workflow, the absorption bands in specific wavelength ranges indicative of hydrous minerals were detected automatically. The distribution of the detected hydrous minerals in the area has been mapped. We also performed the analysis of the SHARAD data covering the same area (7 observation lines). We found dielectric discontinuities in the subsurface of the area. The clustering of such discontinuities using DBSCAN resulted in 9 distinct clusters. The different subsurface layers in the area were visualized in 3D space after Kriging interpolation, also with the altitude model from the MOLA data. The integrated results from the CRISM and SHARAD data analyses confirmed the presence of discontinuities under some spots where hydrous minerals were detected. The results indicate a potential correlation between the distribution of surface hydrous minerals and subsurface water/ice layer. We have succeeded in developing a tool which enables series of integrated analyses.