PolSAR-Based Characterisation of the Lunar Surface using Chandrayaan-2 DFSAR Data
- Paper number
IAC-24,A3,IP,236,x83387
- Author
Ms. Shaifali Garg, Indian Institute of Remote Sensing, India
- Year
2024
- Abstract
Lunar research holds immense significance in advancing our understanding of celestial bodies. This thesis explores how Polarimetric Synthetic Aperture Radar (PolSAR) analysis can be used to characterize the lunar surface. The main goal is determining the lunar surface's dielectric constant, a crucial variable for comprehending its ice traps and other volatiles. The progression of lunar research has been constrained by the lack of fully polarimetric data, with only S-band data accessible until the advent of Chandrayaan-2. This scarcity created a research gap, a challenge this study addresses by harnessing the full polarimetric (fullpol) and additional L and S-band data from Chandrayaan-2. Two craters were chosen: Nobile with the PSR ID: SP_842890_0563440 and Sverdrup with PSR ID: SP_882490_2164550. Along with these two craters, Taurus-Littrow Valley (Apollo 17 Landing Site) was also used to explore the Dual-Frequency SAR (DFSAR) data released by Chandrayaan-2. Integral Equation Model (IEM) was used to achieve this. The statistical parameters like the mean absolute error, coefficient of correlation, etc. involved a meticulous evaluation of the IEM combined with the Artificial Neural Network and Dubois model’s predictive capabilities for the dielectric constant. This entailed rigorous testing against ground truth data and known surface features, confirming the IEM model's predictive power in capturing the dielectric constant with remarkable precision. Scattering decomposition, a pivotal aspect of this research, is introduced. The scattering mechanisms of the lunar surface are analyzed to reveal its structural and compositional characteristics. This step forms the foundation for further investigation, laying the groundwork for two prominent decomposition methods - Barnes decomposition (wave dichotomy-based decomposition) and H/A/Alpha decomposition (eigenvalue, eigenvector-based decomposition). These methods are then applied, leading to insightful results that effectively corroborate the outputs of the dielectric constant estimation. These methodologies produced findings that perfectly agreed with the dielectric outputs, proving the methodology's accuracy and reliability. The effects of the signal frequency could be seen in the dielectric constants, and the decompositions applied. All the results agreed with the conclusion that the crater bases were the sites with the most potential to host volatiles, including the water-ice traps.
- Abstract document
- Manuscript document
(absent)