Detection of water-ice deposits using synthetic aperture radar (SAR) is a cost-effective, and efficient approach to understand lunar water resources. As water is crucial to supporting human-based space exploration, current and near upcoming lunar missions are primary concentrated on mapping and quantification of water ice exposures on surface and subsurface levels. The circular polarization ratio greater than one (CPR >1) derived using the orbital radar observations is considered as an important SAR derived parameter for water-ice detection. This study aims to investigate 14 craters near the lunar poles with high CPR (CPR >1), as identified in previous studies, using the L-band (24 cm) dual frequency synthetic aperture radar (DFSAR) onboard Chandrayaan-2. In addition to CPR, we computed the degree of polarization (DOP) after applying parallax error correction that helps in reducing misinterpretation. Our findings are based on orthorectified DFSAR calibrated data analysis. We found that the CPR of crater interiors is not significantly different from that of their surroundings, and this pattern is consistent throughout all the 14 craters selected. Further, we also found a linear inverse relationship between CPR and DOP for the interior and exteriors of the craters, with R-2 0.99, indicating a strong correlation between these two parameters. We found only 2 % of total pixels are above CPR > 1, which indicates that there is less possibility of homogeneous water-ice but the possibility of water-ice mixed with the subsurface regolith cannot be ruled out.

Understanding the reachability of water ice by future in-situ experiments near the lunar poles is crucial for supporting growing exploration plans and constraining the uncertainties on its genesis and distribution. To achieve this objective, we perform a thorough three-dimensional mapping of the distribution of water ice in the lunar polar regions (70 degrees onward), integrating radar, optical, and neutron detector observations from the Lunar Reconnaissance Orbiter mission (LRO). Our analysis reveals similar to 5-to-8-fold larger expanse of subsurface water ice (similar to 1-3 m depth) compared to surface water ice (up to 1 m depth) for the north and south poles, respectively. Our investigation cannot rule out the possibility of deep-seated water ice deposits in the lunar poles that remains beyond the detection capabilities of existing instruments on LRO. Moreover, we find that the extent of water ice in the northern polar region (similar to 1100 +/- 74 km(2)) is twice that in the southern polar region (similar to 562 +/- 54 km(2)). Our mapping also suggests that the dichotomous latitudinal distribution and the antipodal longitudinal distribution of water ice are likely driven by Mare volcanism and preferential cratering. We provide additional evidence that outgassing during Imbrian volcanism was probably the primary source of subsurface water ice in the lunar poles, which favors larger expanse over meteoritic sources.

High circular polarization ratio (CPR) characteristics were found in permanently shaded regions (PSRs) near the lunar poles. High CPR was regarded as a water ice index. The compact-polarimetric (CP) miniature radio frequency (Mini-RF) radar transmits left-circularly polarized signals and receives horizontally polarized ($S_{\text {HL}}$ ) and vertically-polarized ($S_{\text {VL}}$ ) echoes from the lunar surface. Statistics of the CPR data show its relations with the relative phase ($\delta$ ) between $S_{\text {HL}} $ and $S_{\text {VL}} $ and the degree of polarization ($m$ ) but few interpretations were provided. The average CPR data reach the maximum and minimum at $\delta =\pm 90{\circ }$ , respectively. As $m$ becomes very small, the CPR approaches 1. It has been found that CPR is also affected by surface roughness and incidence angle of radar waves. The CPR is now expressed in CP mode to explain the Mini-RF observation. Full-polarimetric radar echoes and CP parameters of the lunar surface are numerically simulated using the bidirectional analytic ray-tracing method. Single-bounce and multiple-bounce scattering components are included in the simulation. Radar images of the lunar crater are simulated with the digital elevation model (DEM) data. The $H-\alpha $ decomposition derived from the full-polarimetric simulation is presented to analyze $\delta $ and $m$ . Simulated radar images with different surface roughness are analyzed statistically to study the functional dependences of $\delta $ , ${m}$ , and CPR on incidence angle and roughness. Relationships among $\delta $ , $m$ , and CPR are used to analyze the effects of incidence angle, roughness, TiO2, and rock abundance on the scattering components. The CPR, $m$ , and $\delta $ of PSR craters of different ages are compared with those of nonpolar craters. The results indicate that the CPR, $m$ , and $\delta $ are unlikely to be unambiguous evidence of water ice.