Comparison of Polarimetric Decomposition Techniques for Scattering Classification in Cabeus Crater.

The Indian Space Research Organisation’s (ISRO) Chandrayaan-2 lunar mission carries the Dual-Frequency Synthetic Aperture Radar (DFSAR), which acquires fully polarimetric (quad-pol) observations at L- and S-band wavelengths [1]. This format measures the full scattering matrix for each pixel (HH, HV, VH and VV) enabling a range of polarimetric decompositions to be used to characterise surface and near-subsurface scattering mechanisms. At the lunar south pole where low solar incidence angles create many permanently shadowed regions (PSRs), radar provides a key means of investigating these otherwise inaccessible environments.

Given the capabilities of DFSAR, Cabeus crater was chosen for this study as it was designated as a site of interest in a study completed by Lemelin et al. [2], and the selected swath also intersects with a large PSR. The analysed area is close to the LCROSS Centaur stage impact site where near-infrared absorption spectra attributed to water vapour and ice, and ultraviolet emissions attributable to hydroxyl radicals were detected in the impact plume debris [3]. The crater itself is one of the coldest craters on the Moon [3], and has a substantial fraction of water equivalent hydrogen (WEH) content of values around 0.5%wt [4].

The chosen DFSAR single look complex (SLC) swath was processed within PyPolSARPro to generate Pauli, Cloude-Pottier, Yamaguchi 4-component, and target scattering vector model (TSVM) decompositions, enabling the comparison of the scattering responses. A polarimetric whitening filter (PWF) was also used to create an image where heterogeneous areas were isolated and highlighted against the background. For example, the volumetric component of the Y4R decomposition, originally developed to represent randomly oriented vegetation scatterers, was used to understand the volumetric return of surface and near-subsurface regolith. Finally, a ratio of the Y4R odd and volume components was used to generate a surface classification map identifying areas of dominant surface, dominant volume, and mixed return in a discrete and quantifiable way.

As shown in Fig.1, the Pauli RGB image indicates surface scattering as the dominant mechanism across the region. An enhanced volumetric signal is associated with the crater walls, which is consistent with buried boulders, blocky material or potentially volatile-bearing regolith. This is reinforced by the Y4R decomposition, which identifies a reduction in surface scattering and more volume scattering across the crater walls. Regions located south of the crater are characterised by smoother, surface-dominated scattering. In contrast, the region north of the crater exhibits enhanced heterogeneity, with strong volumetric signals associated with the rims of smaller impacts. Analysis of the PWF image further highlights the northern region’s heterogeneity and enhances linear features in the crater walls.

Fig.2 presents the resulting surface classification map. The northern region is predominantly volume dominated, whereas the smoother southern region is consistent with the Y4R surface component as seen in Fig.1. By categorising scattering behaviour into discrete regimes, ratio maps enable easy identification of regions where surface and subsurface heterogeneity may indicate volatile-bearing or structurally complex regolith. This can therefore support site prioritisation for future lunar exploration and in situ investigation.  

We acknowledge the use of data from the Chandrayaan-II, second lunar mission of the Indian Space Research Organisation (ISRO), archived at the Indian Space Science Data Centre (ISSDC). This work was supported by the UK Space Agency and the Science and Technology Facilities Council [grant number: UKRI2555].

 [1] Bhiravarasu et al., 2021, Planetary Science Journal; [2] Lemelin et al., 2021, Planetary Science Journal; [3] Colaprete et al., 2010, Science; [4] Sanin et al., 2015, Icarus