Possibility of existence of water-ice on lunar surface: An application of Mini-SAR data of Chandrayaan-I

Summary form only given. A rigorous effort is required to search the possibility of water-ice deposits on lunar surface. Conceptualization of MiniSAR was a breakthrough in this field for imaging permanently shadowed regions, where possibility of water-ice might be high. Earlier studies have found circular polarization ratio (CPR) greater than unity in regions representing volume scattering phenomenon due to dielectric mixing (or, water-ice deposits). Mini-SAR (Synthetic Aperture Radar) was the first polarimetric imaging system designed to map the lunar surface. It was based on hybrid-polarimetry principle and used frequency was 2.38 GHz. Polarimetry is a special feature of RADAR by which the information about targets like shape, orientation etc can be guessed. The hybrid polarimetry gives Stokes parameters and these Stokes parameters can be used to calculate various derived child parameter which can be further used to classify lunar surface based on surface roughness and other properties. A common practice in classification of surface using satellite image data analysis is assuming pixels distribution of satellite images as Gaussian distribution, which is reasonably correct when target surface has higher diversity as in case of earth surface or where geographical structures are small in size compared to image resolution. However, if image resolution is too high or surface diversity is much lesser this assumption fails, as in case of Mini-SAR images of lunar surface obtained by Chandrayaan-I. It was proposed to study this on the basis of Circular Polarization Ratio (CPR), but is observed that only CPR is not sufficient to characterize the lunar surface. So, a statistical study of Mini-SAR images is required to determine best fit density function for pixel distribution. Pixel-wise analysis of Mini-SAR imagery is generally complicated due to the presence of speckle and requires that statistical modeling methods are employed. It is well known that there are circumstances, where radar complex scattering coefficients are non-Gaussian in distribution. For this reason, various non-Gaussian models have been proposed to represent SAR data and many of these have been extended to the polarimetric SAR (PolSAR) case but not for Mini-SAR images. This study investigates best fit distribution function for obtained images from Mini-SAR for different regions. In this study circular polarization ratio (CPR), a derived child parameter from Mini-SAR data is fitted to see how fitting of density function is working. CPR is ratio of power received in left sense polarization to right sense polarization. Value of CPR>1 represents rough regions as well as regions having buried ice deposits. Moon has very large surface area (-84%) covered by craters, these craters shows abnormal behavior of CPR value. A large area gives CPR>1, So it is important to segregate physical behavior of moon surface on the basis of CPR as well as other parameters, so study of different regions of moon is carried out to analyze distribution behavior of CPR pixels.