ИСТИНА

Abstract: Synthetic aperture radar echoes coming from planetary surface are evaluated numerically for realistic surface topography. Simulation results are validated against true MARSIS radargrams. Applications to ionopheric studies are discussed. Introduction: Space-borne radar sounding is now effectively applied for surface and subsurface probing of planetary bodies. In subsurface sounding with ground penetrating radar (GPR), only nadir echoes from the diurnal surface and buried interfaces are useful. Side echoes coming from rough surface of the planet mask these nadir signals. The first experiment on deep radar probing of Moon, carried from the manned space mission Apollo, showed the ultimate need in the a priori information about the surface (high-resolution images, topography data etc.). This information should be incorporated in the data processing scheme in real experiments. Being started from manual sorting of echoes and visual identification of signatures produces by elements of the surface terrain, now this approach uses computer simulation of the radar wave echoes based on realistic topographic data. These data are now available for Earth, some planets and their satellites. Numerical algorithms. In the real GPR experiment, aperture synthesis and some other processing techniques are typically applied for partial suppression of the off-nadir echoes. All these procedures should be adequately represented in the numerical simulation. For this reason, the simulation algorithm may be quite complicated and resource consuming. This is probably why only a few such radar echo simulators are known, and those few algorithms still exploit great simplifications, which reduce their physical adequacy and limit the numerical accuracy. In the algorithm created by Nouvel et al. [1] the arbitrary rough surface is approximated by a mosaic of tilted square facets. The individual reflections from these facets are evaluated within the Kirchoff approximation and then coherently summed. It is well known that square tile mosaic (rather than triangular) cannot make continuous 3D surface. These artificial discontinuities unavoidably produce unphysical echoes, which are the artifact of this particular model. In addition, no aperture synthesis simulation technique applied in that model is reported in [1], although it might be implemented there. Both continuous approximation of the surface shape and aperture synthesis have been applied in [2], but for the very specific case of the so-called Martian polar valleys (surface features typical for the North Polar Cap of Mars). Here we present the echo simulation algorithm for arbitrary shape of the surface, also based on Kirchoff approximation, but using continuous surface shape approximation on the triangular grid and performing the aperture synthesis correctly. The idea is now to compare, for the same orbit, the new simulation results with results of the facet algorithm [1]. Ionospheric studies with GPR. Frequency dispersion of the ionospheric plasma corrupts phase relations among the spectral components of the signal and therefore destroys matched filtering, normally applied for the radar signal processing. At the same time, ionospheric distortions of the signal carry useful information about height distribution of the ionospheric plasma density. Dispersion curve of the ionospheric plasma, retrieved by adaptive scheme of distortion compensation, can then be used for assessment of basic ionospheric parameters (TEC, critical frequency etc.) For correct separation of the regular ionopheric phase shift from other systematic and stochastic distortions, everything in the radar experiment should be simulated as thoroughly as possible. Therefore, surface clutter modeling is a necessary part of GPR experiment, both for ionospheric and subsurface research. Within the Kirchoff approximation, ionospheric and surface clutter are independent multiplicative contributions to the signal, so that for both of them necessary numerical techniques can be individually developed, tested and optimized. The progress in surface clutter numerical models recently made gives us hope to improve ionospheric correction of real GPR signals, and to improve ionospheric parameter retrievals from these GPR data. Conclusion and remarks. Adequate computer simulation of Martian GPR experiment with the application to the real MARSIS radar sounder data is the principal goal of the work presented here. Synthetic radargrams, obtained with numerical algorithms developed by the authors, are shown. Simulation procedure includes preliminary processing of the MOLA topography data available for every part of the Martian surface, simulation of the rough surface reflection within Kirchoff approximation and unfocused aperture synthesis, exactly as it is applied in real MARSIS signal processing. Comparative analysis with real radargrams obtained from MARSIS instrument during the MEX 9466 orbit is also presented. Acknowledgements. This study has been partially supported by the European Space Agency (ESA) and Russian Fundamental Research Fund with the grant 13-02-12065 ofi-m "Fundamental problems of the microwave remote sensing of the Earth from space". One of the authors thanks the administration of the Scientific Research Computing Center of the Moscow State University for granting the access to the computational resources of the parallel computing system SKIF-GRID "Tchshebyshev". References. 1. Nouvel, J.-F., A. Herique, W. Kofman, and A. Safaeinili (2004), Radar signal simulation: Surface modeling with the Facet Method, Radio Sci., 39, RS1013, doi:10.1029/2003RS002903. 2. Ilyushin Ya.A. Martian northern polar cap: Layering and possible implications for radar sounding Planetary and Space Science 52 (2004) 1195–1207. 5MS3-MS-13