Imaging of Martian surface and subsurface with high-frequency radar sounder

Abstract

A physics model for the electromagnetic scattering from large-scale rough surface of layered medium is established. Stratton-Chu integral equation and Kirchhoff approximation are adopted to calculate the electromagnetic scattering field. The Gaussian random rough surface is selected to simulate the actual rough surface of layered medium. The scattering echo is detected with a single-station nadir-looking radar sounder. A comparison of radar echoes from Martian mare areas with different roughness is demonstrated. A radar sounder echo simulator is developed, which could be used to validate the radar echoes processing and to support the interpretation of Martian exploration data. By using the simulator, the B-scan images of Martian highland areas with different crater density are studied. Besides, special attentions are paid to the effect of radar waveforms on the B-scan imaging of highland areas. Simulation-based results show that the method presented is feasible in the detection of Martian subsurface structure.

Computing low‐frequency radar surface echoes for planetary radar using Huygens‐Fresnel's principle

“…While planetary radar echoes are commonly used like seismograms to visually recover structural features, we propose to work with signal phases and amplitudes using equations derived from boundary integral expressions of the planetary surface scattered electromagnetic fields. Similar works have been carried out previously using Huygens's principle, in the acoustic approximation [e.g., Kobayashi et al , ], for specific setups [e.g., Fa et al , ], using numerical integration [e.g., Plettemeier et al , ], coupling numerical surface integrations and ray tracing [e.g., Jin , ; Zhang et al , ], or with approximate geometries [e.g., Nouvel et al , ]. The latter is currently implemented as a routine to quantitatively assess Mars's ionosphere and Mars's surface permittivity [e.g., Mouginot et al , ] from the increasing amount of data acquired with MARSIS.…”

“…Fundamentally, their differences arise in the geometry of the facet element used (triangular or square mostly) and in the nature of the quantities associated to each facet used to describe the fields on the surface. In the zeroth‐order series expansion case [e.g., Jin , ; Plettemeier et al , ; Zhang et al , ], quantities are sets of constant vector (i.e., independent of the position on a facet element) associated to each facet element, which is fundamentally a numerical integration. Whereas in higher‐order series expansion cases, quantities are vector fields associated to each facet (i.e., vectors' amplitude and phase can vary with the position where they are evaluated on a facet element).…”

“…On the other hand, some authors [e.g., Ferro et al , ; Russo et al , ] have used local backscattered coefficients to account for local unknown perturbations in the topography (but statistically characterized). Scattered fields have also been computed incorporating subsurface features [e.g., Jin , ; Zhang et al , ]. However, the fundamental difference does not lie in the computation of surface fields (which are supposed to be known in Huygens's principle) but rather in the choice of their local facet expressions (i.e., in the nature of the surface quantities).…”