Beyond a Point Source: Realistic Modelling of the RIMFAX Ground Penetrating Radar at Jezero Crater

The exploration of Mars and the Moon has been a primary focus of planetary science for decades. The prospects of resource surveying and extraction, searching for water ice, and finding potential evidence of past life have resulted in multiple missions being sent to uncover what lies within the Martian and Lunar subsurfaces. Ground Penetrating Radar (GPR) is a critical, non-destructive instrument for planetary subsurface exploration, emitting electromagnetic waves to study and reveal structures in the subsurface. The RIMFAX (Radar Imager for Mars' Subsurface Experiment) GPR antenna, aboard the NASA 2020 mission Perseverance rover, has generated approximately 40km of data since February 2021, mapping the complex sedimentary history of the Jezero crater subsurface. The Jezero crater has been an area of fervent study as it preserves a clear paleolake and river delta system. This has also made it a high-priority target for detecting biosignatures within the ancient sedimentary deposits. RIMFAX has been instrumental in this effort, mapping the dielectric properties of the crater floor to depths of tens of meters.

However, interpreting this data is challenged by an absence of readily available, high-fidelity 3D numerical models of the RIMFAX antenna and its interaction with the rover structure. Accurately modelling the geometry and properties of RIMFAX and the local Perseverance rover structure better simulates how the antenna pulse interacts with its complex environment. Approximating RIMFAX to a simple point-source can cause deviations in the waveforms, as well as fail to model the electromagnetic coupling with the rover structure; leading to flawed interpretations of the subsurface.

To address this problem, we present robust and geometrically accurate numerical models of the RIMFAX antenna and the Perseverance rover for use in gprMax, an open source finite-difference time domain (FDTD) solver. Our workflow adapts existing surface mesh models, voxelating them so that they are compatible in an FDTD environment. Material properties and excitation sources are derived from available technical specifications, or constrained through optimization processes, where proprietary data is unavailable. Validation of the models show highly consistent results with both laboratory measurements and in-situ planetary data. These freely available models enable the community to produce more realistic radargrams, leading to more accurate characterisations of the mechanical and mineralogical properties of the Martian subsurface. Furthermore, this modelling workflow provides a scalable framework for future rover-mounted GPR systems across the solar system.