Resolution of Lava Tubes with Ground Penetrating Radar: The TubeX project

Remote sensing surveys of the Moon and Mars show evidence of lava tubes, which are potential safe havens for human crews and their equipment. Ground penetrating radar (GPR) can be used to map tubes because the void/rock interface at tube ceilings and floors strongly reflects radar pulses. We have tested the capacity of GPR to sense lava tube geometry at Lava Beds National Monument in California, USA. GPR and detailed light detection and ranging (LiDAR) data are presented for two tubes: Skull Cave, with a few meters of overburden, diameter similar to 10-20 m, and a rubbly floor; and Valentine Cave, with similarly thin overburden, diameter similar to 1-3 m, and a flatter smoother floor. On both caves GPR clearly resolves the ceiling and permits good estimates of the cave width as validated with LiDAR data. Where GPR fails, the primary cause is inferred to be strong out-of-plane effects due to complex 3-D geometries. Recovery of the floor position requires migrating the GPR data with a 2-D velocity model, as signal velocity is faster in void space. We find that floor position is recoverable in caves whose voids are taller than the radar wavelength (similar to 3 m in this study). Forward modeling assuming planetary parameters suggests that GPR should be similarly successful on the Moon or Mars. Plain Language Summary Lava tubes are tunnel-like caves found in lava flows on the Earth and other planets such as the Moon and Mars. On other planets, lava tubes can offer potential safe havens for human crews and their equipment, so developing methods for identifying and characterizing them from the surface is important. Geophysical methods are ideal tools for exploring lava tubes, among which ground penetrating radar (GPR), which does not affect rocks in the study area, is fast and relatively simple to use. In this study, we have used GPR and other tools to map lava tubes in Lava Beds National Monument, California (USA). We have collected GPR, GPS, and LiDAR data on two tubes. Their depths and widths are relatively simple to find with GPR, while the height and floor are the most challenging characteristics to be determined. Therefore, special numerical modeling algorithms, migration techniques, are used which require a general knowledge of subsurface geometry. Our tests show that with careful algorithm utilization and a good velocity model, GPR data are likely to provide an acceptable tube model.