Unpiloted aerial system-borne ground-penetrating radar for snow depth estimation in mountainous warfighter domain

We demonstrate the latest capability in unpiloted aerial systems (UAS) ground-penetrating radar (GPR) measurements for snow depth estimation in mountainous terrain. Several technical considerations are important for successful data collections such as aircraft radio frequency link interference. We found that a 2.4 GHz ground control station to aircraft link frequency maintains independence from the 0−1.5 GHz bandwidth of the Zond Aero 500 GPR. Between flights at 2 m and 4 m above ground level (AGL), the footprint of the radar signal on the ground increases by approximately 25%, which increases off-nadir scattering and reduces reflection coherence. We developed an automated layer detection tool to identify air-snow and snow-soil interfaces within the radar signal and estimated snow depth using these automated signal interpretation methods and user-guided interpretations for validation. We found that flights conducted at 2 m AGL resulted in more precise snow depth estimates (21 cm [7%] uncertainty) than flights conducted at 4 m AGL (34 cm [12%] uncertainty). We estimated snow depth with UAS-GPR; however, with further development, this technique can inform near-real-time retrievals of additional snow properties critical to vehicle mobility within the warfighter domain.