Multi-Offset Imaging of Bed Topography Using Radio Frequency over Fiber Radar Arrays: Modelling and Initial Field Results

Radio-echo sounding (RES) is a widely used tool in glaciology, providing insight into englacial and subglacial environments. Conventional high-spatial resolution RES surveys typically employ zero- or small-offset configurations with a single transmitter-receiver pair. Such surveys often prioritize spatial coverage over monitoring temporal changes in englacial and subglacial conditions. Stationary radar arrays aimed at providing time series data have been previously deployed in glaciated regions to provide estimates of basal melt rates, infer vertical strain within ice sheets, and image englacial layers in 3D. However, these stationary arrays are unable to image the ice-bed interface with sufficiently high resolution to infer changes in bed geometry over time. This is largely due to hardware limitations in the radar systems used in glaciology which typically support an inadequate number of antenna elements. Unlike in towed or airborne radar systems, where spatial resolution can be improved through synthetic aperture processing techniques, the spatial resolution achieved by a stationary array is proportional to the number of real antenna elements deployed. We overcome limitations in the number of supported antennas by integrating radio-frequency over fiber (RFoF) hardware, typically used in the communications industry, into existing radar systems such as the autonomous phase-sensitive radio-echo sounder (ApRES), as well as software-defined radios (SDRs). By converting RF signals to optical signals, lossy copper-based coaxial cables is replaced by low-loss fiber optic cables, permitting large separations between receive and transmit elements without significant signal attenuation during transmission. Further, the low cost, high switching speeds, and large number of output channels provided by fiber optic switches allows for a cost-effective way to rapidly cycle through 100s of antenna elements using a single radar unit RF input or output port. These modifications allow an ApRES, which traditionally supports up to 8 receive and 8 transmit antennas, to handle 100s of antennas on both the receive and transmit side, offering significant improvements in imaging capabilities. Such a system could support advanced imaging geometries capable of 3D time-lapse monitoring of englacial and subglacial processes, such as seasonal hydrology, subglacial erosion, isostatic rebound, and the evolution of sub-ice shelf features. We demonstrate these imaging capabilities through modelling and initial field results using our modified ApRES and SDR systems.