Continuous multi-offset ground penetrating radar (GPR) observation of the two-layer lake ice and snow system for ice physical properties

The measurement of physical properties and thickness of lake ice is valuable for detecting trapped gas, engineering winter roads, and understanding ice growth processes. Ground penetrating radar (GPR) is an established geophysical method used to rapidly scan frozen lake ice. For bare-ice conditions, often encountered in early winter, this is a relatively straightforward one-layer problem. However, when snow is covering the ice, it becomes important to analyze both the snow and ice layers in order to retrieve dielectric permittivities and thicknesses accurately. Given that these parameters vary spatially, it is valuable to exploit the ability of GPR to acquire data rapidly over long transects, however using a conventional single-offset configuration, physical properties can only be retrieved at points where direct snow depth and ice thickness measurement points are available to enable time-of-flight calculation of the respective velocities. Here we present a geophysical workflow to analyze continuous multi-offset GPR data for ice and snow physical properties. While a multi-offset analysis strategy and semi-continuous single-layer physical property retrieval algorithm are available in the literature, here we integrate both concepts and demonstrate the approach on a lake in Northern Alaska. Using two fixed shielded antennas simultaneously, we acquire four raypaths for each measurement location enabling velocity analysis for both layers at each position. Validation is performed using a dataset of direct snow depth and ice thickness measurements in drilled holes. We find that the retrieved layer thicknesses are within 10% of the direct measured values. These results suggest that extended multi-kilometer acquisition of ice properties to estimate volumetric gas content trapped on frozen lakes with only sparse validation is possible.