Ground-penetrating radar-derived measurements of active-layer thickness on the landscape scale with sparse calibration at Toolik and Happy Valley, Alaska

Abstract

Active-layer thickness (ALT) is an important parameter for studying surface energy balance, ecosystems, and hydrologic processes in cold regions. We measured ALT along 10 routes with lengths ranging from 0.7 to 6.9 km located on the Alaska North Slope near Toolik Lake and the Happy Valley airstrip (between 68.475° and 69.150°N, and [Formula: see text] and [Formula: see text]). Using a ground-penetrating radar (GPR) system in a common-offset configuration, we measured the two-way traveltimes from the surface to the bottom of the active layer at the end of summer, when the thaw depth was greatest. We used 500 and 800 MHz antennas; the 500 MHz antenna provided suitable vertical resolution, while producing more unambiguous active-layer reflections in the presence of nonideal antenna coupling and active layer inhomogeneity. We derived ALT measurements and their uncertainties from GPR two-way traveltimes, with mechanical probing for velocity calibration. Using an empirical relationship between the wave velocity and soil volumetric water content (VWC), we found that the velocities were consistent with soil VWCs ranging from 0.46 to 0.63. In 31% of traces, the permafrost table horizon was identifiable, resulting in ALT measurements with uncertainties of generally less than 25%. The average ALT was 48.1 cm, with a standard deviation of 16.1 cm. We found distinct patterns of ALT spatial variability at different sites and different length scales. At some sites, the ALT at one point was effectively uncorrelated with ALT at other points separated by lag distances as small as tens of meters; for other sites, there was correlation at lag distances up to approximately 400 m. The ALT statistics were similar to nearby long-term in situ ALT measurements from the Circumpolar Active Layer Monitoring Network, through which yearly ALT measurements have been made since 1990.