Observations of geophysical and dielectric properties and ground penetrating radar signatures for discrimination of snow, sea ice and freshwater ice thickness

Abstract

Abstract Separate snow and ice thickness at the same location are important parameters in determining the flux of heat and light between the atmosphere and ocean in Arctic marine environments, but physical sampling requires a great deal of effort to yield few results spatially and temporally. Here, ground penetrating radar (GPR) at 250 MHz and 1 GHz reliably measured snow, river ice and sea ice thickness and compared well with physical and dielectric properties of the media in spite of their spatial and temporal variation. GPR velocities were calculated for snow, river ice and sea ice from both dielectric information and GPR returns for snow river ice and sea ice. In snow on river ice, physical sampling yielded permittivity (e′) of 1.68 at 1 GHz while the apparent e′ (derived from GPR measurements) at 1 GHz from GPR returns ranged from 1.6 to 2.1 resulting in snow thickness overestimations of 6.7% to 11%. River ice e′ at both frequencies from physical samples was 3.17, while apparent e′ ranged from 2.94 to 3.11, underestimating ice thickness by less than 3.6%. Snow on sea ice on 4 March 2006 had a e′ of 1.94 at 1 GHz calculated from physical samples, and 1.9 derived from GPR measurements, resulting in near-identical thicknesses. Snow on sea ice 3 days later had an e′ of 1.67 at 1 GHz calculated from physical samples, while its apparent e′ was 1.9 at 1 GHz, leading to a GPR-predicted snow thickness which was 1 cm (8.3%) greater than physically measured. Physical samples of sea ice on 4 March yielded e′ of 3.8 at 1 GHz, while GPR returns yielded an apparent e′ of 3.35, underestimating sea ice thickness by 5.6%. GPR returns at 250 MHz returned a snow-plus-sea ice thickness which was 5.9% less than physically measured. In cases when the 1 GHz radar did resolve the sea ice–seawater interface, the 250 MHz system was used to measure the snow-plus-sea ice thickness and the snow thickness resulting from the 1 GHz radar was subtracted to make separate snow and sea ice thickness measurements at the same location.