Surface-wave analysis for identifying unfrozen zones in subglacial sediments

To reveal the extent of freezing in subglacial sediments, we estimated S-wave velocity along a glacier using surface-wave analysis. Because the S-wave velocity varies significantly with the degree of freezing of the pore fluid in the sediments, this information is useful for identifying unfrozen zones within subglacial sediments, which again is important for glacier dynamics. We used active-source multichannel seismic data originally acquired for reflection analysis along a glacier at Spitsbergen in the Norwegian Arctic and proposed an effective approach of multichannel analysis of surface waves (MASW) in a glacier environment. Common-midpoint crosscorrelation gathers were used for the MASW to improve lateral resolution because the glacier bed has a rough topology. We used multimode analysis with a genetic algorithm inversion to estimate the S-wave velocity due to the potential existence of a low-velocity layer beneath the glacier ice and the observation of higher modes in the dispersion curves. In the inversion, we included information of ice thickness derived from high-resolution ground-penetrating radar data because a simulation study demonstrated that the ice thickness was necessary to estimate accurate S-wave velocity distribution of deep subglacial sediment. The estimated S-wave velocity distribution along the seismic line indicated that low velocities occurred below the glacier, especially beneath thick ice ([Formula: see text] for ice thicknesses larger than 50 m). Because this velocity was much lower than the velocity in pure ice ([Formula: see text]), the pore fluid was partially melted at the ice–sediment interface. At the shallower subglacial sediments (ice thickness less than 50 m), the S-wave velocity was similar to that of the pure ice, suggesting that shallow subglacial sediments are more frozen than sediments beneath thick ice.