Attribute-Based Interpretation of Ground-Penetrating Radar Data for Improved Sea Ice Thickness Estimation

Abstract Accurately estimating the thickness of Arctic multiyear sea ice is essential for understanding cryospheric dynamics and monitoring climate change. Ground-penetrating radar (GPR) is widely used in terrestrial geophysical applications, including subsurface imaging for geological mapping, permafrost investigations, and structural assessments of the lithosphere. In recent years, GPR has also been applied to sea ice studies, particularly for detecting the snow-ice and ice-seawater boundaries. However, accurately resolving the ice-seawater boundary remains challenging due to the presence of a transition zone, where brine drainage causes gradual changes in dielectric properties. This leads to significant signal attenuation and reflection ambiguities in GPR data. To improve GPR interpretation under such conditions, this study applies attribute analysis techniques specifically instantaneous amplitude, instantaneous phase, and instantaneous Q-factor. These attributes enhance the interpretability of GPR signals and assist in identifying boundaries within brine-affected transition zones. In addition, field data acquired from GPR systems were used to evaluate how attribute analysis contributes to the interpretation of sea ice structure, demonstrating the practical value of these attributes for delineating internal boundaries and transitions within multiyear sea ice. Comparison with direct ice thickness measurements and electromagnetic (EM) surveys conducted along the same transects was used to assess the performance of the attribute-based interpretation. The attribute-based interpretation provided supplementary information that improved the interpreter’s ability to resolve complex boundary conditions. Compared to traditional methods relying on empirical calibration or multi-frequency analysis, this approach offers improved support for ice thickness estimation in challenging sea ice environments.