Key points are not available for this paper at this time.
Snow on sea ice strongly modulates growth, albedo, and air–sea exchange, but it also drives major uncertainties in altimeter-based thickness retrievals. This is critical for Ku-band radar altimeters (e.g., CryoSat-2, CS-2), whose waveforms integrate backscatter from snow-covered sea ice. Contrary to the common assumption that returns originate near the snow–ice interface, complex snow properties (roughness, layering, wetness, ice lenses) can shift effective scattering upward into the snowpack. We analyze Ku-band CS-2 satellite and Ka-band KAREN airborne waveforms over the Weddell Sea to partition contributions from the snow surface, snow volume, and ice surface. Using a physics-based Forward Backscatter Emulation Model (FBEM) and a CNN trained on simulated waveforms, we retrieve geophysical parameters and assess sensitivity to snow conditions. Under typical Antarctic conditions, snow-volume scattering contributes as much as, or more than, the snow–ice interface to CS-2 returns, while Ka-band is dominated by surface/near-surface snow scattering with minimal penetration to the ice surface. Wet snow further amplifies upper-layer backscatter. Sensitivity tests identify volume scattering and ice-surface roughness as primary controls on waveform shape. These results argue for explicit snow-volume terms in waveform models and support dual-frequency strategies relevant to ESA’s upcoming CRISTAL mission. Part I (this study) treats waveform decomposition; Part II evaluates retracking for improved thickness retrievals. • In summer Ku-band, snow volume exceeds interface returns, biasing freeboard. • Physically based FBEM+CNN separates surface, volume, and interface scattering. • Enables snow-aware retrackers for CRISTAL and scale-consistent altimetry processing.
Zhou et al. (Sat,) studied this question.