Sentinel-1 reveals large variability of dominant scattering in a drifting snow-dominated environment of East Antarctica

Assessing the Surface Mass Balance (SMB) of the Antarctic Ice Sheet is crucial for understanding its response to climate change. Synthetic Aperture Radar (SAR) observations from Sentinel-1 provide a potential to monitor the variability of SMB processes through changes in the scattering response of near-surface layers and internal snow layers. However, the interplay between accumulation, wind erosion, deposition and melt is complex, thereby complicating the interpretation of the changes in scattering of the microwave signal. Additionally, the lack of reliable ground truth measurements of snow surface limits our capability to relate the SMB processes to the dominant scattering processes. In this study, we focus on understanding how the surface processes relate to the changes in the dominant scattering mechanism from Sentinel-1 in a drifting snow-dominated region of East Antarctica. We introduce a new parameter, alpha_scat, derived from scattering-type and scattering entropy descriptors from Sentinel-1 SAR observations. This parameter quantifies the continuous scattering response from near-surface layers (i.e., pure scattering) and from internal snow layers (i.e., volume scattering). The changes in alpha_scat are evaluated from the repeated in-situ surface measurements acquired during Mass2Ant field campaigns. These measurements include roughness and accumulation derived from a terrestrial laser scanner, and surface densities from SnowMicroPen. At the field-scale, our analysis shows a strong correlation between surface roughness and alpha_scat (R-squared value of 0.99), thereby indicating the role of roughness on the dominant scattering mechanism. During periods associated with erosion, the vertical component of roughness (Root Mean Squared Height) is found to be more important than the horizontal component (Autocorrelation length) in changing the scattering response. This is also marked by an increase in alpha_scat value, indicating a tendency towards pure scattering. In contrast, accumulation events lead to surface smoothening with dominant scattering from internal snow layers. Looking at the long-term changes in alpha_scat (i.e., period 2017 - 2023), high surface densities are found to be associated with an increase in pure scattering. However, increasing (decreasing) accumulation rates contribute to suppressing (enhancing) the effect of surface density on dominant scattering. The analyses provide new insights into the connection between SMB processes and dominant scattering in Sentinel-1 observations, but more field data is needed from multiple locations to quantify the combined effect of roughness, surface density, and accumulation rates on dominant scattering mechanisms. Such a framework could lead into a better separation between pure scattering and volume scattering, thereby furthering our knowledge on observing the variability of SMB processes from Sentinel-1.