We present methods and results from interferometric data processing of a long-lasting survey campaign monitoring the Planpincieux glacier, located on the Italian side of the Mont Blanc, using a ground-based synthetic aperture radar (GB-SAR). Monitoring a European Alpine glacier during the winter, when the meteorological conditions are highly variable, presents some difficulties in radar data interpretation. The main issues to tackle in interferometric processing are unwrapping errors and high amplitude dispersion (DA), mainly due to the high velocity and dielectric heterogeneity of the backscattering surface. To improve the reliability of the results, a coherence-driven pixel-selection criterion for identifying the glacier area and a simple approach to reduce possible unwrapping errors in interferograms with low coherence are here proposed. The development of a new 2D polynomial regression model, as a function of elevation, for atmospheric phase screen (APS) estimation is also discussed. A comparison with the results obtained with a vision-based approach gave showed good agreement.