Accurate estimation of Snow Water Equivalent(SWE) and Snow Depth(HS) is essential for water resource management and climate monitoring in mountainous regions. The retrievement of these variables at regional scale is challenging due to the complex interaction between snow properties and radar signals. This thesis work investigates the potential of Interferometric Synthetic Aperture Radar (InSAR) data combined with temperature information and deep learning techniques for the estimation of regional SWE and HS variations. The study is structured into two main parts. The first one focuses on dataset construction through a preprocessing pipeline, which integrates multiple data sources: InSAR data (coherence, displacement, elevation, intensity, local incidence angle, phase, unwrapped phase bands) from Sentinel-1 mission, temperature data from Sentinel-3 mission, temporal embeddings for seasonal dynamics and a validity mask to exclude invalid regions.