Understanding snow accumulation and distribution in high mountain regions, as well as snow water equivalent (SWE), is crucial for accurate hydrological forecasting in snow-influenced catchments. A major challenge in these systems is the scarcity and complexity of snow-related observations; furthermore, hydrological models are required to represent snow processes at larger scales. However, both observations and models are subject to different sources of uncertainty, affecting prediction accuracy. Data assimilation techniques help improve predictions in such uncertain scenarios. Among these, particle filters (PF) are particularly suitable for nonlinear systems. This study aims to reduce uncertainty in key snowpack variable predictions in the Andes Mountains by implementing PF and assimilating hydrological variables. Additionally, it explores the variability in spatial precipitation patterns by updating and correcting state variables through discretization into elevation bands. The study domain is El Yeso Basin in Chile, and the hydrological model used is the Canadian Hydrological Model (CHM), a physically based distributed model that represents terrain using irregular triangular elements. Specifically, the work focuses on state variables such as SWE, input variables derived from numerical weather prediction, and the use of observations from remote sensing products such as SCA/fSCA. This approach constrains uncertainty, enabling updated predictions with reduced variability.