Climate change is rapidly altering snow conditions worldwide and northern regions experiencing particularly significant impacts. As these regions warm faster than the global average, understanding snow distribution and its properties at both global and local scales is critical for effective water resource management and environmental protection. While satellite data and point measurements provide valuable information for snow research and models, they are often insufficient for capturing local-scale variability. To address this gap, we integrated UAV LiDAR with daily reference measurements, snow course measurements and machine learning (ML) approach. By using ML clustering, we generated high-resolution (1 m) snow depth and snow water equivalent (SWE) maps for two study areas in northern Finland. Data were collected in four different field campaigns during the 2023–2024 winter season. The results indicate that snow distribution in the study areas can be classified into three distinct categories based on land cover: forested areas, transition zones with bushes, and open areas namely peatlands, each showing different snow accumulation and ablation dynamics. Cluster-based modelled SWE values for the snow courses gave good overall accuracy, with RMSE values of 31–36 mm. Compared to snow course measurements, the cluster-based model approach enhances the spatial and temporal coverage of continuous SWE estimates, offering valuable insights into local snow patterns in the different sites. Our study highlights the influence of forests and forest gaps on snow accumulation and melt processes, emphasizing their role in shaping snow distribution patterns across different landscape types in arctic boreal zone. The results improve boreal snow monitoring and water resource management and offer new tools and high-resolution spatiotemporal data for local stakeholders working with hydrological forecasting and climate adaptation and supporting satellite-based observations.