This study presents a vehicle localization frameworkdesigned for GNSS-limited environments, particularly snow-covered Nordic regions. By integrating georeferenced snow poleswith LiDAR-based odometry, the proposed method ensures betterlocalization even in challenging conditions. The system dynami-cally switches between GNSS data and LiDAR-based localizationdepending on signal availability, leveraging snow pole detectionto refine vehicle position and compensate for sparse or unreliableGNSS signals.Experimental results indicate that the method achieves amedian localization error of 8.39 meters in GNSS-denied environ-ments, significantly outperforming FastReg’s 35.68-meter error.When GNSS signals are partially available, the framework’s ac-curacy improves to sub-meter levels, with a mean error of 1 meterand a median error of 0.52 meters at 50% GNSS availability.This demonstrates its capability to enhance localization precisiondynamically.The framework offers a cost-effective, adaptable approachrequiring minimal infrastructure modifications, advancing local-ization technology for extreme winter conditions. Future workincludes expanding applicability to urban and off-road scenariosusing additional environmental landmarks.