One missing link in ground truth observations for remote sensing data of snow is continuous snowpack monitoring over the course of a season. While conventional snow pits represent only snapshots in time, which not necessarily coincident with satellite overpasses, continuous observations at short time intervals allow for direct relation of current snowpack conditions with recorded remote sensing data. Furthermore, such monitoring enables tracking of changes in adjacently recorded satellite signals to settling or disappearance/ appearance of specific snow layers or liquid water occurrences. The combination of upward-looking ground-penetrating radar (upGPR) and automatic weather station (AWS) allows for continuous monitoring of changes in snowpack stratigraphy, snow water equivalent (SWE) and bulk volumetric liquid water content (Theta_w) within the snowpack. Results thereof are not biased through spatial variability of pit locations, since upGPR is a non-destructive monitoring technique. Other non-destructive instruments recording snow parameters are usually measuring from above the snow surface. Above snow installations, however, are not capable in monitoring layer specific settling and depth of liquid water infiltrations into the snowpack. Even surface wetting cannot be clearly identified by above snow instrumentation. In addition, to monitor changes in Theta_w in snow, only non-destructive methods will produce reliable data. Here, we present upGPR data recorded over three consecutive winter season at the test site Weissfluhjoch, Davos Switzerland together with data from a slope sites above Davos, and a test site above Boise, Idaho, USA. We can show that upGPR continuously monitors major changes in snowpack stratigraphy, liquid water appearance/ disappearance and SWE within the snowpack. SWE determinations by radar were always within or close to a 5% range in comparison to manual measurements. While comparing estimated diurnal liquid water outflow with lysimeter records at the test site WFJ, the knowledge of the variations in residual water content from one day to another from the radar reduces deviations between measured and modeled outflow remarkably. The installation of upGPR systems in sheltered, spatially homogenous high Alpine areas will allow for continuous calibration and/or validation of snow retrieval algorithms for remote sensing (satellite and airborne) data. Data thereof can be used to assimilate model outputs and may help to improve and consolidate remote sensing retrieval algorithms for various kinds of snowpack conditions.