For most of the cryosphere components such as glaciers, ice sheets, sea ice, and snow satellite monitoring and change detection is well established since several decades. For permafrost, however, which represents the largest component of the Arctic cryosphere operational satellite monitoring schemes do not exist so far. Most of the processes which control the Arctic terrestrial ecosystems are related to the thermal state of permafrost and the freeze/thaw dynamics of the active layer. Hence, satellite based permafrost monitoring would be highly beneficial for the impact assessment of climate change in the Arctic. Permafrost monitoring could also be highly beneficial for the risk assessment of infrastructure in the Arctic such as roads, pipelines, and buildings which are directly affected by the thermal stability of permafrost. Increasing thaw depths and prolonged thaw periods can damage pipelines and interrupt the access to vast regions due to road damages. Sustained warming of permafrost can result in thermal erosion and landslides which threaten buildings and other infrastructural facilities. In this study we present a possible permafrost monitoring scheme based on a numerical heat flow model which is forced by multiple satellite products and initialized by weather reanalysis data. The used forcing and initialization dataset includes the land surface temperature (LST), the snow cover fraction (SCF), and the snow water equivalent (SWE). Previous studies demonstrated that MODIS LST products can deliver reasonable surface temperature measurements in tundra landscapes (Langer et al. 2010, Westermann et al. 2011). This study is based on the ten year record of the daily MOD11A1v5 and MYD11A1v5 land surface temperature products with a spatial resolution of 1km. The snow cover evolution is obtained from the daily GlobSnow SWE product with a spatial resolution of about 25km. In addition, the MODIS snow cover products MOD10A1v5 and MYD10v5 with a resolution of 1km are used in order to bridge large scale differences between the LST and SWE datasets. The model is initialized by a twenty year record of weather reanalysis (ERA-interim) and GlobSnow data. The proposed scheme is extensively tested at a typical low land permafrost site in the Lena River Delta in northern Siberia. The forcing data and model results are compared to field measurements of surface temperature, snow depth, permafrost temperature profiles, and active layer depth. The sensitivity of the model is evaluated by comprehensive Monte-Carlo simulations.