L-band passive measurements (1.4 GHz) over continental areas are known to be related to surface soil moisture. Two satellite missions were recently launched to measure land surface emissions at this frequency band (SMOS-Soil Moisture and Ocean Salinity in 2009 and SMAP-Soil Moisture Active/Passive in 2015). In order to improve soil moisture retrievals from satellite data, ground-based radiometer systems operating at the same frequency were deployed over specified areas to investigate the L-band emission of various land covers under various climatological conditions. In this study, three years of L-band passive measurements from a radiometer installed on top of a steep mountain in the French Alps were analyzed and compared to L-band passive simulations. The innovative radiometer location led to large footprints due to the distance between the radiometer and the area under study. This experiment also produced microwave measurements affected by various potential difficulties typically encountered in SMOS/SMAP satellite missions: topography, heterogeneous footprints, dry/wet snow events, dew and vegetation litter. Based on in situ and modeling data, this paper investigates the potential of a radiative transfer model (L-band Microwave Emission of the Biosphere, L-MEB) to simulate L-band measurements and analyzes the differences with ELBARA observations. First, it was found that the topography generated a mixing of the horizontal and vertical polarizations. In addition, a large positive bias was found on ELBARA measurements (31 K and 12 K in horizontal and vertical polarizations respectively). Investigations showed that the sky reflection measured by the radiometer was partially substituted by land reflection coming from the surrounding topography. Second, the low-vegetation emission was investigated and highlighted the inability of the MODIS NDVI product to correctly represent the vegetation dynamics. Finally, dry snow conditions were found to have non-negligible impact at L-band and a particular signature was found during snow melting periods, with potential applications at the SMOS/SMAP spatial scales (~40 km).