Snow can be regarded as one of the most complex materials on earth. Snow is a porous high temperature material which undergoes permanents change due to diagenetic and metamorphic processes. However, snow permanently influences many facets of science and society, as it is immutably bound to climatology, hydrology, natural hazards, numeric weather prediction or public transport. This thesis addresses current methodological limitations that are of critical importance when measuring snow physical properties. These limitations appear with regard to both, in- and ex-situ measurements. In-situ measurements suffer mainly from extensive measurement duration, coarse resolution or the lack of objectiveness of traditional observation techniques. Ex-situ or remote measurements follow the progress towards higher spatial resolution with active instruments, where the modeling of active microwaves has become a serious demand. Consequently, two models were developed within the thesis: first, a statistical model to efficiently derive snow physical parameters at high resolution, with particular focus on practical applicability, and second, the extension of a current microwave model to include backscattering. The first study presents a novel approach to derive snow physical parameters with high vertical resolution and in short measurement times. A statistical model was developed to derive density, specific surface area (SSA) and correlation length from the SnowMicroPen (SMP), a high resolution penetrometer. The model was calibrated using 3-D microstructural data from micro-computed tomography (CT), which lead to an accuracy in the derived parameters of 10.6, 16.4 and 23.1 %, for density, correlation length and SSA, respectively. The potential of the method was demonstrated by the retrieval of a two-dimensional stratigraphy at Kohnen Station, Antarctica, from a 46 m long SMP transect, which clearly revealed past depositional and metamorphic events. The second study systematically assessed bias, precision and spatial resolution of different methods to measure snow density. Different approaches to measure snow density were applied in a controlled laboratory environment and in the field. Overall, the agreement between CT and gravimetric methods (density cutters) was 5 to 9 %, with a bias of -5 to 2 %, expressed as percentage of the mean CT density. Besides the good agreement, it was found that the millimeter scale density variations revealed by the CT contrasted the thick layers with sharp boundaries introduced by the observer. In this respect, the unresolved variation, i.e. the density variation within a layer, which is lost by sampling with lower resolution, was found to be critical when snow density measurements are used as boundary or initial conditions in numerical simulations. In the third study the Microwave Emission Model of Layered Snowpacks (MEMLS) was adapted to active microwave modeling. For the extension, the reflectivity had to be decomposed into diffuse and specular components to compute the specular part of the radiation which is directly scattered back to the scatterometer. Good agreement was found between scatterometer observations and simulations, if the specular snow-ground reflectivity ss0 and the cross polarization ratio q were chosen accordingly. MEMLS3&a (as well as MEMLS) was set up in a way that snow input parameters can be derived by objective measurement methods such as CT and SMP. This avoids fitting procedures of the scattering efficiency of snow and eliminates a main uncertainty in microwave remote sensing. The fourth study presents a simple data assimilation approach to locally recalibrate the SMP derived snow physical parameters. This approach was developed with regard to recent developments of the instrument, but also allows the flexibility to account for local conditions of a specific snow environment. After recalibration with a density cutter, SMP derived and cutter densities agreed within 5 %. After recalibration with the CT, the agreement between SMP and CT derived parameters was 9.5, 12.9 and 17.7 % with respect to density, SSA and correlation length. In summary, this thesis contributes to the progress in snow measurement methods and their applicability to microwave remote sensing. However, even though two models were developed within this thesis, their limited quantitative generality was also demonstrated. In addition, the simplified characterization of snow microstructure and stratigraphy appeared as serious limitation. We believe future progress in snow measurement and modeling will depend on an improved physical process representation, possibly in combination with statistical approaches or FEM modeling.