Accurate snow modeling is the basis for assessing the snowpack stability in support of avalanche forecasting and for many cryospheric studies, including climate modelling and paleoclimate research. However, current snowpack models have difficulties accurately reproducing the temporal evolution of relevant microstructural descriptors of snow. In particular, the evolution of the specific surface area (SSA) of snow is related to the optical, thermal, and mechanical properties of snow and is currently insufficiently represented in snow models. The present thesis improves the understanding of the microstructure evolution in snow in three studies by (i) revealing the minimum-required complexity of a snow metamorphism model to simulate the SSA evolution in snow microstructures under a temperature gradient in the first two studies and (ii) proceeding towards a physics and microstructure-based macroscopic equation for the SSA evolution in snow in the third study. This thesis makes a considerable step forward in tightly linking 4D tomography experiments and 1D snow cover modeling.