A micro-scale modelling approach of snow based on the extension of the classical discrete element method has been presented in the first part of this study. This modelling approach is employed to predict the mechanical response of snow under compression dependent on strain rate, initial density and temperature. Results obtained under a variety of conditions are validated with experimental data for both micro- and macro-scale, in particular the broad range between ductile i.e. low deformation rates and brittle i.e. high deformation rates regimes are investigated. For this purpose snow is assumed to be composed of ice grains that are inter-connected by a network of bonds between neighbouring grains. This arrangement represents the micro-scale of which the interaction is described by inter-granular collision and bonding. Hence, the response on a macro-scale is largely determined by inter-granular collisions and deformation and failure of bonds during a loading cycle. Consequently, validation was first carried out on micro-scale deformations at different loading rates and temperatures. Hereafter, macro-scale simulations of confined and unconfined, deformation-controlled compression tests have been predicted and were successfully compared to experimental data reported in literature.