Measurements of the acoustic surface normal impedance of snow samples that have been fitted tolerably well by predictions of a three-parameter rigid frame porous model, in which a least squares technique with 25 data frequencies was used, are shown to be similarly fitted by an approximation-based method. In this method two approximations of the model are used that are respectively valid for low frequencies and/or high flow resistivities and low flow resistivities and/or high frequencies. Recent interpretation of the classically defined structure factor as tortuosity and effective pore radius in terms of pore shape together with measurements of porosity and air flow resistivity obtained by non-acoustic means enable an examination of the relationships between these quantities to be made. A standard relationship between tortuosity, porosity and a grain shape factor reveals a significant variation in grain shape factor with porosity in snow. This suggests the ability of acoustic techniques to distinguish between high and low porosity snows.