||Graphical rendering of snow poses some specific challenges since it is difficult to formulate general illumination models and rendering techniques that take into account the many different appearances of snow in real nature. A common approach to modeling and rendering of materials is to develop (local) illumination models based on observations (measurements) of the interaction between incident and reflected light at certain surfaces. Other approaches use much simplified models of light propagation in space or they are based on hypotheses regarding the irradiance of certain materials. A physically correct model for the simulation of light transport in a scene is often not feasible and in many cases not even desired. This talk will focus upon one aspect of snow appearance - its diffusely reflected light in and nearby a surface point under evaluation. I will give a brief introduction into methods for modeling and rendering of snow as well as other materials that have some properties in common with snow. I then present the general idea of using surface curvature as an additional parameter for a new local illumination model. Starting from this, two new curvature based factors are developed and added to the existing illumination functions. Those factors represent ambient irradiance and occlusion due to local surface properties in the neighborhood of a surface point under evaluation. Different parameterizations of this model are discussed along with a presentation of first results of this ongoing research.