Rough surfaces have many applications in industry or computer graphics, including quality control, CAD, texture generation and terrain synthesis. Analysing and controlling such surfaces can be tedious, making it difficult to obtain a desired rough- ness. We introduce the Surface Differential Characteristic Function (SDCF), an analytical form that helps characterising and analysing the differential properties of particular non differentiable fractal surfaces generated by tensor product of iterated function systems (IFS). The IFS allows the generation of self-similar multiscale objects, encompassing a large variety of pos- sible roughness. The SDCF approach is an extension of the Differential Characteistic Function DCF model that was defined for analysing and characterising fractal curves. We use the SDCF tp compute the pseudo-curvatures. For smooth models, it corresponds to the curvatures obtained with classical approaches, but for fractal models, we get ranges of curvature due to the complex geometry. The self-similarity property of the surface results in these curvature to be transformed on all dyadic point in a deterministic way.