Noise maps results are usually presented as contour graphs or isoline curves, which describe the sound levels as functions of spatial location. These maps are added to Geographic Information Systems (GIS), allowing sound level evaluation as a function of the continuous coordinates x and y, for a given height above ground. Although the outcome of the system is a continuous variable, the calculations that allow its evaluation are obtained from discrete points structured in a calculation grid. This grid is created by the application of spatial sampling techniques. Using spatial interpolation tools (IDW, krigging...), values are assigned to the locations in which acoustic calculations have not been performed. The application of sampling and interpolation techniques (the type of grid, its density, the interpolation algorithms...) contributes to the uncertainty of the results. This paper describes a calculation method to quantify the uncertainty associated to the spatial sampling and interpolation processes. Despite most of the previous literature refers to the uncertainty in a noise map as the uncertainty of the receivers' results (output of the noise model), in this approach we propose also including the contributions derived by the interpolation and classification processes, so that the uncertainty must be estimated only for the locations on the contour lines.
