Identifying the risk of traffic noise is vital in minimizing traffic noise pollution in urban areas. As noise travels in every direction, 3D visualization of traffic noise is essential, which involves visualising traffic noise along the facades of buildings. A standard traffic noise model is necessary to calculate traffic noise levels, as several factors affect traffic noise. Moreover, designing noise observation points in 3D and spatial interpolation play significant roles in 3D noise visualisation. Therefore, this study demonstrates the results by elaborating on the spatial interpolation and designing noise observation points. A noise observation point consists of four parameters in 3D space. Generally, Inverse Distance Weighted (IDW), Triangular Irregular Network (TIN), and Kriging do not support the interpolation of four parameters in 3D. However, 3D Kriging in Empirical Bayesian Kriging provides significant opportunities to interpolate noise levels in 3D. However, the elements of the function of spatial interpolations are vital for accuracy. The 3D Kriging uses different variograms according to semivariance. This variogram directly impacts the weighting factor of 3D Kriging. Therefore, this study develops a comparison to identify the impact of different variograms on the accuracy of 3D Kriging interpolation on traffic noise.