Noise modelling can range from highly inaccurate, to very accurate modelling. This depends on various modelling assumptions, including:
- The algorithm used (such as Concawe, ISO 9613-2, Nord2000, etc.), with each of these prediction algorithms having certain advantages and disadvantages;
- Assumptions and correction factors included in the noise model (noise sources used, weather conditions, humidity, temperature, ground surface, topography, etc.);
- Time spent on developing the noise model (topographical contours, calculation grids, project layout, screening effects etc.) – as the saying goes: Garbage in, garbage out.
Years ago noise levels were calculated by hand (using Tables and a Slide rule to estimate a noise level), moving to simple calculator and spreadsheet-based modelling, with the latest being Georeferenced 3-D computer models.
Anyone can use a spreadsheet-based simple model considering geometric divergence (highly inaccurate, yet, still used by small companies), free tools such as NoiseModelling, iNoise, dBmap or Noise-Planet, or commercial packages such as NoiseMap, IMMI, CadnaA and SoundPLAN. Each of these options has certain advantages and disadvantages, with EARES selecting to use a commercial package (even though we have developed a noise propagation model in-house).
However, no matter what algorithm or model is selected, experience with the model, the experience of the acoustician and the time used in developing the noise model will and does have a significant impact on the accuracy in calculating the noise levels. While some may think that it is not important, incorrect modelling can have a significant impact on the perception of a project, the selection and the need for mitigation/management measures and the potential authorization of a project.
Unfortunately, spending time to develop detailed topographical contours, adding different noise sources (with spectral emission characteristics, operational characteristics, etc.) and project details (such as opencast mining pits, berms, buildings, etc.) does take significant time, which, unfortunately, increase the cost of a noise study.
While EARES can (and sometimes do) use very simple noise models, we always recommend accurate noise modelling, having seen the benefit to the client and EAP during the EIA stage, with public participation as well as during operation.