The future noise mapping model in Sweden

Sammanfattning

An extensive technological shift is currently taking place to mitigate climate changeand this trend is particularly noticeable in the transport sector. This is interestingfrom an acoustical perspective, since it changes the noise environment in society.For example, a study in Gothenburg has shown that a complete electrification ofthe road traffic would reduce the noise levels by between 2 and 5 dB(A). In Sweden,noise emissions are calculated with a calculation model from 1996, called the Nordiskberäkningsmodell (Nordiska). Given the age of the model it is reasonable to investigate whether Sweden should change completely to the EU-common calculationmodel Common NOise aSSessment methOdS (CNOSSOS), since it is mandatory touse for national noise mapping. This master thesis has performed a computation analysis to compare and discussdifferences between CNOSSOS and Nordiska, to contribute to answering the question whether Sweden should change to CNOSSOS (or perhaps another model). Theresults show that CNOSSOS overall computes higher noise levels than Nordiska andthat the differences between them increase linearly with distance. Farthest from thenoise source the differences are up to 5 dB(A) for the road case and 9 dB(A) forthe railway case. In other words, the differences are larger for the railway trafficmodels than they are for the road traffic models, which is thought to be a result ofthe complexity of the CNOSSOS railway model. Another interesting phenomenonis that the differences behind buildings between the models are different for roadand railway traffic, which can be explained by the fact that the screening effects inNordiska’s road and railway models are different. My conclusion is that CNOSSOS is unsuitable for domestic calculations of noiseemissions. The model does not align with Swedish legislation and there is uncertaintydue to the fact that the differences between the CNOSSOS and Nordiska road andrailway models are different in size. Moreover, CNOSSOS railway model requires alot of computational power, which can delay and increase the costs of noise mappingor reduce the accuracy of the results. However, additional work is needed in whicheach calculation model is compared with measurements in situ to see which modelbest describes reality. If the conclusion thereafter is that CNOSSOS still is not asuitable option, it could be examined whether it is possible to create an updatedversion of Nord2000 (another Nordic calculation model used e.g. in Denmark) toobtain a calculation model that is more suitable for future traffic conditions.