Today's models for road traffic noise are based on the Nordic prediction model from 1996. Since then, several projects have been carried out to develop prediction models, both in Scandinavia and in Europe. In other Nordic countries for example the calculation model Nord2000 Road is used and within Europe the model CNOSSOS-EU will be used for noise mapping. Nord2000 Road has the advantage that the model has a broader scope and can provide more detailed results compared with 1996 model, and there is a willingness to move to that model in Sweden. The model gives slightly higher estimated levels compared to the old model. Input data for the prediction model Nord2000 Road are based on measurements carried out in the Nordic countries and the latest measurements in Sweden are 10 years old. In 2015 new measurements of noise emissions from vehicles in real traffic on Swedish roads were carried out. The results show that the source data to Nord2000 Road needs to be updated to represent current Swedish conditions. The results indicate that Nord2000 Road overestimate levels and suggested adjusted input data are presented. Spectrum adaptation to Swedish conditions are revised. The measurements confirm, however, speed coefficients for Swedish conditions.

Spectrum adaptation terms for other road surfaces than SMA 0/16 (stone mastic asphalt with 16 mm maximum aggregate size, ABS 0/16), which is the most common pavement for high traffic roads in Sweden require more extensive measurement data, but some data for a Swedish drainage asphalt road surface are presented in the report.

Updated input is proposed for Category 1, 2 and 3 for Nord2000 Road. For motorcycles input data can be based on data from previous Harmonoise- and Imagine projects, even if they are designed for Southern European conditions. Some data from the measurement series 2015 are presented in the report, but additional measurements needed when the spread in the results are great. Also for vehicles with alternative powertrains such as electric or hybrid vehicles sufficient information is lacking today to determine reliable input data for the prediction models for Swedish conditions.

CNOSSOS-EU underestimates clearly the emission levels for Swedish conditions, and also have different speed-dependency, so a speed-dependent correction is proposed in the report. Maximum levels are proposed to be calculated in basically the same way in CNOSSOS-EU as is made in Nord2000 Road.

To calculate the traffic noise at different driving conditions such as in urban areas with accelerations or decelerations associated with intersections or roundabouts it is proposed that the methods of CNOSSOS-EU and Nord2000 Road for heavy vehicles (category 2 and 3) is applied. CNOSSOS-EU gives corrections for the total sound power depending on distance from the crossing or roundabout, while Nord2000 Road gives a general increase of the rolling noise with 3 dB for heavy vehicles in urban traffic.

The Nord2000 Road source model allows for separation of rolling and propulsion noise from measurements in real traffic. For the CNOSSOS-EU model this is more difficult because of the co-localization of the two model sources. Separation of rolling and propulsion noise could be beneficial for separation of tire/road noise and vehicle noise for evaluation of the tire noise and vehicle noise regulations.