Road noise mapping with CNOSSOS – the EU calculation method

Authorities must use the EU’s common noise calculation method, CNOSSOS, for strategic road noise mapping. How do the results differ from previous calculations, and what data does CNOSSOS require?

The Danish Environmental Protection Agency initiated a project to help the relevant authorities understand and obtain the input data required for CNOSSOS calculations.

CNOSSOS is a simpler and less accurate method

CNOSSOS (CNOSSOS-EU) is a noise calculation method developed within the EU. Authorities must use it for the strategic noise mapping carried out every five years. From 2021/2022, CNOSSOS must be used for these calculations throughout the EU, pursuant to the Danish Noise Order (Executive Order no. 2092 of 18 November 2021), which implements the EU Environmental Noise Directive and describes the CNOSSOS method.

CNOSSOS is only to be used for strategic noise mapping within the EU framework. In Denmark, authorities must continue to use the Nord2000 method for calculations relating to environmental cases.

In many respects, CNOSSOS is considerably simpler and less accurate than Nord2000, although it is more detailed in a few areas. The transition to CNOSSOS requires source data, or input data, in a different format than previously, and the method will, to some extent, produce different results from those obtained using the familiar Nord2000 method.

In addition to describing the source and propagation method, the Danish Noise Order also includes a new method for counting noise-exposed homes and people. As a completely new requirement, estimates must also be prepared of the number of people at risk of developing serious noise-related diseases. This is not discussed further here.

Vehicle categories in CNOSSOS

Broadly speaking, CNOSSOS divides traffic into the same vehicle categories as Nord2000. However, in the original source catalogue supplied with the CNOSSOS method, there were relatively large differences in vehicle noise compared with Nord2000.

CNOSSOS produces lower values than Nord2000

Force Technology converted Nord2000 source data for use with the CNOSSOS method on behalf of the Danish Road Directorate. Comparative calculations, as shown in Figure 1, for flat terrain without screening from noise barriers, embankments, buildings, or hilly terrain reveal several differences between the values produced by the two calculation methods.

The difference is smallest at short distances from the road, at 25 metres, and for the higher calculation position, at 4 metres. Further away from the road, the difference becomes greater, reaching up to 5 dB. CNOSSOS therefore generally produces lower values than Nord2000.

The differences are smallest at a calculation height of 4 metres. This is consistent with CNOSSOS having been developed for calculations at a height of 4 metres in urban areas, whereas in Denmark calculations are typically carried out at a height of 1.5 metres to determine noise exposure at a home or other building.

Figure 1: Difference between noise levels calculated using Nord2000 and CNOSSOS for a traffic scenario with different receiver heights of 1.5 and 4 metres and different distances of 25, 100, and 400 metres. Blue: 25 metres from the road; orange: 100 metres from the road; black: 400 metres from the road. The dashed curve applies to a receiver height of 4 metres, and the solid curve applies to a receiver height of 1.5 metres.

The corrections for different asphalt types familiar from Nord2000 can also be converted directly for use with the CNOSSOS method.

Motorcycles and mopeds can generally be excluded from noise mapping

As a new requirement in Denmark, both motorcycles and mopeds must be included in noise mapping if they have a material effect on the overall noise situation. In a city such as Rome, it probably makes sense to include mopeds and scooters, but what about Denmark? Traffic specialists from the Danish Road Directorate assisted with the assessment of this issue.

We first examined the number of registered motorcycles and 45 km/h mopeds in relation to passenger cars, as shown in Table 1.

Category	Number registered
Motorcycles	160,000 (6 %)
45 km/h mopeds	32,000 (1 %)
30 km/h mopeds	210,000* (8 %)
Passenger cars	2,600,000

Number of registered vehicles by category. Percentages in brackets are relative to the number of passenger cars.

Source: Danish Road Directorate. *Source: Estimate based on online source (Wikipedia).

The number of motorcycles is approximately 6% of the total number of passenger cars, while 45 km/h mopeds account for approximately 1%. If it is assumed for estimation purposes that motorcycles and 45 km/h mopeds have the same source strength as passenger cars, a share of 1% would mean that their noise contribution is 20 dB lower than the contribution from cars. A share of 6% would mean a noise contribution 12 dB lower than that from passenger cars.

Furthermore, the total distance travelled by motorcycles and mopeds is considerably lower. In other words, on average, they spend much less time on the roads than passenger cars. We concluded that, unless specific local conditions apply, motorcycles and 45 km/h mopeds can be excluded from the mapping without introducing significant errors.

A sample survey carried out by the Danish Road Directorate covering 31 roads, including motorways, showed that motorcycles and Class 45 mopeds accounted for less than 0.5% of passenger car traffic on average on roads outside urban areas. On roads within urban areas, the proportion was slightly higher, but still below 0.5%.

According to online sources, there are approximately 210,000 ordinary 30 km/h mopeds, corresponding to 8% of the number of passenger cars. With a source strength approximately 2 dB higher than that of passenger cars, as shown in Figure 2, this means that noise from mopeds may be up to approximately 8 dB lower than noise from passenger cars. Excluding 30 km/h mopeds may therefore affect the total noise level by a few tenths of a decibel.

Figure 2: Comparison of source data at different speeds. Data for Categories 1–3 have been converted from Nord2000. Motorcycles and mopeds use standard CNOSSOS data.

We concluded that noise from 30 km/h mopeds should be included in the mapping in specific cases, namely when their number exceeds 10–20% of the number of passenger cars (or other vehicles). The decision must be based on local estimates or actual traffic counts.

Source noise must be corrected at traffic lights and roundabouts

Another new feature in Denmark is that CNOSSOS includes methods for correcting source noise at roundabouts and signal-controlled junctions. Nord2000 allows speeds and acceleration towards and away from a signal-controlled junction to be adjusted manually. However, it is uncertain whether this option was used at all in previous strategic noise mapping carried out with Nord2000.

In CNOSSOS, this is modelled as a correction that increases from 100 metres before a signal-controlled junction or roundabout and decreases until 100 metres after it. The correction depends on the vehicle category. However, the effect on total noise exposure is limited, as shown in Figure 3.

The figure shows the difference between results calculated using Nord2000 and CNOSSOS for a signal-controlled junction where two urban roads with a speed limit of 60 km/h intersect. The calculations were carried out using conditions corresponding to average Danish weather.

No changes to speed or assumptions concerning acceleration were applied in the Nord2000 calculation, while the CNOSSOS calculation included the correction for the traffic lights. Positive values, shown as red areas, mean that Nord2000 calculates higher values than CNOSSOS. Negative values, shown as darker green areas, mean that CNOSSOS produces higher values. The large red areas in Figure 3 represent differences generally seen in road noise calculations, where CNOSSOS typically calculates lower noise levels than Nord2000.

Figure 3: Effect of using CNOSSOS for signal-controlled junctions. The colours show the differences between results calculated using Nord2000, without corrections for traffic lights, and CNOSSOS, with corrections for traffic lights. In the yellow, orange, and red areas, CNOSSOS calculates lower noise levels than Nord2000. Calculation scenario: urban road with a speed limit of 60 km/h and “Danish weather”.

In this specific scenario, the effect of the signal-controlled junction is an increase in the noise level of up to approximately 1 dB within a radius of approximately 100 metres around the junction.

Noise propagation decreases with CNOSSOS

As shown in Figure 3, calculated noise levels will typically be lower with CNOSSOS than with Nord2000. One reason is that the weather conditions in CNOSSOS are typically classified as “favourable” or “neutral”, meaning that there is either a following wind or no wind between the road and the calculation position.

The relevant weather classes used in Nord2000 calculations have been converted into proportions of “favourable” sound propagation that can be used directly in CNOSSOS calculations. Under simple propagation conditions, meaning flat terrain without screening, the differences are approximately ±2 dB within 500 metres of the noise source.

Where the propagation path includes screening, the differences may be in the order of 10 dB. CNOSSOS generally calculates lower values than Nord2000 in situations involving screening.

How to approach strategic noise calculations in future

When no new measurements are carried out and existing data are used, source data converted from Nord2000 can be used for road noise calculations. In principle, this provides the same level of uncertainty for source data as conventional road noise calculations using Nord2000. The same applies to the influence of asphalt surfaces. However, propagation calculations are less accurate with CNOSSOS than with Nord2000, meaning that the overall calculated difference between the two methods can be relatively large.

Based on vehicle registration data and counts carried out by the Danish Road Directorate, it will probably only be relevant to include motorcycles and 45 km/h mopeds in calculations in very exceptional cases. 30 km/h mopeds are more difficult to assess, as this type of vehicle cannot be expected to be included in every counting system because they use the road in some places and cycle paths in others. An initial estimate is that, where the proportion of 30 km/h mopeds is below 10% relative to passenger cars, their effect on the noise level will be minimal.

Force Technology recommends using the built-in CNOSSOS methodology for vehicles approaching and leaving signal-controlled junctions and roundabouts. The methodology may produce a small increase in noise levels around signal-controlled junctions and roundabouts of approximately 1 dB.

Differences between the CNOSSOS and Nord2000 calculation methods mean that CNOSSOS produces lower calculated noise levels than Nord2000. The differences are greatest at longer distances from the source, particularly in hilly terrain and where there are barriers and/or buildings that screen or reflect sound.

The article was originally published in the Danish magazine Trafik & Veje.

Related insights

See all

Reach out to Per Finne for more information.