It is important to be aware that even though input data are given as accurately as possible, there will always be a certain insecurity in a noise calculation. Noise values close to limits that decide if measures should be taken, should therefore be specifically considered for inaccuracies.
Inaccuracies in the calculations results can be caused by:
inaccuracy in input data
inaccuracy in the calculation method
inaccuracy in the implementation of the calculation method
Inaccuracy in input data
Examples of inaccuracies in input data are:
The terrain data (map) is not updated
Terrain data is organized badly in the terrain model
Wrong traffic data is entered for a road
Bad accuracy of height data in the terrain model
Too little terrain and road length are included. You should always include terrain and roads a good distance away from the calculation points in all directions. The noise module doesn't assume anything about the road length, so if it only 200m long in the drawing, you will be missing many possible noise contributions from the rest of the road.
The figure to the right shows how big difference small inaccuracies can make.
A: An inaccuracy of maybe only a few centimeters gives this terrain.
B: Correct terrain
a: Noise source
b: Sight line
c: Calculation point
A major question when calculating the noise along a noise profile with the Nordic method, is whether there is a barrier along the profile. The difference between having a barrier or not is quite big (close to 3dBA). However, this will normally be leveled out, since there are many profiles calculated for each calculation point.
Studies show that the deviation between measured values and calculated values normally are within +- 1dBA.
Inaccuracies in the calculation method
There are limits in the Nordic calculation methods. One example is that it calculates only one "barrier" for each terrain profile. If there is both a local noise barrier and a barrier close to the road, the effect will not be calculated correctly. This is a scenario to keep in mind.
Inaccuracies in the implementation of the calculation method
Studies have shown that the calculations in Novapoint Noise are very close to what is calculated manually with the same method.