Noise Modeling, Prediction, and Impact Assessment
Domain 6: Noise Modeling, Prediction, and Impact Assessment
This domain tests whether you can predict noise from sources and interpret models without overstating precision. Expect questions on propagation, model selection, inputs, and validation.
Core rules
- Characterize sources with sound power level or emission level at a reference distance, including octave-band data where the model requires it.
- Geometric divergence: a point source loses about 6 dB per doubling of distance; a line source such as a long roadway loses about 3 dB per doubling in the ideal case.
- Other mechanisms include atmospheric absorption, ground effect, barriers and diffraction, reflections, and vegetation. Do not double-count mechanisms already included in a model's algorithms.
- Common models include ISO 9613-2 for industrial sources and FHWA TNM for highway traffic. Inputs include source strength, distance, terrain, barriers, receptor height, and ground type.
- Validate predictions against measurements where possible and report sensitivity to key assumptions.
Worked example
A point source has Lw = 100 dB. In a free field at 50 m, Lp is approximately Lw - 20 log10(50) - 11 = 100 - 34 - 11 = 55 dB. At 100 m it is about 49 dB, a 6 dB drop. If a barrier provides 10 dB insertion loss, the predicted level at 100 m becomes about 39 dB, subject to diffraction assumptions and receptor geometry.
Traps
Using 3 dB per doubling for a point source; adding barrier loss to a model that already includes it; ignoring meteorology; and presenting modeled results as measurements.