13.2 Noise measurement and assessment methods

Noise Measurement Techniques

Noise measurement gives you the tools to quantify how loud an environment actually is and whether that noise poses a risk to health or violates regulations. Without reliable measurement, noise control is just guesswork. This section covers the instruments, metrics, and modeling methods you'll need to understand.

Principles of noise measurement

Sound Level Meters (SLMs) are the primary instrument for measuring environmental noise. An SLM works by capturing sound with a microphone, amplifying the signal, applying frequency weighting, detecting the level, and displaying the result.

A few key settings determine what an SLM actually measures:

• Type classification reflects accuracy. Type 0 is a laboratory reference, Type 1 is precision grade (used for environmental and compliance work), and Type 2 is general purpose (surveys, screening).
• Frequency weighting adjusts the meter's sensitivity to match how humans perceive sound at different frequencies:
  • A-weighting approximates human hearing sensitivity and is the most common for environmental and occupational noise
  • C-weighting has a flatter response and is used for peak measurements and low-frequency noise
  • Z-weighting (zero weighting) applies no filtering at all, capturing the full unweighted spectrum
• Time weighting controls how quickly the meter responds to changes in level. Fast (125 ms) tracks rapid fluctuations, Slow (1 s) smooths them out, and Impulse (35 ms rise, 1.5 s decay) captures sharp transient sounds like hammering or gunshots.

Noise dosimeters are worn on a person to measure cumulative noise exposure over a work shift. They calculate a noise dose, which is the percentage of the maximum allowable daily exposure. The exchange rate is critical here: a 3 dB exchange rate (used internationally per ISO standards) means every 3 dB increase halves the allowable exposure time, while a 5 dB exchange rate (used by OSHA in the U.S.) is more lenient.

Calibration is non-negotiable. You apply an acoustic calibrator (which produces a known sound level, typically 94 dB or 114 dB at 1 kHz) to the microphone before and after every measurement session. If the readings drift by more than the acceptable tolerance (usually ±0.5 dB for Type 1), the data may be invalid.

Measurement protocols standardize how data is collected. These specify microphone placement (height, distance from reflective surfaces), measurement duration, number of samples, and how to account for background noise. If background noise is within 10 dB of the source you're measuring, you need to apply a correction; if it's within 3 dB, the measurement may not be usable.

Interpretation of noise data

Raw sound level readings aren't very useful on their own. Noise metrics condense complex, fluctuating sound into single numbers that describe different aspects of the noise environment.

Equivalent Continuous Sound Level ($L_{eq}$) is the most widely used metric. It represents the steady sound level that would contain the same acoustic energy as the actual fluctuating noise over a given time period $T$:

$L_{eq} = 10 \log_{10} \left[\frac{1}{T} \int_{0}^{T} \left(\frac{p(t)}{p_0}\right)^2 dt\right]$

where $p(t)$ is the instantaneous sound pressure and $p_0$ is the reference pressure (20 µPa). This is the go-to metric for traffic noise studies, industrial noise assessments, and most regulatory contexts.

Day-Night Average Sound Level ($L_{dn}$) is a 24-hour metric that adds a 10 dB penalty to nighttime noise (10 PM to 7 AM) to account for increased sensitivity during sleeping hours:

$L_{dn} = 10 \log_{10} \left[\frac{1}{24} \left(15 \times 10^{L_d/10} + 9 \times 10^{(L_n+10)/10}\right)\right]$

where $L_d$ is the daytime $L_{eq}$ and $L_n$ is the nighttime $L_{eq}$. This metric is commonly used for community noise evaluation around airports and in urban planning. A related metric, $L_{den}$, adds an additional 5 dB penalty for evening hours (used in EU regulations).

Statistical noise levels describe the variability of a noise environment by indicating the level exceeded for a certain percentage of the measurement period:

• $L_{10}$ is the level exceeded 10% of the time, representing the louder portions of the noise climate (often used to characterize traffic noise peaks)
• $L_{50}$ is the median noise level
• $L_{90}$ is the level exceeded 90% of the time, representing the background noise level with occasional louder events stripped away

Other important metrics:

• $L_{max}$ and $L_{min}$ capture the highest and lowest sound levels recorded during the measurement period
• Sound Exposure Level (SEL) normalizes a single noise event to a one-second duration, making it possible to compare events of different lengths. This is standard for assessing aircraft flyovers, train pass-bys, and similar discrete events
• Long-term averages are built from multiple short-term measurements to characterize noise over days, weeks, or seasons

Noise Assessment and Compliance

Noise mapping and prediction

Noise mapping translates measurement data and model outputs into visual representations of how noise is distributed across an area. These maps are typically color-coded contour maps produced using Geographic Information Systems (GIS), where each color band represents a range of noise levels (e.g., 55–60 dB, 60–65 dB).

Since you can't measure every point in a city, interpolation techniques estimate noise levels between measurement locations. Common methods include kriging (which accounts for spatial correlation in the data) and inverse distance weighting (which assumes closer points have more influence).

Noise prediction models simulate how sound propagates from source to receiver. They all follow the source-path-receiver framework:

1. Source: Characterize the emission (sound power level, directivity, spectrum)
2. Path: Account for geometric spreading, atmospheric absorption, ground effects, terrain shielding, and meteorological conditions (wind, temperature gradients)
3. Receiver: Determine the resulting level at the point of interest

Several standardized models exist for different applications:

• ISO 9613 for industrial and general environmental noise
• CNOSSOS-EU for strategic noise mapping across the European Union (covers road, rail, aircraft, and industrial sources)
• NMPB-Routes for road traffic noise (French method, widely adopted)

Software packages like SoundPLAN, CadnaA, IMMI, and NoiseModelling implement these models and handle the complex calculations. After running a model, you validate it by comparing predicted levels with actual field measurements at selected points.

Scenario analysis is where prediction models become especially powerful. You can test the effect of proposed interventions (adding a noise barrier, rerouting traffic, changing building layouts) before any money is spent.

Compliance with noise standards

Noise regulations exist at multiple levels. International standards like the ISO 1996 series define measurement and assessment procedures, while the WHO Environmental Noise Guidelines recommend exposure limits based on health evidence (e.g., $L_{den}$ below 53 dB for road traffic noise to avoid serious health effects). National and local regulations then set enforceable limits that vary by land use zone (residential, commercial, industrial) and time of day.

A noise impact assessment typically follows these steps:

1. Conduct a baseline noise survey to document existing conditions
2. Predict future noise levels using models that account for the proposed development or change
3. Compare predicted levels against applicable regulatory limits and guidelines
4. If levels exceed limits, identify and recommend noise control measures

Noise control measures fall into three categories, applied in order of preference:

• Source control: Use quieter equipment, modify operations, reduce speeds, or limit hours of noisy activity
• Path control: Install noise barriers, increase the distance between source and receiver, or use terrain to block sound
• Receiver control: Improve building sound insulation, install acoustic windows, or adjust land use planning to keep sensitive uses away from noise sources

Compliance reports must document the measurement methods used, state the measurement uncertainty (typically ±1–3 dB depending on conditions and instrument class), provide clear compliance or non-compliance statements, and recommend noise reduction measures where needed.