Freeway and road noise in Los Angeles

Modeled estimates for general information only — not measurements, not a guarantee, and not for legal, medical or real-estate decisions.

Road traffic is the main layer of the Quiet LA map. Every road is a sound source, and its strength comes from how many vehicles use it, how fast they go and how many are trucks.

See road noise on the map

How Quiet LA models freeway and road noise

The model uses NoiseModelling 6, open-source software that follows CNOSSOS-EU, the European Union’s common method for mapping road, rail and aircraft noise. It runs separately for the day, evening and night periods.

For each period the model calculates how sound travels from every road out to buildings and open ground. It accounts for spreading with distance, the ground, hills, buildings and freeway sound walls. Reflections between buildings are not included yet.

The weather assumed is typical for Los Angeles, with sound carrying farther in the evening and at night than in the day.

Parks, mountains and desert get a coarse estimate from freeways, ramps and major roads only, on a 100 m grid with no buildings and no local streets. Where the detailed model has a value, it always wins. Land with no freeway or major road within 3 km shows the lowest color; that is an assumption, not a computed value. Other blank areas are not modeled yet, and blank never means quiet.

The detailed model looks 1.5 km out from each road. Traffic farther away adds a low hum, so a coarser calculation of freeways and major roads out to 8 km, reduced for the buildings in between, is added to the map colors. It is not in the values for a single address, and it is the least certain part of the map.

What drives road noise in the model

• Traffic volume. Counts come from Caltrans on state highways, the Federal Highway Administration (HPMS 2024) on federal-aid roads and the City of Los Angeles on city streets. Evening traffic is set at 0.6 times, and night at 0.2 times, the daytime flow.
• Typical values where there is no count. Most residential streets have no count, so the model uses typical values for that kind of street. Those streets are estimates only. On the map, dashed road lines mark roads with no count.
• Speed. The model uses the posted speed limit where it is known and a typical speed otherwise, the same all day. It does not model stop-and-go congestion.
• Heavy vehicles. The share of trucks comes from state truck counts and federal vehicle classes. Trucks are louder than cars.
• Ground and terrain. Terrain comes from the 2023 USGS lidar scan at 10 m, and sound bends over hills.
• Buildings. Footprints and heights come from LA County. Every building wall is an obstacle.
• Sound walls and bridges. Freeway sound walls found in the lidar are included, and roads on bridges are modeled at deck height. The wall detection is incomplete, and walls along other streets and private walls are not in the model.

How sound drops with distance from a freeway

This section is general acoustics from Caltrans’ Technical Noise Supplement, not a prediction for any address.

A steady stream of traffic acts like a line source. Spreading alone lowers its sound level by about 3 dB each time the distance from the road doubles. A single point source, such as one stationary machine, drops about 6 dB per doubling.

Soft ground and other effects add to that drop, and the simple per-doubling rule is only a rough guide beyond about 200 feet. Caltrans notes that the first row of buildings beside an at-grade highway typically shields the rows behind it by about 3 to 5 dB. A noise barrier whose top just reaches the line of sight between the road and a listener gives about 5 dB.

A difference of 3 dB is barely perceptible, and a change of about 10 dB is heard as twice or half as loud. See what 45, 55 and 65 dB mean.

How accurate is the freeway and road layer?

At a single spot the model is typically off by about 6 dB, higher or lower. On average it is within about half a decibel of public 24-hour measurements.

Near freeways the model runs about 4 dB loud on average, with a wide spread. For major streets the pattern differs by measurement year. The model has not been tuned to fix this, because the evidence does not yet support one simple correction.

Missing local counts, congestion, building reflections and sound walls that were not detected can all move a real level away from the modeled one. See Limits and How it’s made.

Who to contact about freeway noise

Caltrans runs the state highways. For sound walls and who can ask for one, see who to call about noise in Los Angeles.