How accurate is Quiet LA?

Updated 25 October 2026

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

More accurate than the federal noise map. Checked against 118 real sound measurements across LA County, Quiet LA came closer at 2 out of 3 sites. Typical error: 4.6 dB vs 7.7 dB.

Federal map: US DOT National Transportation Noise Map, 2020 data, built for national trends. Compared 25 October 2026.

How does Quiet LA compare with the federal map?

The table gives the result at the 118 sites that have a measured day, evening and night level. That allows an exact 24-hour average level, the quantity the federal map reports. Each figure is the map’s value minus the measurement, so a plus sign means the map reads louder than the measurement and a minus sign means lower.

For scale, a gap of 8 dB is one your ear hears as roughly half the loudness. The “typical error” is the median: half of the sites are closer than that. Every site is listed at the end of this page.

How was the test done?

Quiet LA and the federal map were compared with the same public sound measurements that Quiet LA is checked against: 160 usable measurements from public environmental reports across LA County. 118 of them have separate day, evening and night values, and the table above uses those. For each site the question is the same: which map’s value is nearer what was measured?

• Same places, same rule. Quiet LA and the federal map were each read in a small area around the recorded position of each site, the same way for both.
• The federal map. The US Department of Transportation’s Bureau of Transportation Statistics publishes the National Transportation Noise Map. The test uses its 2020 data year, the 24-hour average level of aviation, road and rail together. The numbers were read from Esri’s copy of the same data and checked against the Bureau’s own map tiles, where 98.7% of cells agree within one color class.
• The 45 dB floor. The federal map publishes no values below 45 dB. Where it shows nothing at a site, the test counts 45 dB. That is higher than whatever the map would show, so it makes the federal map look closer to the measurement than it is. These sites read “below 45” in the table.
• What Quiet LA’s number is. Its value here is roads and trains at open ground, 1.5 m up, the part of the model that is checked against measurements. Aircraft, the distant traffic hum, sirens and the coarse open-land estimate are left out, so this value is not the map’s color everywhere. Sites where aircraft noise was close to road noise were left out of the test.
• Years. The federal values are for 2020, a year of unusual traffic in Los Angeles that its documentation does not size. The measurements were made between 2001 and 2024, and Quiet LA uses 2024 traffic.

The same comparison in CNEL, the 24-hour level Quiet LA publishes, on all 160 sites gives the same picture. The federal map reports a plain 24-hour average, so 5.0 dB was added to its values: the gap between CNEL and the 24-hour level in the measurements themselves.

How does it compare with HowLoud?

Closer to real measurements than the federal noise map and HowLoud.

Tested 25 October 2026 on 160 public sound measurements in LA County. HowLoud gives a 0 to 100 score, so the maps were compared on how well they rank places from loud to quiet.

HowLoud is a commercial tool that gives an address a score from 0 to 100 instead of a level in decibels, so its error in decibels cannot be worked out. What can be compared across all three maps is rank agreement, which says how closely a map puts the measurement sites in the same order as the measurements: 1 means the same order, 0 means unrelated. It is the order of the sites; no area is rated or ranked.

On this main measure Quiet LA ranked clearly closer than both; “clearly” means the gap is bigger than the chance variation between samples of sites would explain. No map ranked closer than Quiet LA in any variant we tried, but some stricter variants show no clear difference, for example when only the sites where the federal map shows a value are used, or with a second rank measure.

• Order is not level. A map can put places in the right order and still be wrong in decibels, and the reverse. Level error in decibels, as in the first table, can be compared only with the federal map.
• One point against an area. HowLoud was asked for one point at each site, while Quiet LA and the federal map were read over a small area around it, which may favor those two.
• Different ingredients. HowLoud’s score also counts air traffic and local sources such as restaurants and schools, which Quiet LA’s tested value leaves out. A measurement contains whatever was audible, so that could help or hurt HowLoud here.

We do not republish HowLoud’s scores for single sites, so the site-by-site table at the end of this page shows Quiet LA and the federal map only.

What are the limits of this test?

• The federal map is not built for single places. Its documentation says it tracks transportation noise across the country, to follow trends, and should not be used to judge an individual place or time. It leaves out local streets and the shielding of buildings, walls and terrain. This test asks it something it does not claim to do: the result says how each map does at a place, not how well the federal map does its own job.
• Positions are approximate. Each measurement’s position is known only to within 60 to 150 m. Even a perfect model would miss by several decibels at some sites for that reason alone, and so would the federal map and HowLoud.
• Old, single-day measurements. They are 24-hour records from public environmental reports made between 2001 and 2024, not annual averages. Older ones tend to read lower than a model built on 2024 traffic, which counts against Quiet LA here, not for it.
• The sites are near roads. Environmental reports measure near roads, so places far from roads are barely tested.
• Some sites were seen while the model was built. The corridor sites along I-405, Sepulveda Boulevard and Victory Boulevard were looked at during development, so they are not a blind test. The other sites were checked after the model’s choices were made, and nothing was tuned to them. On those, in CNEL, Quiet LA is −0.1 dB from the measurements on average and the federal map −9.6 dB.
• Near freeways the two maps are closer. Their typical errors are not clearly different at freeway sites. Quiet LA tends to read louder than the measurements there and the federal map lower; Limits has the details.
• All the maps are estimates. None is a reading at your door.

What was not tested?

• Other noise maps and reports. Only the federal map and HowLoud’s score have been tested. HowLoud’s paid per-address reports were not, nor was Realtor.com’s noise rating, so this page says nothing about them or about any other tool. See noise maps for Los Angeles, compared.
• The federal map’s newer edition. An edition with 2022 and 2023 data was released in 2026. We could compare it only by its published color classes, not by numbers; that comparison shows the same pattern, a little lower still. The figures on this page are for the 2020 data.
• Aircraft. The federal map’s aircraft layer was not compared with airport monitors. Quiet LA’s own aircraft estimate is checked separately, against 57 permanent airport monitors (see How it’s made and LAX and aircraft noise).
• Indoor sound, and noise from sources the map does not model. Sources & methods lists what it does.

Open the Quiet LA noise map

Site-by-site results

All 118 sites, in the 24-hour average level (dB). “Measured” is worked out from each report’s day, evening and night values. “Below 45” means the federal map shows no value there; the statistics above count 45 dB. No addresses or names are given. Sites are grouped by region only to keep the table readable: they are not a sample of any region and say nothing about how loud an area is.

The method behind the map is on How it’s made and Sources & methods. For the units, see what CNEL means and what 45, 55 and 65 dB mean.