Methodology

HowNoisy estimates outdoor environmental noise exposure around a geographic location — road traffic, railways, aircraft, industry and nightlife — using official noise maps and open geographic data.

HowNoisy does not measure indoor sound levels, building sound insulation, facade acoustic performance, bedroom comfort, neighbour noise or construction quality. It is not an acoustic simulation and not a measurement. For any of those you need an on-site acoustic study.

1. The address is located (CartoCiudad for Spain, OpenStreetMap elsewhere) and the country of the point is identified.

2. Official strategic noise maps that cover that country are queried exactly at the point: in Spain the Ministry for the Ecological Transition (MITECO / SICA), the European Environment Agency (EEA) and, in the Valencian Community, the Generalitat Valenciana; in the other covered European countries, the EEA.

3. OpenStreetMap is queried for the nearest main roads, railways, airports, industrial land and nightlife venues. Distances are measured to the nearest part of each road or railway, not to its centre.

4. The coverage engine decides how much the available information supports, and the scoring engine produces the score.

Every input keeps its provider, dataset, reference year, retrieval date, source URL and licence, shown under “View data sources” on each result.

The score runs from 0 to 100, where 100 indicates lower expected environmental noise exposure and 0 indicates higher expected environmental noise exposure. 100 is not “perfect” and 0 is not “terrible” — they are ends of an exposure range.

Where an official band covers the point, the governing Lden band sets the score. These are HowNoisy classification thresholds, not legal limits:

• Below 50 dB LdenScore 92
• 50–55 dB LdenScore 80
• 55–60 dB LdenScore 66
• 60–65 dB LdenScore 50
• 65–70 dB LdenScore 34
• 70–75 dB LdenScore 18
• Above 75 dB LdenScore 8

Where no official band exists, independent secondary factors are combined with fixed relative weights:

• Roads40%
• Railways15%
• Airports20%
• Industry10%
• Nightlife15%

Secondary factors produce a risk classification. They are never translated into a decibel level, because no source provides one for that point.

HIGH — an official strategic noise map band covers the point. The official Lden band is the primary indicator.

MEDIUM — no official band covers the point, but at least 3 independent datasets (always including the road network) returned usable information, which supports an estimated assessment. It is expressed as a risk class, never in decibels.

INSUFFICIENT — the available information does not support a meaningful assessment. No score is produced. This is a deliberate outcome, not an error.

Confidence is always decided by the coverage engine, never by the interface.

The same point can appear in several official maps (for example, a regional 4th-round road map, the MITECO 3rd-round map and the EEA compilation). Within each source family (roads, railways, aircraft, city maps) the most recent mapping round wins, because it supersedes the older one. Across families, the loudest band governs the score. Every band found is still listed on the result, with its provider and round.

Official strategic maps only draw bands from 55 dB Lden (50 dB at night) and only for major sources: roads with more than 3 million vehicles a year, railways with more than 30,000 trains a year, major airports and cities over 100,000 inhabitants.

In Spain, where all official sources are queried, “no band reaches this point” means no mapped major source reaches 55 dB Lden there. Local streets in towns that are not mapped agglomerations are not covered, and the result says so.

Outside Spain, the EEA compilation depends on what each country reported, so a missing band is stated as “no band in the European data”, never as proof of low noise.

Official noise maps describe a reference year (for example 2017 for the 3rd round and 2022 for the 4th). HowNoisy shows that year separately from the date the data was retrieved. A map with a 2022 reference year is not a measurement taken today.

Conditions may have changed since the source data was created: roads are rerouted, traffic changes, venues open and close, and noise barriers are built. OpenStreetMap is edited by volunteers and its completeness varies by area. A result describes the environment recorded in the referenced datasets, around the point assessed.

Every result shows the engine version that produced it. When thresholds, weights or source rules change, the version changes too, so results remain traceable to the methodology that generated them.