Airport flight-path noise contour mapping for residential land valuation
Satellite-derived ADS-B trajectory data, combined with published noise propagation models, produces property-level noise exposure scores that track real operational change rather than the static contours airports publish every few years.
Sensors
- Spire Global ADS-B (LEMUR-2 constellation): Over 110 satellites in low Earth orbit receiving 1090 MHz ADS-B Mode S transponder messages. Provides position, altitude, speed and ICAO identifier for equipped aircraft globally. Latency from message receipt to delivery typically under 30 minutes; archive from approximately 2016 onward.
- Aireon ADS-B (hosted on Iridium NEXT): 66 operational satellites plus spares hosting Aireon payloads. Designed for oceanic and polar ATC, but the same trajectory feed covers continental airspace. Particularly valuable for approach and departure tracks over terrain with sparse ground-station coverage.
- Sentinel-2 MSI: 10 m resolution optical imagery in 13 spectral bands, 5-day revisit at the equator with both satellites. Used here for land-use and land-cover classification around the airport, distinguishing residential parcels from commercial, industrial and green space to contextualise noise exposure scores.
- OpenStreetMap / cadastral data (ancillary): Not a satellite sensor, but parcel-level boundary data is required to assign noise scores to individual properties. Integrated with the raster noise surface to produce per-parcel statistics.
Why the official contour is already wrong
Airports in most jurisdictions publish noise contours, expressed in day-night average sound level (DNL) or the European Lden metric, on a cycle tied to planning regulations rather than to operational reality. In the United Kingdom, the Civil Aviation Authority updates designated noise contours at Heathrow, Gatwick and Stansted on an annual basis, but many smaller airports operate on far longer cycles. A new runway threshold, a changed preferential runway direction, a seasonal schedule shift, or the retirement of a noisier aircraft type can each move the 57 dB Lden contour by hundreds of metres. Properties that sit just outside the published boundary may be absorbing materially more noise than the planning record suggests.
The discount that noise imposes on residential land values is well-documented in the hedonic pricing literature. Studies using UK Land Registry data have found price reductions in the range of one to two per cent per decibel of additional DNL exposure in the 55 to 75 dB range, though the gradient varies with local housing market conditions, proximity to public transport and the availability of noise-attenuated housing stock. The point is not the precise coefficient; it is that the exposure score feeding into any valuation model needs to reflect what aircraft are actually doing today, not what they were doing when the last contour was drawn.
What ADS-B trajectory data actually contains
ADS-B (Automatic Dependent Surveillance-Broadcast) is a surveillance technology mandated for most commercial aircraft in controlled airspace across Europe, North America and increasingly elsewhere. Each equipped aircraft broadcasts its GPS-derived position, barometric altitude, ground speed and heading at roughly once per second. Satellite-based receivers, unlike ground stations, have no terrain-masking problem and no coverage gap over water or sparsely populated land. Spire Global's LEMUR-2 constellation and the Aireon payload on Iridium NEXT both collect these messages globally.
From a stream of ADS-B messages, it is straightforward to reconstruct individual flight tracks: the sequence of latitude, longitude and altitude points from pushback to gate. Aggregate tens of thousands of such tracks over a defined period and you have a statistical picture of where aircraft fly, how low they are, and how often. That picture is the raw material for noise modelling. The Federal Aviation Administration's Aviation Environmental Design Tool (AEDT) and its European equivalents use exactly this kind of track data as input, applying published noise-power-distance curves for each aircraft type to estimate sound levels at ground points.
One honest caveat: not every aircraft is ADS-B equipped. Older general aviation and some military traffic are absent from the feed. For major commercial airports, where the noise burden is dominated by wide-body and narrow-body jets, ADS-B coverage is effectively complete for the aircraft types that matter most to a residential valuation.
From tracks to a property-level noise surface
The processing chain has four steps. First, raw ADS-B messages are decoded, deduplicated and assembled into flight tracks, filtered to the airspace volume relevant to the airport of interest, typically within 30 nautical miles and below 10,000 feet. Second, each track segment is attributed with aircraft type using the ICAO 24-bit address, which maps to a registration and thus to a published noise certification category. Third, a noise-power-distance model is applied to each track segment to estimate the sound exposure level (SEL) at a grid of ground points, usually at 100 m or 250 m spacing. The FAA publishes NPD curves for hundreds of aircraft types in the AEDT documentation; these are the same curves used in official contour production. Fourth, SEL contributions from all flights over the chosen averaging period are summed logarithmically to produce DNL or Lden at each grid point.
The resulting raster surface is then intersected with a parcel boundary layer. Each residential parcel receives a noise score: the area-weighted mean Lden across its footprint, or the value at its centroid for simpler applications. Sentinel-2 land-cover classification confirms which parcels are genuinely residential and flags mixed-use or commercial parcels that should be excluded from a residential valuation model. The output is a scored property dataset, refreshable whenever a new block of ADS-B data is processed.
Honest limits of the method
Noise propagation is not purely a function of aircraft position. Ground reflection, atmospheric refraction, local topography and building shielding all affect what a resident actually hears. The NPD-curve approach used in AEDT and similar tools is a statistical model calibrated against measurement campaigns; it gives reasonable ensemble averages but will be wrong for individual properties in complex terrain or dense urban canyons. If a valuation requires defensible site-specific accuracy, ground-truth noise measurements remain necessary.
ADS-B altitude accuracy depends on the barometric altimeter setting broadcast by the aircraft. Near the ground, where noise matters most, altimeter errors of 50 to 100 feet are common. That translates to a modest uncertainty in the computed ground-level noise estimate but is unlikely to shift a property across a meaningful contour boundary.
Finally, the method captures what aircraft do in the air. It does not capture ground-run noise, reverse-thrust noise on landing, or the noise from airport ground operations, all of which can be significant for properties immediately adjacent to the airport boundary. Those sources require separate treatment.
What changes when you update the contour dynamically
The practical value for a property investor or lender is the ability to rerun the analysis against a specific historical window or a projected future schedule. An airport proposing a new flight procedure, a change in preferential runway use, or a capacity increase can be modelled before the official contour is published, sometimes years in advance of any planning document. Conversely, a portfolio manager holding residential assets near an airport that has lost significant traffic, as happened at many European airports during 2020 and 2021, can quantify the noise-exposure improvement and its implied valuation effect.
Satellize applies this kind of trajectory-based noise analysis as part of its satellite-data analytics work. The methodology is the same whether the airport is a major hub or a regional field with no published contour at all, which describes a large proportion of airports globally. For clients building or stress-testing residential land valuations, a request for a baseline noise-exposure dataset for a defined airport catchment is a reasonable starting point.
Integrating noise scores into a valuation model
A noise-exposure score is an input, not a valuation. To translate it into a price effect, analysts typically run a hedonic regression against observed transaction prices, using the noise score alongside standard controls: floor area, age, tenure, school catchment, public transport access and so on. The noise coefficient from such a regression is site-specific and should be estimated locally rather than borrowed from studies at different airports in different housing markets.
Where transaction data is thin, the noise surface can still serve as a stratification tool. Properties can be grouped into bands, say below 55 dB Lden, 55 to 63 dB, 63 to 72 dB, and above 72 dB, which correspond roughly to the planning thresholds used in UK and EU noise policy. Comparable sales analysis within each band is more meaningful than comparables drawn from across the full catchment. Even without a regression, the band assignment tells a buyer, lender or insurer something concrete about relative exposure that a postcode or a straight-line distance from the runway does not.
Typical figures
| ADS-B position update rate | Approximately 1 message per second per aircraft; satellite receiver capture rate varies with constellation geometry, typically multiple passes per hour per location |
| ADS-B horizontal position accuracy | GPS-derived; typically better than 10 m for modern avionics, though satellite-received messages may have gaps between passes |
| ADS-B altitude accuracy | Barometric; ±50 to 100 ft common near ground level |
| Sentinel-2 MSI spatial resolution | 10 m (visible and NIR bands used for land-cover classification) |
| Sentinel-2 revisit | 5 days at equator with both satellites; cloud cover can reduce usable acquisitions |
| Noise surface grid resolution | 100 m to 250 m typical, depending on computational budget and parcel density |
| ADS-B archive depth (Spire) | From approximately 2016; coverage density improves markedly from 2019 onward as constellation grew |
| Aircraft type coverage | Effectively complete for commercial jet and turboprop traffic in controlled airspace; general aviation and some military excluded |
| Output noise metric | DNL (US) or Lden (EU), computed per AEDT-compatible NPD curves; SEL also available per flight event |
| Delivery formats | GeoTIFF noise raster, GeoJSON or Shapefile parcel score layer, CSV per-parcel table, PDF summary report |
Analytics Satellize can run
| Baseline noise-exposure surface | ADS-B track aggregation plus NPD-curve noise propagation model (AEDT methodology) | GeoTIFF raster at 100 m or 250 m resolution, Lden or DNL metric, for a defined airport catchment and averaging period |
| Per-parcel noise score dataset | Zonal statistics intersection of noise raster with cadastral parcel boundaries; Sentinel-2 land-cover mask to isolate residential parcels | GeoJSON or Shapefile with Lden score, noise band classification and land-cover type per parcel |
| Schedule-change impact comparison | Paired noise surfaces computed from two ADS-B windows (before and after schedule change); pixel-level and parcel-level difference map | Difference raster and parcel-level delta-Lden table; PDF summary of affected property count by band shift |
| Projected contour under proposed procedure | Synthetic track generation from published instrument procedure waypoints and altitude profiles; NPD modelling as above | Modelled Lden contour polygons (55, 60, 65, 70 dB) as Shapefile, with affected parcel count per band |
| Noise-band stratification for comparable sales analysis | Assignment of transactions in client dataset to noise bands; summary statistics per band | Annotated transaction CSV with noise band column; summary table of median price per sq m by band |
| Portfolio noise-exposure audit | Batch parcel lookup against pre-computed noise surfaces for multiple airports; flagging of assets above planning thresholds | Portfolio-level CSV with per-asset Lden score and regulatory band flag; GIS layer for mapping |
Who does the work
We can get this done for you. Satellize runs its own analyst desk and a strong science team. You do not buy a data feed and work out what it means; our people source the imagery, run the analysis described on this page, and hand you the answer with its confidence limits stated. Discuss this requirement.