Raw observations, freely available
Every observation collected by TrueNoise is available for download — including aircraft identifiers, operator names, GPS coordinates, and full psychoacoustic profiles. We believe open data makes better science. Use it, build on it, challenge it.
One row per observation. Includes every measurement the app recorded — both entry events (first detection of an aircraft entering the monitoring zone) and track events (continuous measurements as the aircraft passes). All psychoacoustic metrics, all aircraft identifiers, all GPS and ADS-B data.
For health impact analysis, focus on entry rows — these represent
each unique aircraft event. Track rows provide the time-series detail within each event.
The Type column distinguishes them.
On 1 June 2026 a +2.9 dB calibration drift was identified in the TA657A reference meter and corrected. Measurements uploaded before that date carry a known +3 dB systematic bias in absolute SPL. The reference meter has been replaced with a calibrator-verified Wintact SLM-30B and the application offset corrected. See Methodology Section 7b for the full account.
How pre-correction data is used in this dataset:
| Chart / use | Treatment |
|---|---|
| Event counts, frequency by aircraft type | Used unchanged — relative counts unaffected by absolute offset |
| Spectral patterns, C-A delta, aft-aspect geometry | Used unchanged — spectral character independent of absolute calibration |
| Relative trends within a single session | Used unchanged — offset cancels in relative comparisons |
| Absolute SPL claims (Lmax, N65, N70 exceedance) | Post-correction data only, or subtract 3 dB from pre-correction values |
| Loudness in sone for health-impact claims | Post-correction only, or multiply pre-correction sone by 0.81 |
| Comparison to FAA/WHO thresholds (65 dBA, etc.) | Post-correction data only |
Correction math (Stevens power law, valid above 40 phon — covers all aircraft observations):
dBA / Lmax / phon → subtract 3 · Zwicker loudness (sone) → multiply by 0.81 ·
Psychoacoustic annoyance → multiply by 0.81 · Sharpness (acum) → approximately unchanged ·
Health classification → recompute from corrected metrics
Full column definitions and health risk scale descriptions are documented in the methodology page. A quick reference is below.
| Column | Type | Description |
|---|---|---|
| Timestamp | datetime | Date and time of observation (local time zone) |
| Type | string | "entry" = aircraft entered zone; "track" = continuous measurement |
| dBA Level | float | A-weighted sound pressure level in decibels (IEC 61672) |
| Loudness (sone) | float | Zwicker perceived loudness (ISO 532). Doubling = doubling perceived loudness. |
| Loudness Health Impact | string | Risk tier label for loudness value |
| Loudness Level (phon) | float | Loudness expressed in phon for cross-spectrum comparison |
| Sharpness (acum) | float | High-frequency spectral content (DIN 45692). Higher = more piercing. |
| Annoyance | float | Psychoacoustic annoyance index (Zwicker & Fastl composite) |
| Onset Rate (dB/s) | float | Rate of noise level increase. High values trigger startle response. |
| Callsign | string | Flight identifier from ADS-B (e.g. SWA2333) |
| ICAO24 | string | Unique aircraft transponder hex code |
| Type Code | string | ICAO aircraft type designator (e.g. B38M, A320) |
| Type Name | string | Full aircraft type name (e.g. Boeing 737MAX 8) |
| Registration | string | Aircraft tail number |
| Operator | string | Airline or operator name |
| Flight Phase | string | Estimated phase: Taking Off, On Approach, Climbing, Cruising, etc. |
| Ground Distance (mi) | float | Horizontal distance from observer in miles |
| Slant Range (mi) | float | 3D distance from observer in miles (accounts for altitude) |
| Altitude (ft) | float | Barometric altitude in feet |
| Bearing | float | Compass bearing from observer to aircraft in degrees |
| Elevation Angle | float | Look-up angle from observer to aircraft in degrees |
| Speed (kts) | float | Ground speed in knots |
| Climb Rate (fpm) | float | Vertical rate in feet per minute (positive = climbing) |
| Observer Lat / Lon | float | GPS coordinates of the measurement location |
| Windshield Config | string | Windshield configuration used during the session: None, Foam, or Fur. Same value for every row in the session. Blank for sessions recorded before this field was added (all outdoor sessions used Fur). |
| Windshield Correction (dBA) | float | A-weighted insertion loss for the windshield configuration: 0.0 (None), 0.7 (Foam), 1.8 (Fur). Actual noise levels at the receptor are this many dB higher than the reported dBA values. Apply this correction for absolute SPL analysis. |
| Measurement Type | string | Acoustic geometry of the measurement position. One of: Standardized Receptor (ISO 1996 compliant, tripod, ≥1 m from walls — regulatory-comparable); Community Receptor (where someone actually lives — porch, balcony, window — includes +3 to +6 dB boundary reinforcement); Facade-Level (0.5–2 m from a single wall, used for sleep-disturbance and indoor-exposure prediction per WHO L_DEN methodology); Field Characterization (mobile or temporary, strategic location); Hand-held (no fixed mount). Blank for sessions before 18 May 2026. |
| Position Description | string | Free-form description of the exact measurement position within the session. May be empty. Examples: "Tripod, 1.5m, middle of patio" · "Mic clipped to bedroom window sill exterior". Quote-escaped per CSV convention if it contains commas. Blank for sessions before 18 May 2026. |
| C-A Delta (dB) | float | Wind contamination indicator per IEC 61672 / ISO 1996 method. Computed as L_C − L_A over the same 2-second peak window. Threshold guidance: <15 dB = clean; 15–25 dB = possible wind contamination; ≥25 dB = strong wind contamination. Observations with delta ≥25 dB should be excluded from SPL analysis by default. Blank for sessions before 18 May 2026 or when C-weighting was not captured. |
| Excluded | string | Wind contamination exclusion flag set by the observer in the iOS app after reviewing flagged observations. Values: Yes or No (default No). Rows marked Yes are filtered by the upload service before transmission — they will not appear in this dataset. The Excluded column is present in local CSV exports for audit and reversibility. Blank for sessions before this feature was added. |
This data is free to use for any purpose. If you publish findings based on TrueNoise data,
please cite the source:
TrueNoise Community Noise Observatory. Aircraft noise observations, BWI departure corridor.
truenoise.org. Accessed [date].
Questions about the data or methodology? Get in touch.