Hourly Global ADS-B Kinematic Sensitivity Dataset (2025)

Creators Eugene PikMevocopter AerospaceContact: [e-mail] Description This Zenodo record provides both a global, hourly ADS-B kinematic sensitivity dataset and the Python and SQL code used to generate it for the 2025 calendar year. No anomaly detection, labeling, attribution, or causal inference is performed. Contents of This Release This DOI contains the following primary artifacts: 1. Dataset sensitivity_analysis_V2.5-2025_hourly.csv A global, hourly, system-level dataset summarizing the distributional properties of successive ADS-B kinematic reports. Each row corresponds to one UTC hour. Group Column(s) Description Unit Temporal Indexing year, month, day, hour UTC timestamp identifying the hourly aggregation window — Sampling Volume processed_adsb_messages Total number of quality-filtered ADS-B state vectors processed in the hour count Speed (Derived from Position) mean_speed, std_speed Mean and standard deviation of inferred ground speed from great-circle displacement m/s sp_25, sp_75, sp_95, sp_99, sp_999 25th, 75th, 95th, 99th, and 99.9th percentiles of inferred speed m/s speed_cv Coefficient of variation of inferred speed dimensionless Vertical Motion mean_vr, std_vr Mean and standard deviation of inferred vertical rate m/s vertical_rate_cv Coefficient of variation of vertical rate dimensionless Altitude Consistency mean_ad, std_ad Mean and standard deviation of absolute difference between geometric and barometric altitude m ad_25, ad_75, ad_95, ad_99, ad_999 25th, 75th, 95th, 99th, and 99.9th percentiles of altitude disagreement m alt_diff_cv Coefficient of variation of altitude disagreement dimensionless Bearing Dynamics mean_bc, std_bc Mean and standard deviation of absolute bearing change between successive segments deg bc_25, bc_75, bc_95, bc_99, bc_999 25th, 75th, 95th, 99th, and 99.9th percentiles of bearing change magnitude deg bearing_change_cv Coefficient of variation of bearing change dimensionless Temporal Spacing median_dt, p95_dt Median and 95th percentile of time difference between consecutive messages s Instantaneous Kinematic Jumps mean_delta_speed, std_delta_speed Mean and standard deviation of absolute change in reported speed m/s delta_speed_p_95, delta_speed_p_99 95th and 99th percentiles of absolute speed change m/s mean_delta_heading, std_delta_heading Mean and standard deviation of absolute change in reported heading deg delta_heading_p_95, delta_heading_p_99 95th and 99th percentiles of absolute heading change deg mean_delta_vertical_rate, std_delta_vertical_rate Mean and standard deviation of absolute change in vertical rate m/s delta_vr_p_95, delta_vr_p_99 95th and 99th percentiles of absolute vertical-rate change m/s Aircraft Sampling Diagnostics mean_pairs_per_aircraft Mean number of valid successive message pairs per aircraft in the hour count p95_pairs_per_aircraft 95th percentile of message-pair count per aircraft count max_pairs_per_aircraft Maximum number of message pairs for any single aircraft count Distribution Tail Structure speed_tail_ratio Ratio of extreme to upper-tail speed percentiles dimensionless alt_tail_ratio Ratio of extreme to upper-tail altitude disagreement percentiles dimensionless bearing_tail_ratio Ratio of extreme to upper-tail bearing change percentiles dimensionless Composite Descriptor stability_score Weighted composite summarizing distributional tail heaviness dimensionless Explanatory Notes Inferred Speed (m/s):speed_m/s = ST_Distance(lon,lat, plon,plat) / Δt — great-circle distance divided by time difference. Vertical Rate (m/s):vertical_rate_m/s = |geoaltitude - previous_geoaltitude| / Δt — change in geometric altitude over time. Altitude Difference (m):alt_diff_m = |geoaltitude - baroaltitude| — absolute difference between geometric and barometric altitudes. Bearing Change (deg):Δθ = θ_t - θ_{t-1} — difference of successive bearings; statistics use |Δθ|. Coefficient of Variation (dimensionless):CV = σ / μ — standard deviation divided by mean; applied to speed, vertical rate, altitude difference, bearing change. Tail Ratios (dimensionless):Tail_Ratio = P_{99.9} / P_{99} — extreme/upper-tail ratio; computed separately for speed, altitude difference, and bearing change. Composite Stability Score (dimensionless):stability_score = 0.4 * speed_tail_ratio + 0.3 * alt_tail_ratio + 0.3 * bearing_tail_ratio — weighted sum describing distributional tail heaviness; purely descriptive. All values are distributional summaries only. No per-aircraft or per-flight labels are produced. 2. Python and SQL Code aircraft_kinematic_trajectory-sensitivity_analysis-V2.5-final.py A fully commented, public Python script that reproduces the dataset exactly when executed against the OpenSky Network historical Trino database, subject to data availability and access policies. Key characteristics of the script: Hourly, UTC-aligned processing for calendar year 2025 OAuth2-authenticated access to OpenSky Trino with resilient re-authentication Conservative data-quality filtering (maximum inter-message gap ≤ 60 s) Great-circle distance computation using a spherical Earth approximation Approximate percentile computation for global tractability Explicit exclusion of anomaly thresholds, alerts, or inference Incremental daily CSV writes with persistent audit logging The script is intended as a methodological reference and reproducibility artifact, not an operational surveillance tool. Methodological Summary ADS-B state vectors are differenced on a per-aircraft basis. Kinematic quantities are derived only from successive reports separated by ≤ 60 seconds to ensure numerical stability. Horizontal displacement is computed using great-circle distance. Hourly distributions are summarized using approximate percentiles, means, and standard deviations. No rolling windows, persistence modeling, clustering, or hypothesis testing is performed in this release. Receiver Coverage Consideration ADS-B observations originate from a heterogeneous, crowdsourced global receiver network. Receiver density, placement, and temporal availability vary geographically and over time, influencing the volume and composition of observed state vectors. The metrics reported here characterize the internal kinematic self-consistency of observed ADS-B data only and do not attempt to correct for, normalize, or model receiver coverage effects. Interpretation Constraints This release must not be interpreted as evidence of: Aircraft anomalies or unsafe behavior Avionics malfunction Interference, spoofing, or cyber activity Air navigation service provider performance Safety incidents or operational degradation All reported variation reflects statistical properties of reported ADS-B data influenced by traffic mix, reporting rates, surveillance geometry, and observational conditions. Intended Use This dataset and code are intended for: System-level surveillance performance characterization Global or regional comparative analyses Temporal sensitivity baselining Methodological grounding for higher-resolution studies Reproducibility The included Python script fully specifies the analytical pipeline.Executing it against the OpenSky Trino database reproduces the dataset contained in this DOI, subject to OpenSky data retention and access policies. License Code: Creative Commons Attribution 4.0 International (CC BY 4.0). Data: Subject to OpenSky Network data usage terms Appropriate attribution is required for any reuse or derivative work.