The Anomaly Severity Index (ASI): A Probabilistic Model for UAP Validation
In contemporary aviation safety analysis, a binary classification, where incidents are simply labeled as "identified" or "unexplained," is inadequate. The complexity of Unidentified Anomalous Phenomena (UAP) requires a method that can quantify degrees of uncertainty. A possible solution is the Anomaly Severity Index (ASI). This method offers this framework: a probabilistic model that transforms uncertainty from a defect into a measurable variable.
By implementing the ASI, data analysts can objectively assess the robustness of evidence ranging from brief visual contacts to high-quality, multi-sensory data streams within European airspace.
The Bayesian Approach: Probability in Context
The core of the ASI rests on Bayesian inference. This tests the probability of an anomalous report against the "background noise" of conventional air traffic and natural phenomena. Instead of looking for an absolute truth, this model calculates a joint Gaussian likelihood for variables such as altitude change, maneuver duration and speed.
When a pilot observes an object, human visual acuity (typically 1/60 degrees) modeled as a truncated Gaussian distribution. This allows one to mathematically combine the observation with sensor data. Environmental factors, such as an angular size of >0.15 degrees and light quality, serve as critical weighting factors that determine the initial confidence score before instrumental data is added.
The Five Dimensions of the ASI
The ASI is composed of five technical dimensions that together determine the severity of an anomaly:
Sensor quality and coupling
The highest validation requires coincidence between at least two independent sensor types (e.g., radar and FLIR/infrared). We aim for a statistical threshold of 5-sigma against the null hypothesis (instrumental noise or cosmic rays) before a report is considered definitively anomalous.
Kinematic Deviation
This measures the deviation from known aerodynamic laws. High scores are awarded for:
Gears: Ranging from 75g to >5000g (well above mechanical limits).
Speeds: Above Mach 1 without the associated sonic boom or heat emission.
Contextual Exclusion
A score only becomes significant when it has been proven that conventional objects have been excluded using databases such as FlightRadar24 (air traffic), N2YO (satellites), SondeHub (weather balloons) and Stellarium (astronomical objects).
Human Observation Integrity
Although pilots are trained observers, the ASI weighs their expertise probabilistically. Factors such as flight hours and the ability to make prosaic statements a priori to eliminate will enhance the score, provided it is consistent with the instrumental limits of the human eye.
Duration and Consistency
The duration is crucial for error elimination. Reports shorter than 10 seconds receive the lowest score; observations longer than 10 seconds significantly increase the ASI score.
Tier Classification: From Data to Policy
The ASI translates complex data into five Tiers:
Tier 0: Event fully explained.
Tier 1-2: Reports with limited data or short duration; large margin of uncertainty.
Tier 3: Significant anomalies with reliable sensor data that challenge conventional explanations.
Tier 4 (The Gold Standard): Long-term observations by multiple independent sensors, supported by physical evidence and peer review, form the basis for international data sharing within the EU.
Conclusion: A Model for the European Research Area
The implementation of the ASI within European airspace management may provide a methodological framework that strengthens flight safety without damaging the reputation of professionals. By treating reports as quantifiable safety insights within a Just Culture, The European Research Area (ERA) can replace stigma with strict scientific protocols. This is important for both national security and the integrity of academic science.
In depth: Read more about our standards and scientific principles in the dossier Scientific Research UAP.