Rating Severity of Aircraft Separation Violations from Logic of Relative Collision Probability in Event Reenactment

Abstract

Rating the severity of losses of standard separation between aircraft in controlled airspace is a critical practice by FAA safety analysts in the measurement of aviation safety performance. The current practice in use by FAA safety analysts of using three categories is simple, but precludes making comparisons from data aggregated across the full range of separations. The proposed model rectifies this shortcoming by providing an approach to rating severity based on a relative collision probability estimate for an event that is replayed with quantitative separation variability based on published estimates. The model incorporates: (1) consideration of event reenactment of given horizontal and vertical separation with specified variability in each dimension, (2) probability density of co-location of the two aircraft to account for the overlap in aircraft variability for the event reenactment (formally the convolution or aggregation of co-location densities), and (3) size of each aircraft involved in horizontal and vertical dimensions. An appendix compares empirical measures of relative severity with those of the model. The form of such a model may be used to evaluate anticipated future changes in surveillance technology, separation standards, aircraft equipage, flight procedures, etc.