Inference of meteoroid characteristics using a genetic algorithm

A methodology is introduced to optimize and extend the inference of pre-entry size, density, strength, and mass of asteroids based on observed light curves. In this development study, a genetic algorithm (GA) approach is coupled with the fragment-cloud model (FCM) to efficiently evaluate entry and breakup for numerous potential asteroid property combinations and determine which case best matches the observed data. FCM produces energy deposition curves based on assumed pre-entry conditions, and the GA finds values for these inputs that minimize an objective function characterizing the difference between the FCM curve and a target curve. We present an overview of the GA approach, and then demonstrate its capability to infer pre-entry properties for three well-characterized events: Chelyabinsk, Lost City, and Benesov. In all cases, our initial mass and size estimates were within the range of published values.