Orbit dissimilarity criteria in meteor showers: a comparative review

Abstract

In meteor science, the identification of meteor showers is a crucial and complex problem. The most common method is to perform a systematic search of a database of observed orbits using an orbit dissimilarity criterion (D-criterion) and an algorithm. D-criteria compare the result of an orbit dissimilarity function (D-function) and a threshold. These D-functions associate one value to two meteoroids orbits. If this value is lower than the threshold, the meteoroids' orbits are considered similar. Group of meteors are thus formed. However, not all D-criteria have been evaluated, and their high number makes it hard to know which should be prioritised. This paper presents a review of each D-function, the tests they passed, the threshold choice, and the algorithms they are used with. We show what methods are currently used in the search for meteor showers, presenting statistics based on papers justifying the existence of established meteor showers. We review D-functions from eight different papers. We describe how thresholds and clustering algorithms are usually chosen. We also analyse tests that were performed on D-criteria. We discover that most of those criteria were not properly tested, and that some have been criticised for their theoretical background. Thus, we recommend performing a post-search analysis of the groups found, both in a statistical sense (to make sure the groups formed could not have been formed randomly) and an orbital dynamics sense (to check whether the group could indeed come from a singular parent body), to present the findings as potential meteor showers.