The non-trivial problem of meteorite pairing

Fairing is the procedure of identifying fragments of a single meteorite fall (that were separated during atmospheric passage or during terrestrial history) by establishing the similarity of two or more meteorite fragments. We argue that pairing is governed by two principles, that only a single mismatch of properties is required to refute a proposed pairing, and that virtually all pairings bear some degree of uncertainty. Using data distributions for modern falls, we take a probability approach to estimate degrees of certainty associated with proposed pairings, emphasizing the importance of unusual features. For new pairing criteria or new analytical additions to old criteria, the degree of variation within individual meteorites must be delineated and the degree of variation within meteorite classes must be quantified. Criteria for pairing can be divided into (1) parent body history indicators, (2) meteoroid space history indicators, and (3) terrestrial history indicators. Included in these categories are 11 specific criteria, including petrographic textures, mineralogy and mineral composition, terrestrial age estimates, cosmic-ray exposure ages, and natural thermoluminescence (TL) levels. Not all criteria are applicable to all meteorite types. About 2275 pairings suggested in the literature have been subjected to this analysis. Many literature pairings, especially those involving common meteorite types, bear large uncertainties due to lack of data.