The petrophysical classification of meteorites

Summary Petrophysical properties (susceptibility, intensity of the Natural Remanent Magnetisation (NRM) and bulk density) of 489 meteorite samples from 368 meteorites are discussed. The samples, obtained from Finnish meteorite collections, represent all chemical-petrological meteorite classes and their groups. This meteorite petrophysical database has many potential applications in the geophysical studies of extraterrestrial bodies (planets and their moons, asteroids, meteorite parent bodies, etc.). Here we use the database to classify meteorites rapidly and non-destructively by applying the petrophysical classification scheme developed by Kukkonen and Pesonen 10]. For example, the main classes and many groups form distinct clusters in petrophysical relation diagrams such as susceptibility vs. density or NRM vs. susceptibility. The petrophysical classification method was tested on 24 meteorites from Czechoslovak, 3 from Swedish and one from Australian collections. The chemical-mineralogical classifications of these meteorites were previously known. The subjective classification method was also compared with a mathematical cluster analysis. The subjective classification technique was successful in 64% to 93% of the cases whereas the mathematical analysis was successful in 57% to 82% of the cases. The failures can be attributed to (i) non-uniqueness problems (cluster plots overlap) and (ii) effects of porosity, self-demagnetisation, electrical conductivity and frequency on measured values, or to biasing caused by small sample size.