A TEST OF STATISTICAL CORRECTIONS FOR THIN SECTION-BASED SIZE MEASUREMENTS OF CHONDRULES

The size distributions of these chondrules, the main components of ordinary chondrite meteorites, are an important, but still relatively unconstrained, source of information on early solar system processes. Because of the destructive nature of disagreggating chondrules in order to measure their sizes (usually maximum or minimum diameters, here we use the maximum), many chondrule size measurements have been made on thin sections. Thin section-based size measurements, however, yield “apparent” size distributions, rather than true size distributions, because thin sections are two dimensional samples of three dimensional objects, and the maximum diameter of a chondrule as measured in thin section is only the true diameter in the case where the thin section has equatorially sampled the chondrule. Statistical corrections have been developed to account for the sampling bias created by using thin sections for measurements and therefore convert the apparent distribution to true, 3D size distributions. However, the validity of these tests has to date been able to be assessed only by comparing corrected thin section data to disaggregation data for chondrules that may or not be from the same meteorite, but are not the same chondrules the thin section data is based on. We have acquired X-ray computed tomography (CT) data for a series of ordinary chondrites using the scanning facilities at the University of Texas at Austin's High-Resolution X-ray Computed Tomography Facility. The X-ray CT data provide the unique opportunity to create both 2D and 3D size distribution datasets for the same meteorite. X-ray CT data are stacks of x-ray attenuation images that combine to form 3D volumetric data. We have written a software routine that extracts 2D slices in different from the volumetric data, mimicking the act of creating a thin section, forming the basis for the 2D size distribution. This dataset is subjected to the same statistical corrections used with true thin section measurements and compared to a size distribution created by finding the maximum diameter of chondrules in all three dimensions. This allows an assessment of how well the statistical corrections for thin section-based size measurements convert “apparent” size distributions to true size distributions. We present the results of this assessment at the conference.