From 2D to 3D chondrule size data: Some empirical ground truths

In order to characterize the relation between apparent chondrule sizes (2D) and true chondrule sizes (3D), three ordinary chondrites of the H, L, and LL group have been analyzed. The diameters of a large number of chondrule cut faces in thin sections (2D; n = 2037) and of separated chondrules from the same meteorites (3D: n = 2061) have been measured. The obtained 2D/3D mean chondrule sizes (μm) for the H, L, and LL chondrite are 450/490, 500/610, and 690/830; the corresponding median values (μm) are 370/420, 450/530, and 580/730. The data show that there is a cutoff for small chondrule sizes in each sample. Possibly characteristic minimum sizes exist for the various groups, increasing in the (3D) sequence H (~90 μm) <L (~180 μm) <LL (~240 μm). No systematics were found for the maximum chondrule sizes. The investigated samples show very similar chondrule volume (mass) distributions relative to the mode (peak) of their size-frequency distributions. About 2.6–2.9% and 97.1–97.4% of the total chondrule volume (mass) is present in chondrule sizes smaller and larger than the mode, respectively. It was found that 2D sectioning consistently results in a shift of the true 3D size-frequency distributions toward smaller sizes. This effect leads to the underestimation of the values for (1) the true mean chondrule size by 8–18%, (2) the true chondrule median value by 12–21%, and (3) the true mode value of the size-frequency distributions by 12–17% (50 μm binning). This is the opposite of what popular 2D/3D correction models predict (e.g., Eisenhour 1996 ).