Systematic study of universal-stage measurements of planar deformation features in shocked quartz: Implications for statistical significance and representation of results

Abstract

The presence of shocked quartz is one of the key lines of evidence for the impact origin of rocks. Crystallographic orientations of planar deformation feature (PDF) sets in shocked quartz have been used to constrain the peak shock pressure that these grains have experienced. So far no systematic and comparative studies of the various orientation measurement methods and their biases are available. Therefore, three shocked-quartz-bearing thin sections from a meta-greywacke clast in breccia, a biotite-gneiss, and a sandstone, respectively, were independently analyzed by three operators (two experienced and one inexperienced) using a four-axis universal-stage (U-stage), in order to evaluate the quality, precision, repeatability, and representativeness of U-stage measurements. Based on the indexing of PDF sets using a new version of the commonly used stereographic projection template, the study of 1751 PDF set orientations in 666 quartz grains in three different shocked rocks shows that differences in abundance and orientation of various PDF sets, as measured by the three separate operators, are rather limited. The precision of U-stage measurements depends mainly on the number of PDF sets investigated, as the ability level of the operator (experienced versus inexperienced) is only responsible for minor deviations in the number of unindexed planes. The frequency percent of dominant PDF planes may vary by up to 20 percentage points (pp) or 81% for a given crystallographic orientation when only 25 sets are measured. When 100 PDF sets are measured, however, this deviation in dominant orientations is reduced to about 7 pp or 28%. We recommend the use of a new stereographic projection template, which plots the pole positions of five additional, commonly occurring PDF orientations, as it can allow indexing of up to 12 pp more PDF planes; these are planes that would previously be considered unindexed and potentially regarded as errors of measurement. Our results suggest that by following a strict measurement procedure, the reproducibility of Ustage measurements is good and the results of different studies can be readily compared. However, it is critical that published PDF orientation histograms clearly define what type of frequency measurement is used, whether or not unindexed PDF sets are included in the frequency calculations, the numbers of grains and sets analyzed, and the relative proportions of each PDF set population that are combined in the histograms. This information appears to be essential for effectively comparing datasets from different studies.