Quantitative characterization of textures in mantle spinel peridotite xenoliths
Datasets usually provide raw data for analysis. This raw data often comes in spreadsheet form, but can be any collection of data, on which analysis can be performed.
A method for quantitative characterization of grain size in thin sections has been established for mantle spinel peridotite xenoliths, using optical scanning of large areas of thin sections, skeletonization of grain-section outlines and computerized measurement of individual grain-section areas. Measurements range from 218 for the coarsest example to more than 3000 in the finest grained. Variability of the samples has been examined in relation to size and number of grain-section areas measured by using multiple and orthogonal sections from several xenoliths. The results show a linear relationship of arithmetic mean against additive standard deviation, including data from coarse-grained protogranular, through porphyroclastic to the finer-grained equigranular examples. This suggests that peridotite textures form a continuous series rather than discrete groups, as suggested by qualitative (subjective) assessment. The observed distributions of grain-section areas have been explored in relation to their description and possible mechanistic origin. By direct measurement and comparison of cumulative number and area distribution curves, we show that qualitatively assessed ‘typical grain sizes’ are influenced by a small number of larger grain sections. Although the arithmetic mean and standard deviation provide a convenient method for comparison, in practice grain-section area distributions show marked positive skewness more consistent with log-normal or power-law functions. Linear log-probability curves also support the existence of a continuous series of peridotite textures, suggesting that the shallow lithospheric mantle has been subject to processes of comminution and/or grain growth dependent on the Law of Proportionate Effect.