Morphological analysis of mineral grains from different sedimentary environments using automated static image analysis

The properties of sediment grain shape provide valuable information about the transport mechanisms in different sedimentary and geomorphological environments. With the emergence of new, high-resolution analytical techniques, it has become possible to quickly examine the grain shape properties of a large number of individual mineral grains. In this study, we used automated image analysis (Malvern Morphologi G3SE-ID) to investigate mineral particles of four sediment types from different depositional environments (sand sheet (1), floodplain (2) and fluvial channel deposits (3), Pleistocene infilling material of sand wedges (4), n = 20) in the Carpathian Basin (Central Europe). Our primary objective was to identify quantitative key variables that can help objectively distinguish certain geomorphological environments located in the Carpathian Basin. In our analysis and data processing (which included techniques such as hierarchical cluster analysis, Wilks' λ, Kruskal-Wallis, multivariate analysis of variance and principal component analysis) we focused on four variables related to grain shape: circularity (form), convexity (surface texture), solidity (roundness) and elongation (form). The form of sedimentary grains depends largely on the physical properties of their source area, while the roundness depends on the energy of the transport medium and the distant of transportation. Surface texture or convexity can change in a relatively short time in fluvial environment. The study revealed that distinguishing geomorphological environments can be achieved by analyzing the circularity, convexity and solidity parameters of the sediment grains. Based on the established grouping, the analyses carried out with hundred repetitions showed that HS circularity, convexity and solidity variables were the most effective attributes regarding Kruskal-Wallis test statistics it provided significant (p < 0.001) results between the analyzed sedimentary environments, while the elongation was not able to provide significant results between the grouped samples. Statistical analyses of the MANOVA test with hundredfold repetitions showed significant differences between the derived groups. Wilks' λ test statistics and PCA showed that convexity and HS circularity discriminate the groups. Separate analyses of aeolian and fluvial sediments have been carried out. The Kruskal-Wallis analysis of variance showed the significant differences considering all four variables, and differences were also significant in the case of MANOVA test. Wilks' λ test statistics and PCA showed that convexity, HS circularity and solidity discriminate the groups. According to our results, the circularity parameter can provide information about the transport distance, while the solidity parameter can indicate the transport energy. The convexity parameter can serve as an indicator of both transport distance, as well as post-depositional processes. Some infilling materials underwent multiple transport processes, including high energy aqueous, wind transport mechanisms, and post-depositional alteration process (frost weathering), while others originated from sand-sheet covered areas (active during Pleistocene glacials).The solidity parameter proved effective in separating sediments with similarly high convexity values (smooth surface), which were, in our case, from recent aeolian and fluvial environments. This result was due to the investigated fluvial sediments inherited their form and low level of roundness from their source area. Our research supports that aeolian transport is more effective in rounding the grains than the aqueous environment. Using automated static image analysis producing statistically stable results with hundreds of analyzed mineral grains provide useful indicators for paleogeographical reconstruction studies by investigating paleo and recent sediments.