QUANTIFYING ROUNDNESS OF DETRITAL MINERALS BY IMAGE ANALYSIS: SEDIMENT TRNSPORT, SHAPE EFFECTS, AND PROVENANCE IMPLICATIONS

In this article we develop a mathematical method based on image analysis to measure the textural properties of heavy minerals in fluvial, colian-dune, and beach sediments throughout the 1800-km-long Orange cell of littoral sand transport, the longest documented so far on our planet. We analyzed the grain size and shape of 12,700 grains, including all major heavy-mineral species contained in 22 selected samples collected along the Atlantic coast, from Namibia to southern Angola. In this unique natural laboratory, where the Orange River represents a single dominant sediment source characterized by diagnostic compositional fingerprints, hyperarid climate ensures limited chemical alteration of even unstable ferromagnesian minerals. This case study led us to: a) monitor changes in grain size and roundness of various detrital minerals during ultralong transport in high-energy shallow-marine and colian environments; b) determine the relative durability of various detrital minerals to mechanical wear as a function of their mineralogical properties (e.g., hardness, cleavability); c) compare roundness of grains in sands of different facies to establish the relative efficiency of mechanical processes as a function of depositional environment and transporting medium (i.e., water versus air); d) compare roundness of the same detrital species in beach deposits enriched in heavy minerals to different degrees and evaluate whether rounded grains are preferentially entrained or left behind during high-energy storm events; e) test the use of textural properties to distinguish between coarser or angular grains of local provenance from smaller or rounded grains derived from distant sources. The rigorous definition of particle shape and the measurement of grain roundness is a necessary step to achieve a full understanding of sedimentary processes, evaluate mechanical effects during long-distance transport, challenge untested assumptions, and obtain useful complementary information on sediment provenance.