Comparing spatial grain-size trends inferred from textural parameters using percentile statistical parameters and those based on the log-hyperbolic method

摘要

The Folk&Ward (F&W) and the log-hyperbolic methods are applied to a small - and easy to overlook - number of typical sand sized grain-size distributions from the Danish Wadden Sea. The sand originates from the same source, and the pattern of change in the grain-size distributions is, therefore, exclusively linked to dynamic sorting. In general, the F&W parameters reflect the observed grain-size trends far better than the corresponding log-hyperbolic parameters. The log-hyperbolic "typical log grain size", v, is sensitive to changes in skewness and cannot replace mean grain size, Mz, in grain-size trend analysis. The four log-hyperbolic parameters describing dispersion δ, τ~(-1), k~(-1) and ζ are not able to give an unambiguous picture of spreading trends. The F&W sorting parameter, Sd, can therefore not be replaced by any of the log-hyperbolic parameters in grain-size trend analysis. In skewed grain-size distributions, there seems to be a general defect in the log-hyperbolic approximation of the best represented tail. This makes χ less sensitive to indicate the correct sign of slightly skewed grain-size distributions than Sk. However, when examining trends, the relative change between χ and Sk by and large seems to be the same. The log-hyperbolic peakedness parameter, ξ, is preferable to the F&W peakedness (kurtosis) parameter, Kg, because the latter is hypersensitive to small, insignificant variations in a grain-size distribution. By contrast, the concept of peakedness as defined by ξ, seems to relate more directly to sedimentary environmental conditions. In conclusion, the log-hyperbolic method has both advantages and disadvantages when compared with the F&W method. It is thus suggested that a sensible combination of the two methods could be advantageous to sediment trend analysis.