The statistical modelling of the grain size distributions of cave sediments using log skew laplace distributions: Creswell Crags, near Sheffield, England

Abstract

Abstract This paper describes the statistical modelling of cave sediment size data using log skew Laplace distributions. The sediment samples were obtained from excavations in three archaeologically-important, palaeolithic deposits in the Creswell Crags Caves, which occur in the infrequently studied Lower Magnesian Limestone (Permian) bedrocks, east of Sheffield, England. Theoretical and practical justifications for employing this and other types of statistical distribution to model grain size data are briefly reviewed. Three independent and one summary descriptor based upon parameter measures derived by fitting log skew Laplace models to cave sediment grain size data are then used to explore the relationships between sediment location—specifically distance of the sample from the cave entrance(s) in three caves which differ in their geomorphic complexity. In the most complex cave—Robin Hood's Cave—broadly defined “groupings” or “sediment trajectories” were detectable in 3D displays of the parameter values for samples from individual “columns” of the sampled cave sediments. The relatively simple “sac” cave (Pin Hole Cave) appears to show relatively simple “zonal” trends in the descriptors, with variation in the outermost part of the chamber being primarily responsible for the trends observed. Zonal trends were also suggested, but less clearly, by some of the independent descriptive parameters in the more complex Church Hole Cave, but in this case it was the properties of the innermost samples that were responsible for much of the apparent variation detected. On the basis of this statistical modelling and the associated exploratory data analysis, it is hypothesized that textural variations in these three caves can be provisionally “disaggregated” into at least two complex elements. The first manifests itself in zonal variations which appear to relate, directly or indirectly, to the effects of increasing distance from these cave mouths (essentially akin to the ideas of Schmid, 1969 , Science in Archaeology, pp. 151–166. London: Thames and Hudson); the second discloses itself in more location-specific variations which themselves reflect the local sedimentary circumstances of that particular site within Robin Hood's Cave (similar to the ideas of Butzer, 1971 , Environment and Archaeology: An Ecological Approach to Prehistory. Chicago: Aldine).