The study is concerned with the parametric description of a selected size class of the gravel bedload, collected from a number of sample points, of the Tambo River in Eastern Victoria. Lithological composition of the samples is investigated and related to the lithological nature of the basin. Measurements of form, sphericity and roundness are made for samples of the two most abundant lithological types - rhyodacite and sandstone. Changes in mean values of these parameters between stations are related to operative processes, some of which can be characterized by the physical and lithological characteristics of the basin. kdetailed description is presented of the physical environment of the catchment. Consideration is given to sampling procedures and, the selection of suitable parameters to describe pebble morphometry. It can be shown that Wadell and Maximum Projection sphericity are partial expressions of form. R-mode factor analysis is used to compare the relative merits of four measures of roundness, on the basis of which the Kuenen and Kaiser measures.are used for further work. Subsequently the Kuenen roundness method is shown to be the most suitable for the purpose of this study. The quantitative relationships between the lithological nature of the basin and the composition of gravel samples are investigated using methods of analysis first introduced by Tricart in France in 1959 but apparently not previously applied in the English speaking world. The description of the mean form of a sample requires two parameters but values of chi-square can be used as a single parameter measuring the magnitude of change in form between two sampling stations. Changes in form, sphericity and roundness can be related to four basic processes: abrasion, shape sorting, dilution and breakage. In the case of form and sphericity it is shown that all four can be expressed in quantitative form. Changes in mean roundness of sandstone pebbles show a marked trend with respect to distance downstream, but no such tendency can be observed in the case of Thyodacite particles. Mean Kuenen roundness of samples can be replaced, with very little loss of information, by abrasion and breakage indices devised by the writer. Matrices of correlation coefficients are used to investigate relationships between variables both within and between samples and in the latter case also between the two dominant lithologies. They assist in reaching a better understanding of the nature of inter-relationships. Finally, a multiple correlation and regression model is employed to assess quantitatively the relative importance of different processes in producing change in form and sphericity between samples. In the case of rhyodacite highly significant multiple correlations are obtained and it is found that nearly all change can be explained in terms of processes and variations in mean sample size. In the case of sandstone the technique is less successful but even here changes in form at least have a significant multiple correlation with parameters representing processes and sample size.
History
Publication status
Unpublished
Rights statement
Copyright 1972 the Author - The University is continuing to endeavour to trace the copyright owner(s) and in the meantime this item has been reproduced here in good faith. We would be pleased to hear from the copyright owner(s). Thesis (M.Sc.) - University of Tasmania, 1973. Bibliography: l. 174-180