Doctor of Philosophy
Department of Geography
Roberts, Richard G., Sediment budgets and quaternary history of the Magela Creek catchment, tropical Northern Australia, Doctor of Philosophy thesis, Department of Geography, University of Wollongong, 1991. http://ro.uow.edu.au/theses/1387
The Magela Creek catchment is situated in tropical northern Australia. This thesis examines the contemporary sediment transport regime of the sand-bed Magela Creek, the Holocene alluviation of its palaeochannel, the Quaternary history of extensive colluvial sand aprons developed along the flanks of the Arnhem Land plateau, and the chronology of isolated alluvial and lacustrine deposits on the plateau surface. Catchment sediment budgets are constructed at contemporary (AD 1950-1989), Holocene (0-7 kyr) and Pleistocene (7-20 kyr) timescales, and a sedimentation chronology for the sand aprons extended back to ∼240 kyr.
Contemporary sediment and solute transport rates in the Magela Creek were measured between 1986 and 1989. The measured bedload transport rates compare favourably with those computed by the van Rijn (1984c) model and the mean computed yield for 1980-1989 (1800 m3/yr) corresponds closely to the rate of infilling of Mudginberri Billabong determined for the same period from air photo and bathymetric surveys (1400-2100 m3/yr). Annual yields of washload and solutes are estimated using rating curves. Over the 1971-1989 period, the total terrigenous load transported by the Magela Creek (∼12,000 t/yr) consisted of approximately 30% bedload, 15% suspended sand, 45% washload and 10% solutes. The sand fraction is larger than that commonly reported for tropical rivers. It is derived almost entirely from the Arnhem Land plateau and corresponds to a plateau erosion rate of ∼5 m3/km2/yr. Washload and solutes are derived mostly from the lowlands and their yields equate with a lowland denudation rate of ∼28 m3/km2/yr, which lies at the low end of the range reported for other tropical savanna regions. The difference between the two rates suggests that the relative elevation of the plateau is increasing with time.
Sand transported along the Magela palaeochannel prior to ∼8 kyr was discharged into the deep-water estuary that existed at what is now Mudginberri Billabong. Infilling of the palaeochannel began at ∼8 kyr with the downstream progradation of a sand wedge. This sand probably was supplied by gullying of nearby sand aprons, associated with the return of enhanced monsoonal activity to the region during the post-glacial marine transgression. The palaeochannel subsequently accreted vertically, although the rate of infilling over the last 3 kyr was double that of the preceding 4 kyr.
Sand aprons, dated by thermoluminescence, began to develop at the foot of the Arnhem Land escarpment at 220-230 kyr and 100-120 kyr: these ages coincide with the start of the penultimate and last interglacials respectively. Since then, the aprons have accumulated at a fairly constant rate (30-70 rnm/kyr). The basal ages of the aprons and their volumetric rates of accumulation imply an escarpment retreat rate of 20-200 rnm/kyr, in contrast to a plateau lowering rate of ∼5 rnm/kyr (inferred from the rate of infilling an enclosed ephemeral lake in the catchment headwaters). The dominance of escarpment retreat over plateau lowering by a factor of 4-40 accords with the classic theories of parallel retreat of slopes rather than the downwasting of interfluves.