The depositional history and evaluation of two late quaternary, diamondiferous pocket beaches, south-western Namibia

Milad, Micael George

Title: The depositional history and evaluation of two late quaternary, diamondiferous pocket beaches, south-western Namibia
Creator: Milad, Micael George
Subject: Uncatalogued
Date Issued: 2004-03
Date: 2004-03
Type: Academic theses
Type: Master's theses
Type: text
Identifier: http://hdl.handle.net/10962/420934
Identifier: vital:71795
Description: The two Late Quaternary, diamondiferous pocket beach deposits studied here are situated along a 10 km stretch of the storm-dominated, Atlantic coastline of the Sperrgebiet, south-western Namibia. The pocket beaches are approximately 130 km north of the Orange River mouth, which is widely accepted as a long-lived point source for diamonds sourced from the interior of southern Africa. A total of fourteen pocket beach deposits were recently evaluated in this area, but only two of these, namely Site 2 (to the south) and Site 3 (to the north), are considered here. The main diamondbearing horizons are beach gravels, which occur within, and form part of, the pocket beach sequences. The beach gravels are mostly blanketed by sand overburden, meaning that exposures available for study were limited, and much reliance was placed on borehole logging and observations of evaluation sample tailings. The main aims are to unravel the depositional history of the pocket beach sequences, identify the controls on diamond mineralisation in the beach gravels, and critically examine two different methods of estimating average diamond size for the deposits. In pursuit of these aims, sedimentological characteristics of the unconsolidated pocket beach deposits were recorded using small diameter drill holes, hydraulic grab bulk samples, trench exposures and surface outcrops. The surface geology, geomorphology and modern wave patterns were mapped using high-resolution, Airborne Laser Survey imagery coupled with extensive field checking. Three-dimensional geological modeling software was used to gain insight into the subsurface morphology of the deposits. Fossil shell samples were used to aid interpretation of ancient depositional environments and to date parts of the pocket beach sequences. Variations in diamond concentration and the size of diamonds were recorded using bulk samples, some of which were taken from a trench, but most of which were excavated using a hydraulic grab tool called the GB50. Finally, by using diamond size data from Site 3, sample data from diamondiferous beach gravels to the south of the study area and sample campaign simulations, two alternative methods of evaluating average diamond size in marine gravel deposits were appraised.The pocket beach sequences occur within north-south trending valleys of a major deflation basin and are separated from one another by rocky headlands. The ridge-and-valley topography of the deflation basin has resulted from differential erosion of Late Proterozoic basement rock units, alternating layers of which differ greatly in their resistance to the long-lived, local denudationalprocesses. On the basis of the stratigraphic information collected from the unconsolidated pocket beach valley fills, interpreted within the context of global, Late Pleistocene sea level records, the following depositional history is deduced : a) Deposition of sheetflood gravels by ephemeral streams, activated during a regressive phase. b) Transgression, culminating in the deposition of a gravel beach, representing a sea level highstand of +4 metres above mean sea level (mamsl) at between 120 000 and 130 000 BP. c)A regressive phase, resulting in deflation of former valley fills to the bedrock valley floor and accompanied by re-activation of ephemeral stream activity to form sheetflood deposits; this represents a protracted period of subaerial exposure of the +4 m gravel beach deposit. d) Deposition of a great volume of sediment in the valleys during the latter stages of the transgression from the Last Glacial Maximum (LGM). The sequence generated during this phase, which started at ca. 9 000 BP, contains : i) pan/coastal sabkha sediments, ii) shallow, sheltered bay sediments, iii) back-barrier lagoonal sediments, iv) a gravel beach deposit representing a sea level stillstand at -5 mamsl, laid down between 7 600 and 5 600 BP, v) another gravel beach deposit representing the well-known Middle Holocene sea level highstand at +2 to +3 mamsl, laid down at ca. 5 000 BP, and which terminated the transgression from the LGM. e) A minor regression to the current sea level, accompanied by progradation of the shoreline to its current position. This progradational marine unit consists almost entirely of sand and grit, reflecting the lack of gravel supply to this part of the coastline in the most recent past. f) Deposition of modern coastal dunes, which cap the pocket beach sequence and are the youngest sediments in the study area. Using trench and hydraulic grab evaluation sample results, in combination with analysis of wave patterns and field observations, the following local controls on the density distribution (ie. concentration) and size distribution of diamonds in the gravel beach deposits (+4, -5 and +2 to +3 mamsl stands) are recognised: a) Gravel beach depositional processes, which are responsible for clast sorting on the beach, have influenced the density and size distribution of diamonds. The infill zone, or beach toe, favours maximum diamond concentration while diamond size decreases from the imbricate zone (intertidal) to the infill zone (subtidal). b) Wave energy is identified as the dominant local control on diamond size distribution, but has also influenced diamond concentration to a limited degree. Larger diamonds are intimately associated with coarser beach gravels, both of which are a reflection of increased wave energy. Higher concentrations of diamonds are sometimes associated with zones of coarser gravel and therefore greater wave energy. c) The time of deposition of the host gravel beach is seen to be the dominant controlling factor with respect to diamond concentration. This is seen as evidence of significant temporal variation in the availability of diamonds in the littoral evironment. A significant reduction (20%) in average diamond size from Site 2 to Site 3, over a distance of only 6 km, is evident. The following were identified as reasons for this reduction in diamond size : a) Longshore sorting processes, of which the long-lived northerly littoral drift is a key part, are known to have played a role in the diminution of diamond size northwards from the Orange River mouth point source. However, it is believed that this can only partly account for the observed 20% reduction in diamond size. b) Input of sediment and smaller diamonds at Site 3, reworked out of an older, Eocene-aged marine succession in the hinterland, is recognised as a possible additional reason for the large reduction in diamond size from Site 2 to Site 3. It is also speculated that the large size of the pocket beach at Site 3, relative to Site 2, may have resulted in lower average wave energy at Site 3, with consequent reduced average diamond size. Diamond size in the beach gravels of Site 3, as well as in beach gravels elsewhere in the Sperrgebiet, is seen to be lognormally-distributed within geologically homogeneous zones. In theory, lognormal mean estimators represent the best method of estimating average diamond size in such cases, whereas the arithmetic mean estimator has the tendency to overestimate when large outlier values occur. Lognormal mean estimators have the added benefit of providing for the calculation of confidence limits, which are becoming increasingly more important as financial lending institutions insist on better quantification of the risk involved in resource estimates. Sample campaign simulations demonstrate, for the kinds of diamond size-frequency distributions typical of beach gravel deposits at Site 3, that there is no significant improvement in the accuracy of average diamond size estimates when lognormal mean estimators are used instead of the arithmetic mean estimator. This is because the variance (a ) of the diamond populations is low, and large outlier values are extremely unlikely to occur. However, simulation of a diamond population with high variance, drawn from a sample of beach gravels near the Orange River mouth, shows that lognormal estimators produce significantly more accurate results when a is large. Since individual diamond weights were not recorded during evaluation sampling of Site 3, numerical solution of lognormal estimators is not possible, and these would need to be solved using a less accurate graphical method. It is therefore recommended that individual diamond weights are recorded in future sampling campaigns, allowing for the use of lognormal mean estimators, and the calculation of confidence limits for average diamond size estimates.
Description: Thesis (MSc) -- Science, Geology, 2004
Format: computer
Format: online resource
Format: application/pdf
Format: 1 online resource (197 pages)
Format: pdf
Publisher: Rhodes University
Publisher: Faculty of Science, Geology
Language: English
Rights: Milad, Micael George
Rights: Use of this resource is governed by the terms and conditions of the Creative Commons "Attribution-NonCommercial-ShareAlike" License (http://creativecommons.org/licenses/by-nc-sa/2.0/)

Hits: 373
Visitors: 385
Downloads: 15

Collections

RU Department of Geology

		Thumbnail	File	Description	Size	Format
View Details Download			SOURCE1	MILAD-MSC-TR04-999.pdf	17 MB	Adobe Acrobat PDF	View Details Download