Holocene development of multiple dune generations in the northeast Rub’ al-Khali,United Arab Emirates

Abstract

It has been hypothesised that in sand seas where multiple dune generations occur, each generation represents a distinct terrestrial response to contrasting palaeoatmospheric circulation conditions (Lancaster N, Kocurek G, Singhvi A, Pandey V, Deynoux M, Ghienne J-F et al. (2002) Late Pleistocene and Holocene dune activity and wind regimes in the western Sahara Desert of Mauritania. Geology 30: 991–994). However, reconstructing dunefield accumulation and preservation chronologies has often utilised a limited suite of samples because of the difficulties realised in accessing unconsolidated aeolian sands, which has limited the capacity to test this hypothesis. In the eastern United Arab Emirates, artificial excavation for quarrying and construction have generated a unique opportunity to examine and sample the internal structures of dune systems, and to generate data to test the hypothesis of dune generational development. This paper presents new data and chronologies from a multigenerational analysis of dune accumulation in the northeast dunefield of the Rub’ al-Khali. A complex developmental history during the Holocene is revealed, which does not fully conform to the notion of each dune generation forming in distinct palaeoclimatic phases, since the age ranges represented in the accumulation of secondary dunes is also found within a longer suite of accumulation ages from the underlying megaridges.

Keywords

Arabia Indian Ocean Monsoon late Quaternary megadune multiple dune generations OSL dating Rub al-Khali Shamaal

Introduction

In the last decade, the capacity to investigate and date the Late Pleistocene and Holocene development of continental dune systems has been greatly facilitated by the application of luminescence dating, especially optically stimulated (OSL) dating. Published studies from dunefields including those in Australia (Fitzsimmons et al., 2007), southern Africa (Telfer and Thomas, 2007), the USA (Clarke and Rendell, 2003; Forman et al., 2001) and other regions are revealing histories of dune development that are complex and which challenge earlier assumptions regarding the environmental conditions, and timing of dune accumulation (e.g. Glennie, 1998; Sarnthein, 1978).

Some dunefields are relatively simple and display a single system of dunes, often linear in form. Such systems, for example in the Kalahari and in parts of Australia, have in recent years been intensively investigated, and their analysis has revealed that the evolution of dunes in some cases has occurred through multiple accumulation events, separated in the sedimentary record by palaeosols (e.g. Fitzsimmons et al., 2007), while in others the dated record appears to suggest almost continuous accumulation through the Late Pleistocene and Holocene (e.g. Stone and Thomas, 2008).

Other dunefields however display multiple generations of overlapping dunes (e.g. Beveridge et al., 2006), and in these cases it has been hypothesised that each generation may represent a distinct dune formation phase, with different pattern orientations resulting from differences in palaeocirculations at the times of formation. Some dunefields in the Sahara, Arabia and North America (Arbogast, 1996) display these characteristics. In Mauritania, direct dating of sediments from dunes in each generation appear to confirm the hypothesis of a different wind directional regime being responsible for the development of each dune generation (Lancaster et al., 2002). In this paper we report analysis of another of these multigenerational dune systems, in the northeast Rub’ al-Khali, United Arab Emirates. Here, major linear dune ridges are overlain by smaller linear dune forms that are hypothesised to be distinctly younger than the major underlying mega ridges. The secondary dunes are the focus of this paper, which aims to establish the timing of their formation.

Study area

The Rub’ al-Khali in the Arabian Peninsula is one for the world’s largest sand seas, within which marked dune morphological variability occurs (Breed and Grow, 1979). This variability is systematic with zonal differences occurring in dune types as a function of sediment availability and the dominance of different sand transporting regimes. In the western United Arab Emirates (UAE), the transverse ridges of the al-Qafa region give way to the megabarchans of the Liwa Depression, which, at up to 160 m high with average crestal spacing of 2.6 km, are recognised as some of the largest in the world (Stokes and Bray, 2005). Further east, the dunefield is dominated by large linear ridges orientated approximately NE–SW (Figure 1) They are typically up to 70 m high with interdunal spacing up to 1 km and extend over tens of kilometres. Although it is widely believed these features were formed under southwesterly winds due to a northward incursion of the summer Indian Ocean Monsoon (IOM) system across continental Arabia (e.g. Glennie, 1998), today they are transverse to the prevailing atmospheric circulation regime which is dominated by northwesterly trades, locally referred to as shamaal winds. These are responsible for the development of an asymmetric lateral profile with steeper southeastern flanks and the superposition of a secondary longitudinal dune generation. Consequently, surface layers are active, and the aeolian system is dominated by reworking and sediment transport while conditions conducive to preservation are limited.

Figure 1.

Location of the western UAE and study sites. Trends of megaridges and major alluvial fans are also shown

A secondary generation of linear dunes is superimposed on the principal linear megaridges in the UAE. The secondary dunes are orientated approximately NW–SE to WNW–ESE, at right-angles to the underlying parent features, and longitudinal to the current prevailing wind direction (Figure 2). They are considerably smaller in scale than the major ridges, typically attaining heights not exceeding 20 m. Interdunal spacing ranges from tens of metres to several hundred metres. Minor variability in the orientation of the northwesterly winds has resulted in the secondary dunes assuming locally sinuous tracks with frequent coalescing of adjacent ridges at ‘Y’-junctions, extending from the interdunes between the primary linear megaridges on to their northwestern flanks. Local reworking of both these dune generations has resulted in the formation of additional minor forms resulting in a ‘fish-scale’ planiform (Stokes and Bray, 2005).

Figure 2.

Secondary dune ridges, orientated NW–SE/WSW–ESE overlying mega ridges orientated NE–SW. Location of site UAE07-6 is shown. Other features are man-made including a camel race track to the north of the E–W highway. Image from Google Earth

The chronological relationship between the accumulation and preservation of these secondary features and that of the underlying larger-scale linear ridges is of particular interest. It has been proposed that distinct but contiguous dune generations developed under distinct palaeoatmospheric (and therefore potentially alaeoenvironmental) conditions prior to their preservation (Lancaster et al., 2002) In the northeasternmost Rub’ al-Khali of the eastern UAE, the existence of these multiple dune generations, in addition to wadi, alluvial fan and lacustrine sedimentary archives, suggests a complex Quaternary history of palaeoenvironmental change due to the region’s location at the interface between northwesterly and monsoonal wind systems.

Assessing the nature, magnitude and timing of Quaternary palaeoenvironmental change in desert regions has previously been complicated by three key factors. First, the lack of a robust dating protocol for sediments that typically contain little organic material has restricted the accuracy of age determinations. Second, access to deep dune sediments for systematic chronological investigations, in order to obtain more than the most recent history of dune activity available from upper sediments, requires some form of drilling equipment. This needs to be portable if a regional rather than a localised chronology is to be established. Third, where sediment ages have been determined from limited sampling, there have been issues concerning the interpretation and utility of the resulting data. The validity of extrapolating a record of aeolian activity or dune formation has been questioned, since such ages may primarily represent the timing of conditions favourable to dune preservation (see Stone and Thomas, 2008; Telfer et al., 2010; and Thomas and Burrough, 2010, for discussion of issues surrounding dune record interpretion). This may not be representative of periods of dune activity as a whole, either because dynamic activity itself limited net accumulation, or because older deposits have been eroded by later periods of dune activation.

Recent advances in optically stimulated luminescence (OSL) dating protocols, however, have allowed the ages of aeolian dune sand samples to be determined with greater confidence. Portable hydraulic drills for coring through dune sections mean that samples can be extracted from greater depths and systematic sampling strategies may be deployed, affording more comprehensive, regional coverage. In the eastern UAE, an additional advantage is that an extensive infrastructure development programme has resulted in the exposure of many sections through dune cores which are available for sampling, allowing the context of deposits to be better assessed that in situations where coring alone provides access to dune bodies.

In this study, OSL dating is used to develop chronologies of secondary dune formation in the northeast Rub’ al-Khali, in particular, in relation to the corresponding chronologies determined for the underlying megaridges (Atkinson et al., 2011). Such a chronological framework will provide an appropriate context for palaeoenvironmental inferences to be discussed.

Southeast Arabia regional setting

The Rub’ al-Khali extends over an area of approximately 600 000 km² (Goudie et al., 2000), extending from the Yemeni highlands in the southwest to the Oman mountains and Arabian Gulf coast in the northeast. Emanating from the southwestern flanks of the Oman mountains are a series of substantial alluvial fans that have coalesced to form a nearly continuous belt of gravels fringing, and progressively overlain by, the northeastern margin of the dunefield. The shallow Arabian Gulf basin to the north with its asymmetric bathymetry and shallow southern coastal zone has been a significant sediment source at times of sea level low stands, such as during the last glacial maximum, when regression eastwards towards the Straits of Hormuz resulted in the complete exposure of the Gulf’s floor (Lambeck, 1996; Sarnthein, 1972). Combined with the dominance of northwesterly atmospheric circulation (Glennie, 1998; Teller et al., 2000), this resulted in dune construction along the coastal zone of the dunefield. Subsequent transgression and flooding of the Arabian Gulf basin removed this sediment source, with coastal dune deflation resulting in the formation of a network of barrier islands and lagoons with sediment being transported southeastwards.

The sedimentology and provenance of the Rub’ al-Khali dune sands has been widely studied (Abu-Zeid et al., 2001; Besler, 1982; El-Sayed, 1999; Garzanti et al., 2003; White et al., 2001) and are not repeated in this paper. Three key sand source regions have been identified: mature, red Fe-stained quartz grains originate from the Arabian continental interior; carbonates are contributed from the Arabian Gulf basin and the limestones of the R’as al-Jebal mountains; and mafic and ultramafic grains are derived from the ophiolites of the Oman mountains. Today, the local wind regime is broadly bimodal, with 52% of sand-transporting winds blowing from the west and northwest (the ‘shamaal’ wind) and approximately 28% blowing from the southeast (Goudie, 2002). The region is hyperarid with mean annual precipitation of 120 mm recorded at Ras al-Khaimah airport, however, this is highly variable. The significance of the orographic contribution to local rainfall is supported by a precipitation gradient decreasing westwards towards Dubai (approximately 80 mm/yr).

Sampling and experimental procedures

Sampling strategies and sample retrieval

Owing to the largely unconsolidated nature of aeolian landforms and the difficulty in creating continuous stratigraphic sections through dunes, early studies typically relied on shallow subsurface sampling from hand-excavated pits (e.g. Stokes et al., 1997), resulting in only the most recent record of aeolian activity being retrieved. Portable drilling apparatus has overcome this issue and allowed sampling at regular intervals to greater depths within a dune body (e.g. Telfer and Thomas, 2007). This method nonetheless creates a number of issues that require care in the field and in subsequent data interpretation. First, repeated insertion and removal of drilling rods could contaminate sediment being sampled at depth with shallower material falling down the borehole: considerable care is therefore required in the field. Second, it is common for water to be poured down the borehole to encourage the sediment to collect in the sampling head. This prevents in situ water content from being determined later in the laboratory with implications for subsequent dose rate determinations. Third, and arguably most important, is that while samples may be retrieved from depths of over 20 m (especially when drill rigs are used, e.g. Bristow et al., 2006) and over 150m in extreme cases (Preusser et al., 2002), access to the detailed down-profile stratigraphic information that would allow samples to be placed into depositional context is not available.

In the eastern United Arab Emirates, the large artificial exposures in sand quarries yielded continuous stratigraphic sections up to 80 m high. These sites, albeit emphemeral (most had disappeared by 2010, because quarrying ceased) has several advantages. The depth into dune cores that may be sampled is often greater than that afforded by portable drilling equipment and the process of extracting samples is more rapid and less likely to result in contamination. Most significantly, the internal dune stratigraphy is viewable and a sampling strategy tailored to each site is possible, accounting for the presence of features such as bounding surfaces, textural changes and sedimentary horizons of particular interest.

Sand quarries were present during the field period in both underlying mega ridges and overlying secondary dunes. This paper reports the investigations into exposures in five quarries in the latter, plus data from one drilled site where exposures were not present. Though the sites are thus dependent on the distribution of quarrying activities, this situation presented a unique (and ephemeral, since many quarries has been abandoned and graded by 2010) opportunity to view and sample dune interiors. Thus the paper does not present data from multiple blindly sampled sites through the whole region, but targets sites where detailed sampling could occur in the knowledge of sedimentary contexts.

OSL dating

The aim of the preparation procedure prior to OSL measurement is to extract pure quartz of a defined size fraction from the bulk sediment sample. The first stage involves the removal of the carbonate fraction by immersion in 10% concentration HCl until reaction completion. Following rinsing, H₂O₂ is added to remove any organic content. After sieving, the 90–150 µm fraction is suspended in sodium polytungstate (ρ = 2.72 g/cm³) to allow heavy minerals to be separated out. Following a 40 min etch in 40% HF and a final sieve to remove any fine grains generated during etching, the dried, refined quartz is mounted on 10 mm diameter aluminium discs using a silicone adhesive spray.

OSL measurements were conducted at the Oxford Luminescence Dating Laboratory on TL/OSL-DA-15 Risø readers equipped with 90Sr/90Y beta sources for sample irradiation with optical stimulation by blue LED arrays. To determine the efficacy of the preparation protocol described above at removing feldspar from the bulk samples, an infrared stimulation step was included. The equivalent dose, D _e, for up to 24 aliquots of each sample was determined by using the Single Aliquot Regeneration (SAR) protocol of Murray and Wintle (2006). The initial preheat (PH1) was selected after evaluating the results from preheat plateau tests, the value of PH1 being typically 240°C or 260°C and was administered for 10 s. A cutheat (PH2) of 220°C for 10 s was used. OSL measurements were made at 125°C for 80 s, with an optical bleach at 280°C for 80 s following administration of the test dose. To minimise cross-talk, 24 discs were mounted at alternate positions on a 48-position carousel (Bray et al., 2002). From the D_e values obtained, the central age model of Galbraith et al. (1999) was used in age calculations.

To determine dose rates, NERC ICP-MS and ICP-AES facilities at Royal Holloway – University of London and Kingston University were used to measure U, Th and K concentrations. Accounting for the moisture content value of the bulk sediment samples and variation through time since deposition in dose rate calculations is inherently difficult since it is rarely possible to know with any certainty the changing water content history of a sediment since deposition. Given the evidence for widespread aridity in the region, and the likely limited variation in water content of sands in a raised dune body, ‘as-found’ water contents have been used in dose rate calculations. The effect of doubling these values was found to change dose rates and consequent age determinations by <5%, and for most samples, the change was <2%. The variation in ‘as-found’ water contents is minimal with the water/wet sediment values averaging 0.1 or 0.2 for all sections and the mean value for all samples being 0.1. Because the samples originated from similar geomorphological settings this value has been used in dose rate calculations for UAE07-6 samples that were retrieved by coring.

Zander et al. (2007) investigated samples taken from sediments around Dubai, UAE, that contained a carbonate fraction of up to 83%. They point out that it is not feasible to establish the geochemical conditions of the sediment throughout its burial history. However, they showed that by using an appropriate model, it is possible to obtain OSL ages that conform to independent age controls despite uranium-series disequilibrium affecting dose rate calculations derived from ICP analyses.

The samples in this study area have a carbonate component typically around 40%. Three samples distributed north–south were examined in detail in terms of potential disequilibrium effects. The concentration of ²²⁶Ra was determined by measuring activity (Bq/kg) and establishing the corresponding activity of ²³⁸U assuming secular equilibrium. This was compared with the ²³⁸U concentration obtained from ICP-MS data and the effect on subsequent dose rate and age determinations analysed. It can be seen from Table 1 that the difference in dose rates calculated using ²³⁸U concentrations derived from ICP analysis and from ²²⁶Ra measurement does not exceed 9% while the difference in age calculations does not exceed 8%. As a result, a 10% random error component has been incorporated into final age calculations.

Table 1.

Uranium series disequilibrium analysis

Sample	UAE06–2–8	UAE07–7–5	UAE07–9–9
U conc - ²²⁶Ra (ppm)	1.24	0.89	1.18
U conc - ICP (ppm)	0.94	1.03	0.71
ΔU conc (ppm)	−0.3	0.14	−0.47
ΔU conc (%)	−23.97	16.25	−39.9
D′-²²⁶Ra (Gy/ka)	1.26	1.11	1.35
D′-ICP (Gy/ka)	1.19	1.14	1.24
ΔD′ (Gy.ka-1)	−0.07	0.03	−0.11
ΔD′ (%)	−5.56	2.7	−8.15
Age-²²⁶Ra (ka)	4.88	12.63	4.02
Age-ICP (ka)	5.17	12.26	4.36
ΔAge/ka	0.29	−0.37	0.34

Radiocarbon dating

¹⁴C samples were collected from three of the quarry sites (Table 2). The material submitted for dating comprised marine shell (Terebralia palustris and Pinctada radiata) or charcoal. The ¹⁴C samples were dated at the Oxford University Radiocarbon Accelerator. The ages were calibrated using CALIB OxCal 4.05 and for shells corrected for the Arabian Gulf Marine reservoir effect using the ΔR value of 180±53 (Southon et al., 2002) using the Marine04 curve (Hughen et al., 2004).

Table 2.

Radiocarbon dates for sites discussed in this study

Site	Depth (m)	¹⁴C age	1σ	Material	Cal. age BP (1σ)	Lab code
Umm al-Qawain airstrip	0.2	5909	36	shell	6235–6063	OxA–16858
al-Jazirat al-Hamra	0.6	342	26	charcoal	460–319	OxA–16863
al-Jazirat al-Hamra	1.75	4870	35	charcoal	4993–4883	OxA–16862
al-Jazirat al-Hamra	2.47	5220	36	seashell	5457–5309	OxA–16861
Ras al-Khaimah South	1.3	732	27	charcoal	685–666	OxA–16864
Ras al-Khaimah South	1.84	4449	36	charcoal	5271–4973	OxA–16915
Ras al-Khaimah South	2.35	5824	34	shell	6134–5965	OxA–16865

Results

At all the sample sites, described below, there was a distinct absence of evidence of depositional hiatuses/unconformities. In some sections thin horizons of shells, representing midden deposits, were present, with the upper part of UAE6-10 being marked by a distinctive thick midden unit. At all sites where structures could be observed, the dune deposits possessed thin laminations, consistent with aeolian deposition. Figure 3 shows sections from all sampled sites; age data are presented in Tables 2 and 3.

Figure 3.

Stratigraphic columns and OSL ages for the sampled sites referred to in this study

Table 3.

OSL age data for sites analysed in this study

Sample	Depth	K	K error	Th	Th error	U	U error	D′	D′ error	n	De	De error	Age	Age error +/− 1 sigma
	(m)	(%)	(%)	(ppm)	(ppm)	(ppm)	(ppm)	(Gy.ka-1)	(Gy.ka-1)		(Gy)	(Gy)	(ka)	(ka)
10/6/2001	0.3	0.68	0.005	1.72	0.08	1.17	0.01	1.24	0.04	11	16.56	11.68	9.44	0.5
10/6/2002	1	0.67	0.013	1.88	0.04	1.4	0.03	1.27	0.04	14	15.63	11.68	9.19	0.39
10/6/2003	2	0.67	0.001	1.64	0.01	1.31	0.03	1.22	0.04	12	14.78	11.76	9.68	0.47
10/6/2005	4	0.67	0.006	1.47	0.02	1.18	0.05	1.12	0.04	16	14.76	11.56	10.36	0.49
10/6/2006	5	0.65	0.004	1.83	0.04	1.24	0.03	1.11	0.04	14	15.41	16.54	14.85	0.64
10/6/2007	6	0.61	0.003	1.41	0.03	1.06	0.05	1	0.04	14	13.9	15.89	15.94	0.68
10/6/2008	6.7	0.72	0.002	1.66	0.06	1.23	0.03	1.16	0.04	14	12.11	17.23	14.81	0.81
06-13-1	1	0.76	0.002	1.46	0.01	1.21	0.04	1.31	0.05	14	13.46	0.5	10.31	0.54
06-13-2	2	0.63	0.013	1.76	0.03	1.22	0.02	1.18	0.04	16	11.47	0.5	9.71	0.55
06-13-3	3	0.6	0.006	1.41	0.05	1.03	0.02	1.06	0.04	15	11.83	0.55	11.12	0.66
06-13-4	4	0.69	0.003	1.76	0.01	1.23	0.04	1.21	0.04	10	13.13	0.31	10.87	0.46
06-13-5	5	0.72	0.003	1.43	0.03	1.05	0.02	1.16	0.04	14	12.92	0.33	11.17	0.5
06-13-6	6	0.76	0.005	1.69	0.02	1.13	0.02	1.22	0.04	13	14.78	0.4	12.09	0.55
06-13-7	6.7	0.71	0.004	1.84	0.04	1.21	0.02	1.18	0.04	16	17.04	0.66	14.39	0.75
06-14-1	1.5	0.66	0.005	1.64	0.02	1.07	0.04	1.2	0.04	10	5.4	0.2	4.52	0.23
06-14-2	2.6	0.7	0.004	1.8	0.04	1.16	0.02	1.18	0.04	15	12.74	0.37	10.78	0.49
06-15-1	1	0.91	0.01	1.39	0.03	0.78	0.02	1.35	0.06	13	0.31	0.03	0.27	0.02
06-15-2	2.5	0.85	0.005	1.42	0	0.89	0.02	1.31	0.05	14	3.56	0.39	2.72	0.32
06-15-3	3.5	0.75	0.004	1.49	0.02	0.99	0.02	1.21	0.04	12	6.94	0.2	2.95	0.34
06-15-4	4	0.86	0.004	1.84	0.05	1	0.04	1.32	0.05	15	10.24	0.38	7.73	0.41
06-15-5	4.7	0.83	0.01	1.48	0.04	0.97	0.04	1.25	0.05	16	13.48	0.11	10.77	0.43
06-18-2	1.5	0.87	0.002	1.27	0.03	0.83	0.04	1.3	0.05	14	1.44	0.15	1.11	0.12
06-18-3	2.5	0.84	0.001	1.3	0.02	0.96	0.02	1.28	0.05	14	12.07	0.33	9.41	0.44
06-18-4	3.5	0.8	0.006	1.72	0.01	1.1	0.03	1.29	0.05	16	13.2	0.23	10.23	0.41
06-18-6	5.5	0.63	0.001	1.06	0.02	0.85	0.01	0.99	0.04	16	13.52	0.25	13.64	0.56
06-18-7	6.5	0.87	0.001	1.23	0.02	0.86	0.03	1.23	0.05	15	14.9	0.45	12.11	0.6
06-18-8	7.25	0.84	0.006	1.32	0.02	0.89	0.02	1.21	0.05	20	13.46	0.43	11.16	0.56
6/7/2003	2	0.77	0.009	1.77	0.03	0.97	0	1.19	0.04	15	0.65	0.05	0.56	0.05
6/7/2004	2.5	0.71	0.004	1.79	0.04	1.1	0.03	1.1	0.04	13	1.06	0.08	0.96	0.08
6/7/2005	3	0.84	0.005	1.28	0.01	0.78	0	1.28	0.05	15	2.34	0.14	2.07	0.14
6/7/2006	3.5	0.76	0	1.57	0.04	1	0	1.11	0.04	12	2.69	0.18	2.66	0.22
6/7/2007	4	0.68	0.003	1.83	0.02	1.27	0.02	1.15	0.04	15	6.4	0.36	6.02	0.43
6/7/2008	4.5	0.83	0.004	1.32	0	1.03	0	1.14	0.05	16	6.11	0.25	5.25	0.28
6/7/2009	5	0.78	0.001	1.2	0.2	1.06	0.01	1.08	0.04	15	8.49	0.56	7.32	0.88
6/7/2010	5.5	0.87	0.001	1.45	0.01	0.77	0.03	1.14	0.05	15	30.44	2.05	27	2.07
6/7/2011	6	0.67	0.003	0.32	0.01	0.66	0.01	0.84	0.04	15	34.75	1.03	26.65	1.31
9/7/2001	0.5	1	0.001	0.4	0.04	0.6	0	1.32	0.07	16	1.76	0.04	1.33	0.08
9/7/2002	2	0.81	0.002	0.91	0.02	0.02	0.01	1.01	0.05	16	3.42	0.1	3.4	0.18
9/7/2003	4.5	0.84	0	0.67	0.01	0.73	0.02	1.13	0.05	16	3.33	0.1	2.94	0.15
9/7/2004	6.5	0.97	0.007	0.58	0.03	0.71	0.01	1.24	0.05	16	5.07	0.14	4.09	0.21
9/7/2005	8.5	0.88	0.002	0.49	0.03	0.74	0.02	1.14	0.05	16	4.47	0.11	3.93	0.2
9/7/2006	10.5	0.83	0.001	0.59	0.02	0.78	0	1.1	0.05	16	4.37	0.1	3.99	0.19
9/7/2007	11	0.84	0.003	1.22	0	0.68	0	1.1	0.05	16	4.46	0.1	4.03	0.2
9/7/2008	11.5	0.83	0.003	0.6	0.02	0.74	0.01	1.06	0.04	16	4.6	0.11	4.33	0.21
9/7/2009	13.5	0.99	0.008	1.43	0	0.71	0	1.24	0.05	16	5.42	0.08	4.36	0.2
07-9-9A	14.8	0.73	0.003	1.35	0.05	0.86	0.02	1.03	0.04	16	6.42	0.19	6.22	0.3
9/7/2010	15.3	0.56	0.001	1.35	0.03	0.68	0.03	0.82	0.03	16	5.86	0.3	7.14	0.46

Umm al-Qawain airstrip: UAE06-10 (25°34.9411′N, 55°39.641′E)

Section UAE06-10 is located near the border between the emirates of Ras al-Khaimah and Umm al-Qawain, less than 1 km from the Arabian Gulf coast. Shallow excavation into the coastal parabolic dunes yielded a 7 m section with the base obscured by recent slumping, and consisting of two units. The overlying unit is 4.8 m thick, comprising very fine to fine pale-grey/buff sand with little internal structure visible. It is capped by a poorly cemented shell midden horizon of variable thickness, typically 0.2–0.6 m thick, dominated by various bivalve and Terebralia palustris bioclasts (Figure 4). The underlying unit is darker and displays shallow inclined thick laminae dipping north to north-northwest. There are no discontinuities in the section.

Figure 4.

Shell midden layer capping section UAE06-10

The ages of seven samples were determined ranging from immediately below the shell midden layer at 0.3 m to 6.7 m. The shell midden was dated to 6.15±0.07 ka cal. BP (6235–6063 cal. BP 1σ, OxA-16858). The upper unit was preserved between approximately 10.4 and 9.2 ka with the lower sequence preserved between approximately 15.9 and 14.9 ka, suggesting rates of net accumulation of 3.16 m/ka and 1.50 m/ka, respectively. In general, both palaeodose and dose rate show a decreasing trend with depth.

al-Jazirat al-Hamra: UAE06-13, UAE06-14 (25°41.838′N, 55°48.313′E)

Sections UAE06-13 and UAE06-14 are located in an artificially excavated pit approximately 17 km southwest of Ras al-Khaimah city, near the village of al-Jazirat al-Hamra. Samples were collected from a 6.7 m section on the western sides of the site. The excavation surface is highly undulating and, in places, the uppermost 0.2–0.3 m contains abundant Terebralia palustris and bivalve shell fragments, suggesting a locally preserved capping shell midden. Further shell-rich horizons and lenses, and charcoal layers, are found within the upper unit which varies in thickness from approximately 1.0 to 3.0 m around the excavation. Within the western section, two such shelly horizons dip shallowly north with one outcropping at the surface. The repeated occurrence of these features suggests the site has a complex history of human occupation. The sedimentology shares many of the characteristics of the upper unit at site UAE06-10, comprising very fine to fine pale buff-coloured homogenous sand with little clear internal structure visible.

Section UAE06-14 is located approximately 50 m south of UAE06-13, within the same excavation, where Terebralia palustris samples from three shell-rich horizons were retrieved at 0.6, 1.75 m and 2.47 m. These were dated to 0.39±0.07 (460–319 cal. BP 1σ, OxA-16863), 4.94±0.06 (4993–4883 cal. BP 1σ, OxA-16862) and 5.38±0.07 ka cal. BP (5457–5309 cal. BP 1σ, OxA-16861), respectively. OSL samples were also taken at 1.7 m and 2.75 m. Age determinations for seven samples were made from UAE06-13, revealing that conditions favouring preservation of the material existed between approximately 9.7 and 14.4 ka.

al-Dhaith: UAE06-15 (25°43.051′N, 55°54.177′E)

Five samples were taken from a 5 m section through the upper section of a secondary dune near the village of al-Dhaith, c. 11 km southwest of Ras al-Khaimah city. The secondary dunes in this location are minor features, typically attaining crestal heights not exceeding 7–10 m (Figure 5). They feature greater vegetation cover than comparable dunes to the southwest, possibly because of thinning sand cover overlying the gravels of the coalesced outwash fans emanating from the mountain front to the east of Ras al-Khaimah city. The section exposed here consists of what appears to be a thick, structureless sandy layer (Figure 6). However, the material noticeably pales down-section so it is suggested that a darker layer overlies a paler unit with an indistinct gradational contact between approximately 3.7 and 4.0 m down-section. It is noticeable that the carbonate content of samples increase down-section from 46.5% at 1.0 m to 51.8% at 4.0 m, with sediments at 15-3 and 15-4 being finer than those above. This variability may explain the graded nature of the contact. This interpretation is supported by subsequent age determinations which show an abrupt increase in age down-section around the contact from 2.95 ± 0.34 ka at 3.5 m to 7.73 ± 0.41 ka at 4.0 m and a further sample dating at 10.77 ± 0.43 ka at 4.7 m.

Figure 5.

Vegetated secondary dunes south of Ras al-Khaimah near site UAE06-15

Figure 6.

Section at site UAE06-15 – al-Dhaith

Ras al-Khaimah South: UAE06-18 (25°43.551′N, 55°52.822′E)

This site is an 8 m section on the north side of an artificially excavated pit in a secondary dune on the southern edge of Ras al-Khaimah city from which six samples were collected. The upper 2 m consist of pale sands underlain by a thin layer of charcoal and shell fragments that marks the contact with an underlying darker red–buff sandy unit that makes up the remainder of the exposed section. Throughout the profile, there is evidence of inclined laminae dipping eastsoutheast to southeast, while localised midden deposts are also found in the upper sections (Figure 7). Three samples associated with midden contexts were dated via ¹⁴C. At 1.3 m charcoal was dated to 0.68±0.01 ka cal. BP (685–666 cal. BP 1σ, OxA-16864), at 1.84 m charcoal yielded an age of 5.12±0.15 ka cal. BP (5271–4973 cal. BP 1σ, OxA-16915) and shell at 2.35 m to 5.82±0.08 ka cal. BP (6134–5965 cal. BP 1σ, OxA-16865).

Figure 7.

Section at site UAE06-18 – Ras al-Khaimah south. Note the midden deposit at the same level as the figure. The pit in which UAE06-18-7 and 18-8 were sampled had not been deepened at the time of the photograph being taken

Six samples from this section have generated OSL ages, revealing extensive preservation during the Late Pleistocene–early Holocene transition. Although measured dose rate appears to decline down section from 1.30 Gy/ka at 1.5 m to 1.21 Gy/ka at 7.25 m, the apparently anomalous dose rate value of 0.99 Gy/ka measured at sample UAE06-18-6 may explain the age–depth inversion observed from 5.5 m. A total of 2.75 m of sediment was preserved between 9.41 ± 0.44 ka and 12.11 ± 0.60 ka, suggesting a net accumulation rate of approximately 1 m/ka.

Tawi Asmar: UAE07-6 (25°25.347′N, 55°45.787′E)

Site UAE07-6 is situated approximately 16 km inland from the Arabian Gulf coast near Wadi Lamhah (sometimes called Wadi Dhaid). Aerial imagery clearly shows the northwest–southeast secondary dune generation superimposed over the underlying primary ridges that have been truncated by episodic outflow through the wadi (Figure 2). Here, the secondary dunes form relatively minor topographic features, typically attaining heights of <10 m and frequently coalescing at ‘Y-junctions’. Interdunal spacing typically ranges between 40 and 80 m. The dune sampled near Tawi Asmar is approximately 5–6 m high and is situated on the rising northwestern flank of one of the primary dunes. Eleven samples were retrieved by coring from the dune crest, nine of which were dated (the uppermost two being <0.56 ± 0.05 ka). Preservation of aeolian accumulations at this site has occurred throughout the Holocene from 7.32 ± 0.88 ka to 0.56 ± 0.05 ka, with samples from the uppermost 2 m generating D_e values <0, suggesting that this unit remains active today. Between 5.0 m and 5.5 m, an abrupt increase in sample age from 7.32 ± 0.88 ka to 27.00 ± 2.07 ka is observed. This is commented on below in the Discussion.

al-Ain: UAE07-9 (24°21′4.33"N, 55°46′23.69"E)

UAE07-9 is an approximately 20 m high section taken from a quarry in the core of a megadune ridge situated approximately 3 km west of the Dubai–al-Ain highway on the northern edge of al-Ain city. To the south and east, the dunefield is truncated by a westward expansion of the coalesced alluvial gravel fans extending from the mountain front, and underlying the site are multiple gravel wadi terraces. Dune width and interdunal spacing vary considerably, ranging between 700 and 1500 m. The profile at UAE07-9 is characterised by soft red-yellow sand with little internal structure. Eleven samples from this profile were collected.

Section UAE07-9 dates exclusively from the Holocene with the oldest sample at 15.3 m below surface having an age of 7.1 ± 0.5 ka. In approximately 0.27 ka, between 4.4 ± 0.2 ka and 4.1 ± 0.2 ka, 7 m of sediment was preserved at a net accumulation rate in the region of 26 m/ka. This sequence is underlain by a marker horizon containing gravel and bioclasts, below which an earlier phase of accumulation is identified by two samples dating from 7.1 ± 0.5 ka to 6.2 ± 0.3 ka.

Discussion

Earlier work (e.g. Glennie, 1998) has established that during the Late Pleistocene and Holocene, conditions favouring development and preservation of aeolian activity persisted across southeast Arabia, while a range of studies have since produced dune accumulation ages from individual sites in the region (Figure 9). In the northeastern sector of the Rub’ al-Khali, our data show (Table 3) that the periods 16–9 ka and 7–2 ka were significant for secondary dune accumulation and preservation. There are some site-to-site differences in recorded preservation of aeolian accumulation. This is particularly noticeable at sections 06-13 and 06-14, which are located at one site and laterally displaced by approximately 100 m. Preservation during the end-Pleistocene into the early Holocene in the 16–9 ka window is evident in section 06-10 near Umm al-Qawain airstrip, section 06-13 at al-Jazirat al- Hamra and section 06-18 to the south of Ras al-Khaimah city. There is also limited evidence of preservation within this period at section 06-15 near al-Dhaith with sample 06-15-5 generating an age of 10.77 ± 0.43 ka. Although this 4.7 m section through a northwest–southeast orientated secondary dune was sampled from crest to base, given similarities in the geomorphological setting of this and the other sites mentioned above, it is possible that deeper excavation would have yielded further samples dating to earlier in this preservation window. Indeed an earlier study in the Al-Daith area yielded ages between 18 and 14 ka between 10 m and 4.6 m below the secondary dune crest surface (Parker and Goudie, 2007; see Figure 9). At sites 06-10 and 06-13, the uppermost samples in the sections give ages of 9.44 ± 0.50 ka at 0.3 m and 10.31 ± 0.54 ka at 1.0 m, respectively. This strongly suggests that at these locations, conditions favouring preservation of aeolian sediment accumulation have not prevailed during the Holocene and that the dune surfaces have largely remained active. If sediment deposition did take place, it seems likely that preservation was limited in duration and that the material was subsequently removed or reworked. Similar conditions may be suggested for site 06-18, although sample 06-18-2 at 1.5 m generates an age of 1.11 ± 0.12 ka, suggesting limited or ephemeral sediment accumulation and preservation since the underlying sample 06-18-3 was deposited at 9.41 ± 0.44 ka.

Figure 8.

Age–depth profiles for sites analysed in this study. (a) Umm al-Qawain airstrip UAE06-10; (b) al-Jazirat al-Hamra: UAE06-13, (c) al-Jazirat al-Hamra UAE06-14; (d) al-Dhaith: UAE06-15; (e) Ras al-Khaimah South: UAE06-18; (f) Tawi Asmar UAE07-6; (g) al-Ain: UAE07-9. Circles denote OSL ages, crosses denote ¹⁴C ages

Figure 9.

Dune accumulation ages from this paper (1-6) related to previously published data from the Rub’ al-Khali (7-10) and Wahiba Sands (11). Sources: (7) Goudie et al. (2000); Parker et al. (2004, 2006); (8) Parker and Goudie (2007); (9) Goudie et al. (2000); (10) Bray and Stokes (2004); (11) Preusser et al. (2002); Radies et al. (2004). Note OSL ages only are shown, with associated 1 sigma errors

This is not the case at 06-15 (al-Dhaith) where samples 06-15-4, 06-15-3 and 06-15-2 give ages of 7.73 ± 0.41 ka, 2.95 ± 0.34 ka and 2.72 ± 0.32 ka, respectively. In an earlier study in the al-Daith area Parker and Goudie (2007) noted dune ages of 7.58±0.59 ka at 3.8 m depth. A shell midden horizon at 2.85 m was dated by ¹⁴C to 5.06±0.12 ka cal. BP representing a former occupation surface. Further ages 4.18±0.72 ka and 0.54±0.17 ka were recorded at 2.6 m and 1.2 m, respectively. At 07-6 near Tawi Asmar, where a more complete record of aeolian accumulation exists between 7.32 ± 0.88 ka and 2.07 ± 0.14 ka. It should be noted that at this site, there is an abrupt discontinuity down core from sample 07-6-9 at 5.0 m (7.32 ± 0.88 ka) to the underlying sample 07-6-10 at 5.5 m (27.00 ± 2.07 ka). The major episode of preserved aeolian accumulation between approximately 16 and 9 ka recorded at the sections described earlier is not recorded here. Since the samples from this site were retrieved by coring rather than from an exposed section, no access to internal dune stratigraphy and structure was possible. However, it may be assumed that this hiatus in the preserved record of approximately 20 ka represents a significant transition between two distinct phases of deposition. Given the estimated 5 m height of this secondary dune, it is likely that these two underlying samples (07-6-10, 27.00 ± 2.07 ka and 07-6-11, 26.65 ± 1.31 ka), the only two in this study that yield pre-LGM ages, represent the underlying primary dune generation upon which the secondary features are superimposed. The lack of stratigraphic information for the site makes this difficult to confirm.

Estimated net accumulation rates at sites 06-15 and 07-6 during this second phase of deposition are more rapid than those for the 16-9 ka window, ranging between 2.25 and 2.49 m/ka. The geomorphological settings of these sites are not significantly unlike those of other locations described where a considerably different aeolian record is preserved. Further, there is no obvious convergence in their geographical positions within the wider study area compared with other sites. As such, it seems reasonable to suggest that localized or site-specific factors rather than broader subregional processes may be responsible for the variance observed.

Near al-Jazirat al-Hamra, samples were retrieved from two artificially exposed sections in an excavated pit. The two sections, 06-13 and 06-14, are laterally displaced by approximately 100 m with samples taken from 06-14 adjacent to various horizons of archaeological interest consisting of charcoal pits, lenses and more laterally continuous units to allow comparison between OSL and radiocarbon dating methods. Despite the proximity of the sections within the same site, there are significant differences in sample ages near the tops of the profiles. Samples 06-13-1 and 06-13-2 yield ages of 10.31 ± 0.54 ka and 9.71 ± 0.55 ka at 1.0 and 2.0 m down-section, respectively, the apparent age inversion possibly due to variability in the ICP-MS determined Th and U concentrations and the subsequent effect on calculated dose rates. Initially, this seems to suggest either that the upper metre of sediment has been active during most of the Holocene, or that any material deposited has been subsequently eroded and transported southeast. However, in comparison, samples 06-14-1 at 1.5 m down-section and 06-14-2 at 2.6 m give ages of 4.52 + 0.23 ka and 10.78 ± 0.49 ka, respectively. A plausible explanation may be that, in a dynamic aeolian environment where dune crests are highly mobile, differences in sample ages reflect the variability of dune mobility and preservation of sands in upper dune locations.

Conclusions

Much previous work has correlated records of ages from aeolian archives either with phases of aridity or with phases of enhanced aeolian activity. Such conclusions have often been arrived at without access to the stratigraphic context from which the samples are retrieved. In this study, the employment of a rigorous and geographically intensive sampling programme in combination with access to exposures through dune cores and a robust dating protocol has added to the record of aeolian accumulation and preservation chronologies in this region and furthered our understanding of the palaeoenvironmental significance of age–depth profiles through dune features.

The ages determined for samples collected from secondary dune ridges in this region of the northeast Rub’ al-Khali demonstrate the significance of the 16-9 ka and 7-2 intervals for conditions favouring preservation of aeolian sedimentary accumulations that developed under the influence of sand transport by northwesterly winds. Although differences between the extent of preservation during these two windows are recorded at the sites described in this study, the high degree of consistency in the data suggests that local complexity on a subregional scale reflects site-specific processes, such as topography, coastal proximity (and water-table variability), anthropogenic activity and vegetation cover.

The convergence in recorded preservation ages between the northwest–southeast orientated secondary dune features described here and the underlying primary ridges orientated northeast–southwest (Atkinson et al., 2011) has significance for the interpretation of the palaeoenvironmental history of multiple overlapping dune generations. Previous work from Mauritania has hypothesised that superimposed dune generations each represent, and were formed under, discrete episodes of contrasting atmospheric circulation dynamics. In the Rub’ al-Khali at least, this interpretation does not account for the pattern of sediment and dune form activity and preservation. This complexity is revealed here because of the large number of dated samples used to characterise the record in the region, whereas previous studies (e.g. Lancaster et al., 2002 from Mauritania) relied on a smaller number of samples to draw conclusions about the development of different dune generations.

In the study area, the northeast–southwest primary dune ridges may have largely formed prior to the 16-9 ka period, as suggested by the age–depth profile of the exposed dune section at Shu’ayb (UAE07-7 (Atkinson et al., 2011). It seems that re-working of upper units and the development of their common asymmetric profile with steeper southeastern flanks coincides with the preservation of the overlying secondary dune generation during the phases highlighted here. Similarly, the primary dune section dated at al-Ain, underlain by wadi gravels, proved to be an exclusively Holocene feature, forming within the 7–2 ka window during which it has been shown that significant accumulation of secondary dune features was being preserved further north in the study area. Therefore it can be seen that it is possible for preservation of different dune morphologies to take place during phases of consistent atmospheric circulation regimes and that re-working of underlying dune generations can be synchronous with the preservation of an overlying subsequent dune generation. Thus whilst there are two sets of dune features in the region (megaridges and overlying secondary forms) their development is not independent, but interrelated.

Footnotes

Acknowledgements

We are grateful to Christian Velde and Imke Moellering, Department of Antiquities and Museums, Government of Ras al-Khaimah, and Dr Asma al-Farraj, United Arab Emirates University, for field assistance and support. We are grateful to Dr Matt Telfer, formerly of School of Geography and the Environment, University of Oxford, for assistance and advice during the optical dating programme.

David Thomas was in receipt of the Royal Geographical Society’s Thesiger Oman International Fellowship, which provided fieldwork costs for this research project. Dose rate determinations were possible thanks to successful NERC applications OSS316/0506 and OSS/327/1206 and radiocarbon dates were funded by NERC allocation No. 2006/11/07.

References

Abu-Zeid

Baghdady

El-Etr

(2001) Textural attributes, mineralogy and provenance of sand dune fields in the greater Al Ain area, United Arab Emirates. Journal of Arid Environments 48: 475–499.

Arbogast

(1996) Stratigraphic evidence for late Holocene aeolian mobilization and soil formation in south central Kansas, USA. Journal of Arid Environments 34: 403–414.

Atkinson

OAC

Thomas

DSG

Goudie

Bailey

(2011) Late Quaternary chronology of major dune ridge development in the northeast Rub’ al-Khali, United Arab Emirates. Quaternary Research 76: 93–105.

Besler

(1982) The north-eastern Rub-Al Khali within the borders if the United Arab Emirates. Zeitschrift für Geomorphologie 26: 495–504.

Beveridge

Kocurek

Ewing

Lancaster

Morthekai

Singhvi

. (2006) Development of spatially diverse and complex dune-field patterns: Gran Desierto Dune Field, Sonora, Mexico. Sedimentology 53: 1391–1409.

Bray

Stokes

(2004) Temporal patterns of arid-humid transitions in the south-eastern Arabian Peninsula based on optical dating. Geomorphology 59: 271–280.

Bray

Bailey

Stokes

(2002) Quantification of cross-irradiation and cross illumination using a Risø TL/OSL DA-15 reader. Radiation Measurements 35: 275–280.

Breed

Grow

(1979) Morphology and distribution of dunes in sand seas observed by remote sensing. US Geological Survey, Professional Paper 1052: 253–302.

Bristow

Duller

GAT

Lancaster

(2006) Age and dynamics of linear dunes in the Namib Desert. Geology 35: 555–558.

10.

Clarke

Rendell

(2003) Late Holocene dune accretion and episodes of persistent drought in the Great Plains of Northeastern Colorado. Quaternary Science Reviews 22: 1051–1058.

11.

El-Sayed

(1999) Sedimentological characteristics and morphology of the aeolian sand dunes in the eastern part of the UAE: A case study from Ar Rub’ Al Khali. Sedimentary Geology 123: 219–238.

12.

Fitzsimmons

Rhodes

Magee

Barrows

(2007) The timing of linear dune activity in the Strzelecki and Tirari Deserts, Australia. Quaternary Science Reviews 26: 2598–2616.

13.

Forman

Oglesby

Webb

(2001) Temporal and spatial patterns of Holocene dune activity on the Great Plains of North America: Megadroughts and climate links. Global and Planetary Change 29: 1–29.

14.

Galbraith

Roberts

Laslett

Yoshida

Olley

(1999) Optical dating of single and multiple grains of quartz from Jinmium rock shelter, northern Australia: Part I, Experimental design and statistical models. Archaeometry 41: 339–364.

15.

Garzanti

Ando

Vezzoli

Dell’Era

(2003) From rifted margins to foreland basins: Investigating provenance and sediment dispersal across Desert Arabia (Oman, U.A.E.) Journal of Sedimentary Research 73: 572–588.

16.

Glennie

(1998) The desert of southeast Arabia: A product of Quaternary climatic change. In: Alsharhan

Glennie

Whittle

Kendall

(eds) Quaternary Deserts and Climate Change. Rotterdam: Balkema, 279–291.

17.

Goudie

(2002) Great Warm Deserts of the World. Oxford: Oxford University Press, 444 pp.

18.

Goudie

Colls

Stokes

Parker

White

Al-Farraj

(2000) Latest Pleistocene and Holocene dune construction at the northeastern edge of the Rub Al Khali, United Arab Emirates. Sedimentology 47: 1011–1021.

19.

Hughen

Baillie

MGL

Bard

Bayliss

Beck

Bertrand

. (2004) Marine04 Marine radiocarbon age calibration, 26–0 ka BP. Radiocarbon 46: 1059–1086.

20.

Lambeck

(1996) Shoreline reconstructions for the Persian Gulf since the last glacial maximum. Earth and Planetary Science Letters 142: 43–57.

21.

Lancaster

Kocurek

Singhvi

Pandey

Deynoux

Ghienne

J-F

. (2002) Late Pleistocene and Holocene dune activity and wind regimes in the western Sahara Desert of Mauritania. Geology 30: 991–994.

22.

Murray

Wintle

(2006) A review of quartz optically stimulated luminescence characteristics and their relevance in single-aliquot regeneration dating protocols. Radiation Measurements 41: 369–391.

23.

Parker

Goudie

(2007) Development of the Bronze Age landscape in the southeastern Arabian Gulf: New evidence from a buried shell midden in the eastern extremity of the Rub’ al-Khali desert, Emirate of Ras al-Khaimah. UAE Arabian Archaeology and Epigraphy 18: 132–138.

24.

Parker

Eckersley

Smith

Goudie

Stokes

White

. (2004) Holocene vegetation dynamics in the northeastern Rub’ al-Khali desert, Arabian Peninsula: A pollen, phytolith and carbon isotope study. Journal of Quaternary Science 19: 665–676.

25.

Parker

Goudie

Stokes

White

Hodson

Manning

. (2006) A record of Holocene climate change from lake geochemical analyses in southeastern Arabia. Quaternary Research 66: 465–476.

26.

Preusser

Radies

Matter

(2002) A 160,00-year record of dune development and atmospheric circulation in southern Arabia. Science 296: 2018–2020.

27.

Radies

Preusser

Matter

Mange

(2004) Eustatic and climatic controls on the development of the Wahiba Sand Sea, Sultanate of Oman. Sedimentology 51: 1359–1385.

28.

Sarnthein

(1972) Sediments and history of the postglacial transgression in the Persian Gulf and northwest Gulf of Oman. Marine Geology 12: 245–266.

29.

Sarnthein

(1978) Sand deserts during Glacial Maximum and climatic optimum. Nature 272: 43–46.

30.

Southon

Kashgarian

Fontugne

Metivier

Yim

WW-S

(2002) Marine reservoir corrections for the Indian Ocean and Southeast Arabia. Radiocarbon 44: 167–180.

31.

Stokes

Bray

(2005) Late Pleistocene eolian history of the Liwa region, Arabian Peninsula. Geological Society of America Bulletin 117: 1466–1480.

32.

Stokes

Thomas

DSG

Washington

(1997) Multiple episodes of aridity in Africa south of the equator since the last interglacial cycle. Nature 388: 154–158.

33.

Stone

AEC

Thomas

DSG

(2008) Linear dune accumulation chronologies from the southwest Kalahari, Namibia: Challenges of reconstructing late Quaternary palaeoenvironments from aeolian landforms. Quaternary Science Reviews 27: 1667–1681.

34.

Telfer

Thomas

DSG

(2007) Late Quaternary linear dune accumulation and chronostratigraphy of the southwestern Kalahari: Implications for aeolian palaeoclimatic reconstructions and predictions of future dynamics. Quaternary Science Reviews 26: 2617–2630.

35.

Telfer

Bailey

Burrough

Stone

AES

Thomas

DSG

Wiggs

GSF

(2010) Understanding linear dune chronologies: Insights from a simple accumulation model. Geomorphology 120: 195–208.

36.

Teller

Glennie

Lancaster

Singhvi

(2000) Calcareous dunes of the United Arab Emirates and Noah’s Flood: The postglacial reflooding of the Persian (Arabian) Gulf. Quaternary International 68–71: 297–308.

37.

Thomas

DSG

Burrough

(2010) Interpreting geo-proxies of late Quaternary climate change in African drylands: Implications for understanding environmental and early human behaviour. Quaternary International doi: 10.1016/j.quaint.2010.11.00110.1016/j.quaint.2010.11.001.

38.

White

Goudie

Parker

Al-Farraj

(2001) Mapping the geochemistry of the northern Rub’ Al Khali using multispectral remote sensing techniques. Earth Surface Processes and Landforms 26: 735–748.

39.

Zander

Degering

Preusser

Kasper

Brückner

(2007) Optically stimulated luminescence dating of sublittoral and intertidal sediments from Dubai, UAE: Radioactive disequilibria in the uranium decay series. Quaternary Geochronology 2: 123–128.