Landscapes

Abstract

Namibia has a wide range of landscapes and the classification provided in the Atlas of Namibia is adopted by way of introduction to their nature and diversity. This is followed by a description of three main landscape regions: the Namib Desert, the Kalahari Desert, and the Great Escarpment. Finally, the chapter includes a regional analysis of the river systems and the coastline of Namibia, with feature boxes on two iconic landscapes, the Fish River Canyon and Sandwich Harbour.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Landscape Types

Namibia is a vast and varied country with wonderful landscapes and landforms, many of which have been engagingly portrayed in words and pictures by Swart and Marais (2009). It is particularly notable because of the richness and beauty of its desert landforms, and because of what it can tell us about the long-term tectonic history and climate of this part of Africa. The major controls on landscape evolution are tectonics (and its influence on geology ) and climate (and its influence on ecology) and their dynamic interrelationships over a range of timescales. However, before introducing the outlines of tectonic and geological histories (Chap. 2) and the dynamics of climate and ecosystems (Chaps. 3 and 4) this chapter introduces the major characteristics of Namibian landscapes and their diversity, including the nature of the two great deserts (the Namib and the Kalahari ), the Great Escarpment which runs down its spine, its rivers , and its long coastline .

In the second part of the book we present a series of regional studies illustrating some of the most dramatic and interesting landforms and landscapes of the country, and these are approximately arranged from north to south (Fig. 1.1). Chapter 25 is not marked on Fig. 1.1, as the phenomena it describes occur over much of Namibia. We have chosen to include landscapes and landforms for which there is a good array of scientific literature and which reflect the diversity of landscape types in Namibia. Other important and much visited landscape features, such as the Fish River Canyon and Sandwich Harbour , are described in boxes in Chap. 1, while another characteristic landform type—sandstone-capped mesas—is featured in box 3 in Chap. 2.

Fig. 1.1

The approximate locations of Chaps. 5–24 and the relief of Namibia (from Mendelsohn et al. 2002, p. 39 in http://www.uni.koeln.de/sfb389/e/e1/download/atlas_namibia/) (accessed 30th January 2014)

Covering an area of about 823,680 km², Namibia is up to 1,320 km long, and 1,440 km wide. It is, however, sparsely populated with only around 2 million inhabitants. To the north it is bounded by Angola and Zambia, to the east by Botswana, to the south by South Africa , and to the west by the cold waters of the Atlantic Ocean . Much of Namibia consists of a wide plateau at 900–1,300 m above sea level (Fig. 1.1). This plateau is bounded on the west by a large escarpment and on the east by the Kalahari Basin. Bordering the Atlantic in the west is the lower-lying coastal plain of the hyper-arid Namib Desert (Van Zyl 1992). Van der Merwe (1983) estimated that plains were the dominant landscape of Namibia, covering over 45 % of the country, with mountains covering c 19 %, dunes just under 14 %, plains with scattered hills over 13 %, and hills just under 8 %. Wellington (1967) divided the Namibian landscape into three main types—the Namib Desert , the Plateau Hardveld and the Kalahari Sandveld, within which he identified a number of more specific landscape types.

Recently, The Atlas of Namibia (Mendelsohn et al. 2002) has identified a range of landscape regions in the country, providing a useful framework to describe the geomorphological diversity (Fig. 1.2). These regions are described here heading from north to south, apart from two areas (the Kalahari sandveld and the Great Escarpment ) which span large distances from north to south in the eastern and western parts of the country respectively, with which we begin. The Kalahari sandveld occupies a huge part of northern and eastern Namibia. It is a generally monotonous, flat, basin of sedimentation, much of which is characterised by aeolian landforms, including linear dunes and pans (Thomas and Shaw 1991). It is discussed further in Chap. 21. The Escarpment (or Great Escarpment ), discussed later in Chap. 1, runs roughly parallel to the coast and divides much of the country up into two general landscapes: the low-lying coastal plain to the west, and the higher inland plateau to the interior. It is not a continuous feature, and is largely absent from the Central-Western Plains.

Fig. 1.2

The landscape divisions of Namibia (from Mendelsohn et al. 2002, p. 14, in http://www.uni-koeln.de/sfb389/e/e1/download/atlas_namibia/) (accessed 30th January 2014)

In the far north east is a small area called the Caprivi Floodplains, created by the Zambezi and Kwando rivers and consisting of a network of channels, spectacular oxbow lakes and grasslands. In the late Pleistocene it may have been occupied by a lake, called Lake Caprivi (Shaw and Thomas 1988). Further east, the Okavango Valley occurs as a narrow strip along Namibia’s northern border. The Karstveld of northern Namibia covers a scatter of areas in the east and west and is underlain by soluble carbonate rocks, including limestones and dolomites , and has an array of karstic forms including caves and sinkholes, of which Guinas and Otjikoto are the most dramatic examples (see Chap. 6). Pans are a typical Namibian landscape element, represented most notably in the north by Etosha Pan (see Chap. 6). The Cuvelai system which lies between Etosha Pan and the Angolan border, is dominated by a network of curious, shallow channels, called ‘Oshanas ’, which in wet years obtain much of their water from the Angolan Highlands. The Kunene Hills in the far north west of Namibia, sometimes called the Kaoko Highlands, are a rugged area of dissected ancient rocks, commonly 1,000–1,900 m above sea level (Sander 2002). The hills include the Baynes , Steilrand , and Zebra Mountains . The Zebra Mountains are so called because they consist of a mass of interlayered, relatively unaltered dark leucotroctolite with relatively altered, “white,” anorthosite (Maier et al. 2013). Glacial features, originating in the Dwyka phase (c 200–270 million years ago), have been exhumed and are widespread. These include U-shaped valleys with striated walls. The Etendeka Plateau (discussed further in Chap. 9) consists of flat-topped hills underlain by great expanses of volcanic lava and some sedimentary Karoo age rocks. The lavas were spewed out when Africa and South America split apart some 132 million years ago (as discussed in more detail in Chap. 2). The Kamanjab Plateau, mostly underlain by ancient granites and gneisses, is in the north west of the country and is drained and dissected by the Huab and Ombonde rivers .

A large landscape unit within central Namibia is the Central-Western Plains, much of which lies between 500 and 1,000 m above sea level, stretches inland from the Atlantic coast and has been formed by rivers such as the Khan, Omaruru, Swakop and Ugab cutting back eastwards into higher ground. The area is studded with upstanding granite hills called inselbergs (Mabbutt 1952). These Inselbergs are large, free-standing mountain masses that punctuate the Central-Western Plains, and include Brandberg (at 2,579 m, the highest point in Namibia), Erongo (see Chap. 11), Paresis and Spitzkoppe (see Chap. 10). The Khomas Hochland Plateau is located in the centre of the country around Windhoek, and consists of a ridge of rolling hills and deep valleys. Much of it lies at altitudes between 1,700 and 2,000 m above sea level and it receives sufficient rainfall to feed such rivers as the Nossob, the Kuiseb and the Swakop. It is the remnant of a once-great mountain chain created towards the end of the Damara stage (c 550 million years ago) as a result of the collision of continents. This area contains one somewhat anomalous landform curiosity, a field of late Pleistocene dunes neither connected with the Namib or the Kalahari , up to c 7 m tall, located at Teufelsbach, some 20 km south of Okahandja (Eitel et al. 2004). The Naukluft Mountains , which largely consist of limestones and shales, occur on the edge of the Great Escarpment within central Namibia. The mountains are highly dissected by small, steep valleys in which extensive spreads of calcareous tufa occur (see Chap. 20). The Rehoboth Plateau lies in the centre of the country to the south of Windhoek at an altitude of between 1,500 and 1,700 m above sea level. It is an area of inselbergs and rolling terrain underlain primarily by granites and complexes of metamorphic rocks.

The western parts of central and southern Namibia are dominated by sand and plains. The Namib Sand Sea (see Chap. 18) stretches for 400 km north from Lüderitz (now officially known as #Naminus) to Walvis Bay and is up to 100–140 km in width. It contains a wide array of large and mobile dunes . The Namib Plains (see Chaps. 12 and 14) consist of gravel and gypsum covered surfaces, rocky outcrops and hills, which together with the Namib Sand Sea make up a large proportion of the coastal plain seaward of the escarpment. Sand ramps are often banked up against hills (Bertram 2003). The Karas Mountains of southern Namibia consist of uplifted blocks of sandstones, limestones and shales that rise up above the surrounding plains. The highest peak in the Gross Karas Mountains reaches 2,203 m above sea level. The Gamchab Basin is an area to the north of the Orange River , with large valleys created by river erosion. Over much of the area drainage densities (the amount of stream channel per unit area) are high and there are extensive fan systems. The Islands , of which there are 12 main ones, occur just offshore between Walvis Bay and the Orange River and have been noted for their rich guano resources (Watson 1930). Whilst they are small and inconspicuous features, they have been given intriguing names such as Plumpudding and Roastbeef. The Nama-Karoo Basin is a predominantly flat-lying plateau underlain by sedimentary rocks, which slopes from 1,400 m above sea level in the north to 900 m in the south. This region includes the Schwarzrand to the south of Maltahöhe. It is drained by rivers such as the Fish, which flows to the Orange. Some ancient inselbergs have been exhumed from beneath the former Late Proterozoic to Cambrian Nama sedimentary cover (Stengel 2000; Stengel and Busche 2002) either because of river erosion or groundwater-related weathering effects (Twidale and Maud 2013). Fossil landslides have been extensively developed (Stengel 2001), probably as a result of higher precipitation amounts than today. Brukkaros forms the only major mountain in this area (see Chap. 24). Finally, the Weissrand Plateau is an intriguing area of solution hollows (dayas ), calcrete , aligned drainage and old dunes sandwiched between the Nama-Karoo Basin and the Kalahari sandveld (see Chap. 22).

2 The Namib and the Kalahari Deserts

Of the composite landscape types of Namibia, two of the largest and most important are the two great deserts, the Namib in the west and the Kalahari in the east. The Namib Desert landscape comprises a range of landscape types from hills to gravel plains and dune fields, whereas the Kalahari Desert is dominated by stabilised dunes (Fig. 1.3).

Fig. 1.3

Google Earth image of gramadullas to the north of the Kuiseb River . Scale bar 2 km (© 2013 GeoEye, Google)

2.1 The Namib Desert

The Namib, one of the world’s driest and most beautiful deserts, extends for more than 2,000 km and eighteen degrees of latitude along the Atlantic coast of southern Africa from the Olifants River in South Africa (latitude 32°S) to the Carunjamba River (latitude 14°S) in Angola . Being on the west side of the continent, in a zone of subsiding anticyclonic air, and bounded by the cool Benguela current offshore (Dingle et al. 1996), the Namib is hyper-arid (see Chap. 3). On its inland side it is bordered by a portion of the Great Escarpment which forms the western edge of the interior plateau and basin of southern Africa. Thus the Namib Desert forms a rather narrow strip some 120–200 km wide.

The geomorphology of the Namib Desert has been described by a number of workers (e.g. Gevers 1936; Cloos 1937; Logan 1960; Spreitzer 1965; Beaudet and Michel 1978; Hövermann 1978; Wilkinson 1990; Lageat 1994, 2000; Besler et al. 1994) and its context within the Cenozoic history of southern Africa is treated in Partridge and Maud (2000). The landforms in proximity to Gobabeb , the base for much of the work that has been done on the desert, are described in Eckardt et al. (2013). The Namib Desert can be subdivided into four main landscape types. In the area south of Lüderitz there is ‘The Southern or Transitional Namib’, which includes coastal Namaqualand and the diamond mining lands of the Sperrgebiet (Pallett 1995). This zone is cut through by the Orange, the last perennial river until the Kunene is reached on the Angolan border. It includes the rugged terrain of the Richtersveld and some areas of dunes —the Obib Dunes—to the north of the Orange. The area around Lüderitz and Elizabeth Bay has high velocity winds and there is extensive yardang development, rock fluting and deflation (Lancaster 1984; Corbett 1993).

The second Namib Desert landscape is that of the ‘Namib Sand Sea’ (see Chap. 18) which extends between Lüderitz and Walvis Bay and contains some of the world’s biggest dunes. This area has been the subject of a detailed review by Lancaster (1989). The third Namib Desert landscape is the ‘Central Namib Plains’. These lie between the Kuiseb River and more dissected terrain that lies to the north of the Brandberg . The plains have a low gradient of only 1° between the coast and the 1,000 m contour, and are studded with marble and dolerite ridges, some isolated inselbergs and complexes of shallow pans (Eckardt et al. 2001; Eckardt and Drake 2011). The plains show many windstreaks oriented with the easterly ‘berg’ winds which generate some dust plumes, which head out across the South Atlantic. Although the area is hyper-arid, the plains are also crossed by a very dense and intricate network of shallow drainage lines. This is very evident, for example, on the gently sloping rock surface immediately behind the coastal dunes between Walvis Bay and Swakopmund . Locally, ephemeral rivers such as the Swakop are more deeply incised into the plains, producing gorges and areas of badlands (called gramadullas or a moon landscape) (Fig. 1.3). In places these gorges truncate groundwater aquifers so that seepage occurs. This produces tufas made either of lime or of halite. The fourth landscape—‘The Northern Namib and Skeleton Coast ’—includes a dissected area of sandstone and lava hills, known as the Kunene Hills (also called Kaoko Highlands or Kaokoveld), together with some coastal dunefields (Lancaster 1982). The Kunene Sand Sea, which extends into Angola , is cut through by the perennial Kunene River (see Chap. 5).

It is also possible to divide the Namib Desert on the basis of its climate . Besler (1972), for instance, related weathering and other phenomena to the fog environment, introducing three divisions: the cool fog desert at the coast, the desert steppe in the east and the warm (alternate) fog desert in the middle.

One important landscape type, related to desert conditions, is found in between the Namib and Kalahari deserts in northern Namibia. Here, there are some quite extensive deposits of loess . Loess is a largely non-stratified and non-consolidated silt, containing some clay, sand, and calcium carbonate. It consists chiefly of quartz, feldspar, mica, clay minerals and carbonate grains in varying proportions. The grain size distribution of typical loess shows a pronounced mode in the range 20–40 μm and is generally positively skewed towards the finer sizes. It was the great German geographer, Ferdinand von Richthofen (1882, pp. 297–298), who had travelled to the classic deposits in China, who cogently argued that these intriguing deposits probably had an aeolian origin and that they were produced by dust storms transporting silts from deserts and depositing them on desert margins. Thus, it is likely that the Namibian loess deposits on the margins of the Kalahari and Namib Deserts have been produced from dust originating in these arid environments.

Namibian loess locations include the Opuwo basin and Omungunda in the Kaokoland area, where they were originally thought to be of late Holocene age (Brunotte and Sander 2000). However, Brunotte et al. (2009) have recently asserted that in the Opuwo area loess deposition commenced around 55,000 years ago (i.e. in the Pleistocene rather than the Holocene ). Loess, up to 5 m in thickness, also forms a fill in large basins in the valleys of the Huab and Hoanib rivers in the Khorixas district (Eitel et al. 2001), and appears to be of largely late Pleistocene age. It is believed that the loess is formed from material transported by westward moving dust storms from the eastern Kalahari under drier conditions than today. Even today, dust is generated in substantial quantities from the surfaces of the Mkagadikgadi depression in Botswana (Washington et al. 2003), and the Etosha Pan in Namibia (Bryant 2003). The loess is now being eroded by water to give areas of badlands.

2.2 The Kalahari Desert

In the interior of southern Africa, much of it in Botswana but a substantial part in eastern Namibia, lies the Kalahari Desert (Thomas and Shaw 1991). This area was the subject of a major study by the Prussian geographer Passarge (1904), though most of his observations took place in what in his day was called Bechuanaland. However, he did describe the stratigraphy and landforms of the Gobabis area in eastern Namibia.

It is difficult to say what precisely the borders of the Kalahari Desert are, not least because it has expanded and contracted during the last few million years. Much of it is a relict of a more extensive desert that once extended equatorwards well into the Congo Basin. It also merges with the Namib in the west and the Karoo in the south, and its boundary with the latter is often taken as the Orange River . The Kalahari, most of which lies at an altitude of around 1,000 m, derives its name from the Setswana word ‘Kgalagale’, which means ‘always dry’, but there are in a sense three Kalaharis, some drier than others:

• (a) The Kalahari dune desert in the arid south west interior of Botswana and adjoining parts of Namibia and South Africa . The primarily summer rainfall is less than 200 mm per annum and is just sufficient to stabilize the plinths of a major field of dominantly linear dunes . The dune crests are often active.

• (b) The Kalahari region (or thirstland) approximately delineated in the north by the Okavango Swamps and in the south by the Limpopo and the Orange rivers . This is an area of little or no surface drainage despite a relatively higher rainfall (c 600 mm per annum). Rates of groundwater recharge are very low (De Vries et al. 2000). It is almost entirely covered with grass and woodland, and has extraordinarily low relief.

• (c) The Mega-Kalahari, which is an extensive area consisting of a basin filled by the continental sediments of the Kalahari Beds. This extends from the Orange River as far as the Congo. Precipitation may be as high as 1,500 mm, but it displays the evidence of former aridity in terms of the development of ancient dune systems, drainage alignment, and pans (Shaw and Goudie 2002).

The Kalahari contrasts with the Namib Desert because of its relatively high rainfall and because of its basinal form. Because the climate of the Kalahari is semi-arid to sub-humid, most of it is not a true desert but an extensively wooded ‘thirstland’. Over enormous distances the relief is highly subdued and the landscape monotonous. The Kalahari owes its gross form and subdued morphology to the fact that following the break-up of Gondwanaland it became an area of down-warping bounded on the west by the highlands of Namibia and Angola , and on the east by mountains such as the Drakensberg and Lubombo (Haddon and McCarthy 2005) (Fig. 1.4). It became a basin of sedimentation and this largely accounts for its flatness. The Kalahari Beds that fill this basin are often over 100 m in thickness and in parts of the Etosha region of northern Namibia they are over 300 m thick (Fig. 1.5). Differences in thickness are related to graben (fault) structures. The sediments consist of terrestrial conglomerates, breccias, clays, dune sands, diatomaceous interdune deposits, alluvium, calcretes, silcretes and marls (Wanke and Wanke 2007). Depositional settings included braided rivers , sheet flood areas, shallow lakes and pans , and dune systems.

Fig. 1.4

The structural context of the Kalahari

Fig. 1.5

The extent and thickness of the Kalahari Beds (from Haddon and McCarthy 2005, Fig. 1)

Apart from its relict linear dunes (Thomas 1984) (see Chap. 21), the Kalahari contains large numbers of pans and associated leeward lunette dunes (Goudie and Thomas 1985) (see Chap. 23), together with two large closed depressions—the Etosha Pan of Namibia (see Chap. 6) and the Mkgadikgadi Depression of Botswana. Today these are major sources of dust plumes (Washington et al. 2003). In the mid to late Tertiary, palaeolake Etosha received water via the Cubango , Kunene and Cuvelai drainage systems. It largely dried up at about 4 million years (Ma) under conditions of progressively increasing aridity, though it still occasionally floods (Miller et al. 2010). The Kalahari shows excellent development of calcrete , silcrete and combinations of the two (Watts 1980; Nash et al. 1994; Shaw and Goudie 2004). The reason why calcretes in particular are so well developed probably relates to the structural context. The long history of gentle sedimentation within the Kalahari basin has created suitable conditions for the preservation of calcrete sequences, and the presence of ancient limestones and dolomites on the basin margins has supplied the necessary solutes.

3 The Great Escarpment

A third great landscape unit in Namibia is the Great Escarpment (Fig. 1.6), which lies inland of the Namib plains and rises up above them. This is one of the most important and conspicuous landforms of southern Africa (Kempf 2010). Such great escarpments ‘dominate the landscape of many rifted margins and are among the largest topographic features on Earth and compare with orogenic belts in majesty, although they are associated with plate divergence rather than subduction’ (Garzanti et al. 2014, p. 17).

Fig. 1.6

The Great Escarpment (from Goudie and Eckardt 1999, Fig. 1)

The Namibian Great Escarpment is part of a feature that stretches all around Southern Africa from Mozambique in the east to southern Angola in the west. In Namibia it comprises ranges such as the Naukluft Mountains (see Chap. 20) and the Gamsberg and contains some deep gorges. Of these the most impressive is the Fish River Canyon (Mvondo et al. 2011). This is one of the largest canyons in the world, being well over 500 m deep and extending for over 50 km. It is partly an ancient graben feature but is also the result of uplift and incision following the breakup of Gondwanaland (Grünert 2000).

The flat-topped Gamsberg is part of the Great Escarpment that separates the Khomas Hochland to the east from the low-lying Namib plains to the west. With an elevation of 2,347 m above sea level it towers above the Khomas Hochland by 450 m and above the Namib Plains by 1,100 m. It consists largely of Mesoproterozoic granite, but is capped by silicified aeolian sandstones of the 180 million years old Etjo Formation. These are in the form of a 30 m caprock which gives the Gamsberg its famous tabular summit. The nature of the underlying granite slopes of the Gamsberg is described by Moon and Selby (1983).

The age of the escarpment is a matter of debate (Partridge and Maud 1987) and its form and persistence are variable. Its development, in common with other passive tectonic plate margins (such as eastern South America and Western India), is probably closely related to continental fragmentation and rifting , but the timing of escarpment formation, retreat and uplift is more contentious. Kempf (2010) identifies two main theories in the literature. The first is that the Great Escarpment is the erosional remant of the rift shoulder that developed and was uplifted as a marginal bulge through the break up of Gondwana, and which since then has been worn back by as much as 100 km to its present position by erosional processes. This might have been enhanced by the continent rising up like a cork as material was eroded from the land and deposited in the ocean. This process is called ‘erosional isostasy’. Attempts to model the evolution of the Great Escarpment in terms of rifting , denudation and isostasy are provided by Gilchrist et al. (1994) and Dauteuil et al. (2013).

The second hypothesis is that the escarpment came into being a long time after the rifting and is thus the outcome of more recent tectonic events. Testing these hypotheses relies upon having well-dated histories of uplift and erosion, as well as good theoretical models of isostatic behaviour. There remains considerable disagreement.

That considerable denudation has occurred since the Cretaceous is indubitable, but whether there were periods of major denudation or more consistent trends is more debated. The elevations of the Damaraland complexes (such as Brandberg and Erongo ) above the surrounding plains suggests that well in excess of 1.5 km of denudation has occurred around these intrusions. Large amounts of Etendeka lavas have also been removed (Gilchrist et al. 1994).

The fact that Tertiary deposits such as the Tsondab Sandstone overlie the Namib plains suggests that a large amount of denudation occurred relatively soon after the early Cretaceous tectonic events. It is also conceivable, though not proven, that rates of denudation were reduced by progressive Cainozoic aridification (Gilchrist et al. 1994, p. 12220).

On Southern Africa’s south-western margin as a whole there is some evidence from apatite fission track analysis and the offshore sedimentary record that the early Cretaceous was a time of rapid denudation and offshore sedimentation (Brown et al. 1990; Rust and Summerfield 1990; Gallagher and Brown 1999). This could imply that following continental fragmentation there was indeed a phase of early and rapid uplift and scarp erosion, which led to the stripping of large expanses of Karoo strata (Gilchrist and Summerfield 1994; Van der Wateren and Dunai 2001) and thus supports the first hypothesis. Comparative fission track and cosmogenic data assembled by Cockburn et al. (2000) also suggest low rates of denudation affecting the central Namib over the last 36 million years. Sediment budgets constructed by calculating amounts of sediment deposited on the ocean floor offshore from Namibia and South Africa confirm this, demonstrating that rates were high in the Lower Cretaceous but have been considerably lower during the Tertiary and Quaternary (Guillocheau et al. 2010). However, sediments from offshore of the Orange mouth suggest that in addition to high rates of denudation during the post-rift Lower Cretaceous , there was another phase of rapid denudation in the Upper Cretaceous , 50 million years after the rifting event, perhaps as a response to a significant rejuvenation of relief at that time (Rouby et al. 2009) thus supporting the second hypothesis.

Burke (1996, pp. 364–369) has also cast some doubt on whether the Great Escarpment is as old as some studies imply. He queried the idea that because the Great Escarpment is parallel to the rifted continental margin it necessarily formed at the same time. Moreover, under erosive rainfall regimes he suggested it could still be evolving quickly. Burke suggested that much of the escarpment’s development has taken place over the last 30 Ma, thus contributing evidence in support of the second hypothesis. On the other hand, cosmogenic nuclide studies by Bierman and Caffee (2001) indicated that there has been significant landscape stability over at least the past million years. Similarly, Van der Wateren and Dunai (2001), also using cosmogenic nuclides, found that long-term rates of denudation have been very low (c 5 m per million years), especially over the last 5 million years thus supporting the first hypothesis. However, Codilean et al. (2012) have pointed out that the rates obtained by this method depend to a substantial degree on the grain size characteristics of the material that is sampled and that results based on pebble-sized clasts may underestimate palaeo-denudation rates. So, at present neither hypothesis can be ruled out until better dating evidence is available from a wider range of locations.

Another important question that needs to be answered is why the Great Escarpment is relatively ill-developed over much of the Central Namib (Birkenhauer 1991) and why there is what Kempf (2010) called ‘The Escarpment Gap’. One possible reason is that it is traversed by five rivers , including the Swakop and the Ugab, that have catchments that are longer and larger than those to the north and south. They derive power and discharge from areas that have relatively high rainfall and so may have concentrated erosion on this zone (Gevers 1936).

Alternatively, Hüser (1989) postulated that the gap is essentially lithological in origin, and is caused on the one hand by the belt of Cretaceous Damara granites and on the other by the fact that the well-developed escarpment to the north and south is the result of the presence of resistant Permo-Triassic Karoo sediments and the remains of Cretaceous Etendeka lavas. The granites, he argues, are not part of the escarpment but ought to be considered “foreign objects” which do not develop steps and terraces and are as such not able to form a classic escarpment. Spönemann and Brunotte (1989), on the other hand, favoured tectonism as being responsible for the break of the escarpment in the Central Namib Desert . They sought the cause in continental and regional scale deformations. The argument rumbles on, and it is still not clear which explanation is correct.

4 The Rivers

Namibia is a predominantly dry nation, and its rivers reflect this fact. On its borders there are three perennial rivers that reach the sea, the Kunene, the Zambezi and the Orange. Two other perennial rivers, the Okavango and the Kwanda flow into the inland Okavango Delta and the Linyanti Swamps of northern Botswana. All these rivers gain their flow from relatively humid and mountainous areas in Angola , Zambia and South Africa . The rivers that rise in Namibia itself are all ephemeral and seasonal. The great majority are dry for most of the year. Some, such as the Kuiseb, may not flow for several years and only occasionally reach the sea, while others, such as the Tsondab (Stone et al. 2010) and Tsauchab never manage to reach it under current conditions, terminating in the Tsondab and Sossus vlei s respectively.

Details of the flows and catchment areas of the Namibian river catchments are provided by Strohbach (2008). The following are the largest catchments (areas in km²):

Omatako Omuramba	61057
Etosha	57030
Fish	54326
Orange	44068
Nossob	37904
Ugab	29355
Auob	24540
Swakop	21010
Cuvelai	20730

However, there is some variability in estimates of catchment extent and, for example, Jacobson et al. (1995) estimate the area of the Swakop to be 30,100 km² and that of the Ugab to be 28,400 km². In the following paragraphs we discuss the nature of the perennial and ephemeral river systems of Namibia, starting with the ephemeral river systems that drain into the Atlantic, followed by the Orange and its tributaries, and ending with the internally-draining (endoreic) river systems of eastern Namibia.

The ephemeral rivers of the Namib, between the Orange in the south and the Kunene in the north (Jacobson et al. 1995; Jacobson and Jacobson 2013), rise in the interior mountains of Namibia and when they flow it is generally because of summer storms. In the south the Tsondab and the Tsauchab empty into pans within the Namib Sand Sea , and sometimes cause them to flood. In the past they flowed further west (Stone et al. 2010) and the Tsondab may have reached the Atlantic early in the Pleistocene (Seely and Sandelowsky 1974). The Tsauchab is notable for the way it has incised down into fan gravels and Tsondab Sandstone to produce the Sesriem Canyon (Grünert 2000, p. 133).

The Kuiseb, which has a catchment area of 15,500 km² and a length of c 420–440 km, rises in the Khomas Hochland of central Namibia and becomes incised into a canyon tract (Huntley 1985), which is up to 250 m deep and 1,000 m wide at its deepest part some 100 km from the coast. Incision occurred in the late Neogene (Van der Wateren and Dunai 2001). Dating and analysis of flood deposits indicate that this tract has received some 35 major floods over the last 1,300 years, with one attaining a discharge of 1,350 m³/s (Grodek et al. 2013). The river is lined by various terrace fragments, including the Pleistocene Oswater Conglomerate and the Gobabeb Gravel Formation (Ward 1987). Lower down its course, the Kuiseb becomes a braided, sandy alluvial channel, forms the northern boundary of the Namib Sand Sea (Fig. 1.7), except at its most seaward point, and disappears into a delta behind Walvis Bay . It only reached the sea three times in the Twentieth Century, in 1933, 1962–1963, and 1985. However, it welcomed the new millennium by flowing to the coast in 2000 and it flooded the salt works in Walvis Bay in 2011. These floods stop the northward movement of dunes and help to account for the largely sand-free nature of much of the Central Namib plains. In flood, it also carries prodigious amounts of woody debris (Jacobson et al. 1999) (Fig. 1.8). When it does flow, its discharge decreases rapidly downstream because of transmission losses into its bed. At Gobabeb (some 50 km from the coast) the volume of an event that statistically occurs one year in ten (known as the 10 year event) is 10 million m³, whereas at Swartbank (around 30 km from the coast) it is 3 million m³, and at Rooibank (c 20 km from the coast) a mere 0.15 million m³. The same applies to peak discharges which for a 10 year event are 90 m³/s at Gobabeb , 25 m³/s at Swartbank, and 0.9 m³/s at Rooibank (Heidbüchel 2007). Between Natab and the sea there is a series of buried palaeochannels of the Kuiseb, the southernmost of which enters the Atlantic just north of Sandwich Harbour (Klaus et al. 2008).

Fig. 1.7

The Kuiseb at Gobabeb , 2010. It passes through a snaking green line of riparian forest

Fig. 1.8

Woody flood debris in the Kuiseb bed, Gobabeb , 2011

To the north of the Kuiseb there is a small river which fails to reach the sea. This is the Tumas (Wilkinson 1990). In turn, to the north of that is the Swakop River , with a length of 460 km, which rises in the Khomas Hochland and enters the Atlantic at Swakopmund . Like the Kuiseb it is prone to flood sporadically and in 1931 it demolished the railway bridge linking that town with Walvis Bay (Figs. 1.9 and 1.10). As a result of another flood event in 1934 it transported sediment which built the coastline out by over 1 km (Massmann 1983). More recently, the Swakop reached the sea in February 2009 and in March 2011. It carries a substantial sediment load, thus justifying its local name, which derives from the local Nama words Tsoa (anus) and Xou (excrement). The major tributary of the Swakop is the Khan. This has its origin near the settlement of Otjisemba north-west of Okahandja. From there the river course passes westwards to the town of Usakos, and has its confluence with the Swakop 40 km east of Swakopmund . The Swakop canyon contains some interesting minor landforms. For example, just 8 km upstream from the Khan confluence, there is a large spring tufa , with a high halite content which is notable for enveloping and preserving ostrich (Struthio camelus) feathers. The present course of the Swakop runs further north than it did in the past for there is sedimentological evidence that it used to flow into the present Tumas, located between Walvis Bay and Swakopmund (Van der Wateren and Dunai 2001).

Fig. 1.9

The Swakop bridge during the 1931 flood (from Digital Namibian Archive)

Fig. 1.10

The remains of the same bridge at the mouth of the Swakop which was demolished by a flood in 1931

The Omaruru River , the bed of which is an important aquifer, reaches the Atlantic just to the north of Henties Bay (Stengel 1966). It has a catchment area of c 13,100 km² and a length of c 330 km. Like the larger Ugab it has a history of flooding. The Ugab is flanked by three conglomerate terraces composing the Bertram Conglomerate Formation at 160, 100 and 30 m above the modern river (Mabbutt 1951; Grünert 2000, p. 69) of which the famous Vingerklip (Fingerklip ), located c 80 km southwest of Outjo and 45 km west of Khorixas, is an erosional remnant (Fig. 1.11).

Fig. 1.11

Google Earth image of a terrace remnant on the Ugab River . Scale bar 0.25 km (© 2013 Digital Globe)

Further north, other ephemeral rivers enter the Skeleton Coast (see Chap. 8) and one of these, the Hoarusib, has a suite of major silt terraces that formed as a result of flood flow being ponded up behind a dune cordon. It reaches the sea most years. The Uniab , Hoanib and Hunkab are other rivers that are periodically ponded up by the Skeleton Coast erg (another name for a sand sea). Many of these rivers have comparatively lush riparian vegetation along their channels, a stark contrast to the adjacent sand and rock desert.

The Kunene rises in the Bié highlands of Angola , where it is called the Cunene, and flows into the Atlantic on the border between Angola and Namibia (Nicoll 2010). For much of its course it flows southwards, as if towards the Etosha Pan , an ancient structural basin (Buch and Trippner 1997), but then it turns sharply westwards and enters a tract with steep falls and rapids (e.g. the Caxambue rapids and the Epupa and Ruacana Falls ). The Ruacana Falls are c 120 m high, while at Epupa the river forms a series of riffles and cascades that drop a total of around 60 m over c 1.5 km. Good photos of the river at Ruacana prior to dam construction are provided by Kanthack (1921). Between Ruacana and the Atlantic the altitude of its bed drops by more than 1,100 m over a distance of 340 km. These characteristics seem to indicate that this is a case of a river capture by a stream eroding backwards from the coast and capturing the interior drainage (Wellington 1955, p. 65). Wellington also suggests that the conditions for a similar process of capture are present in the headwaters of the Rio Coroca to the north of the lower Kunene. This river, having eroded headward through the Sierra de Chela of the Great Escarpment , is threatening to behead the upper Caculuvar River, a tributary of the upper Kunene.

The timing of the Kunene capture is not well constrained (Moore and Blenkinsop 2002) but lowering of the base-level associated with the opening of the Atlantic may have initiated a period of rapid erosion, which may have exploited a Permo-Carboniferous glacial valley from which Karoo sediments were stripped. Miller (2008, Chap. 24) suggests that the capture of the Kunene took place towards the end of a widespread phase of drainage incision during the Miocene -Pliocene . It has also been argued that the capture, which perhaps caused the shrinkage of a postulated large Lake Etosha, occurred only c 35,000 years ago (Buch 1997). At the far western end of its course, as it passes through the coastal sand sea, there are signs that the Kunene formerly entered the sea considerably to the south of its present mouth (Sander 2002), and that it may have been forced northwards by dune encroachment. At Serra Cafema the Kunene has a 5 m high terrace with a sprinkling of unremarkable Middle Stone Age flakes on its surface (Nicoll 2010).

The Okavango is a perennial, endoreic river with ephemeral tributaries (Seely et al. 2003; Strohbach 2013). It rises in the south-western Angolan highlands, near and just east of the source of the Kunene and Cuvelai rivers . It flows for more than 600 km from the upper catchment in a southerly direction until it reaches border between Angola and Namibia. From that point, the river forms the border between Angola and Namibia over a distance of some 400 km. Its channel, which lies c 40–60 m lower than the surrounding sand plateau, is characterised by scroll bars, abandoned channels, oxbow lakes etc. (Fig. 1.12). The river is notable for the fact that it carries a very low silt and clay load, with the bulk of its sediment appearing to be re-worked fine aeolian sand. This is transported as bed load and creates sub-aqueous dune bedforms. Just before it enters its panhandle in Botswana, the river crosses the Popa Falls, which in the dry season have a visible height of c 3.5 m. The month of maximum flow is April. In the far east of the Caprivi , the Zambezi forms the border with Zambia. The Caprivi is often subject to severe flooding (Skakun et al. 2013).

Fig. 1.12

Google Earth image of the Okavango Floodplain. Note the oxbow lakes, abandoned channels, point and scroll bars, etc. Scale bar 0.5 km (© 2012 Google, US Dept. of State Geographer, Digital Globe)

The Cuvelai River is also an endoreic river, rising in the southern foothills of the Sierra Encoco in southwestern Angola (Mendelsohn and Weber 2011). It drains southwards towards the Etosha Pan and is perennial for about 100 km before it ramifies into a delta of ephemeral watercourses (Lindeque and Archibald 1991) which cross a broad plain of low relief; this delta converges again to terminate in the ephemeral Etosha pan . The watercourses (Fig. 1.13), called oshanas , appear to be a complex network of flood channels, most of which are oriented from northwest to southeast. They are the lifeblood of an area where just less than half of the population in Namibia live. However, severe floods can cause many deaths as in March 2009 and March 2011.

Fig. 1.13

Google Earth image of Oshanas . Scale bar 5 km (© 2014 Digital Globe)

The perennial Orange River , which occurs on the southern border of Namibia, originates in the Maloti Highlands of north-eastern Lesotho and after flowing through regions of steadily increasing aridity reaches the South Atlantic at Alexander Bay (Bluck et al. 2007). It is the largest catchment in southern Africa, and about one quarter of its area is within Namibia. Although its flow is now heavily regulated by dams, the river has been subject to major floods . A peak in flood activity may have occurred around 500 years ago in a phase contemporary with the Little Ice Age (Heine and Völkel 2011). The sediment load of the Orange, now greatly reduced compared to natural levels because of entrapment in reservoirs, has included diamond-rich gravels (Corbett and Burrell 2001; Spaggiari et al. 2006), and it has also been a major source of sand for the dunes of the Namib Sand Sea (Garzanti et al. 2012). Its terraces include the Arries Drift Gravel Formation, (which is 19–17 Ma old), also known as the Proto-Orange terrace, and a lower Meso-Orange terrace, which is thought to be of Plio-Pleistocene age (Jacob et al. 1999; Jacob 2005). A map of these terraces appears in Miller (2008, Fig. 25.16). The mouth of the Orange River is an example of a delta dominated by wave action and longshore drift. Rather than accumulating at the river mouth as a classic and visible delta, the sediments have been carried up the coast and onshore by the strong swells and onshore winds.

Box 1: The Fish River and its world class canyon

The Fish River rises in Namaqualand and flows south across the Great Namaqualand plateau, where it cuts a spectacular gorge or canyon before emptying into the Orange River . It is about 600 km long and is intermittent. The river often flows in the summer months (especially January to March) and severe flash floods can pose problems for the tourist resort at Ai-Ais. In 1973/74 and 1975/76, two very wet seasons, flows at the Hardap Dam exceeded 6,000 m³/s. Its Kam, Schlip and Kalf tributaries originate in the central highland area south of Rehoboth whilst the Narub and Usib Rivers flow from the eastern foothills of the Naukluft Mountains . The Hutup, Lewer and Kanibes Rivers drain from the northern and eastern parts of the Schwarzrand Mountains. The Löwen and Gaub Rivers originate in the Groot Karas Mountains and the Konkiep in the western Schwarzrand . It is possible that prior to its present course, the Fish flowed southeastwards from its southerly bend at Ganikobis past Tses and in the direction of the line of pans extending along the sandy lowland towards Aroab (Wellington 1967, p. 26).

The Fish River canyon is one of the greatest spectacles in Namibia (Fig. 1.14). Its scale is stupendous. It is often said to be the second greatest canyon in the world after the very much larger Grand Canyon of the Colorado in the USA. Whether this is true is a matter of debate for the Blue Nile gorge in Ethiopia is very much deeper. Other enormous canyons include the Tsangpo in Tibet and the Copper Canyon in Mexico. For the first 450 km of its course the Fish has a limited gradient and flows within a broad valley. However, after its confluence with the Löwen River, it begins to become incised and eventually enters a canyon tract (Simpson and Davies 1957). This is located about 80 km west of Grünau, and starts about 30 km upstream of the Ai-Ais hot springs, and extends for about 50 km. The gorge is between 160 and 550 m deep, and 5–8 km wide, and is incised into flat-lying Nama sediments and into the underlying Namaqua Complex gneisses, which themselves are some 1,800 million years old. The river, in incising, has created some enormous entrenched meanders. The river must initially have flowed over a flat land surface where it could develop its bends freely. Then, continental uplift associated with the breakup of Gondwanaland in the Lower Cretaceous occurred and this was a major factor that caused the incision to occur. The canyon is also associated with some major fault and graben structures and is thus in part a rift valley (Mvondo et al. 2011; Kounov et al. 2013). It has been of particular interest recently as an analogue of some of the great valley networks on Mars (Petau et al. 2011).

Fig. 1.14

The Fish River Canyon

The endoreic Kalahari catchments, such as the Omatako Omurambo, are prone to lose much of their discharge in the Kalahari sands. Some of them extend into Botswana and across to the margins of the Okavango Delta and Lake Mkgadikgadi. These are the dry valleys or mekgacha systems and are thought to be relicts of formerly more humid conditions (Thomas and Shaw 1991, p. 136). Examples include the Okwa and the Groot Laagte. The Auob and the Nossob, partially incised into calcreted valley sides, negotiate the linear dunes and eventually flow southeastwards into the Molopo. The Nossob has its origin in two main tributaries, the Swart-Nossob and Wit-Nossob, meaning black and white respectively. Both tributaries have their origins in the eastern slopes of the Otjihavera mountain range, east of Windhoek. Their sources are at 1,800 m and over 2,000 m above sea level respectively. The two river beds have their confluence some 80 km south of Gobabis, which is situated on the bank of the Swart Nossob.

One of the most intriguing features of the Namib rivers is that many of them, in contrast to most ‘normal’ rivers in temperate climates , display convex long profiles over all or much of their courses. This is the case for the Kunene, Kuiseb, Omaruru, Swakop, Tumas, and Ugab (Dauteuil et al. 2013), as well as some of the southern Angolan rivers , and the lower course of the Orange. Whether this is due to the nature of the uplift on this passive tectonic continental margin, or to the fact that in this arid environment river discharges diminish downstream, or to a combination of both, is a matter of debate. The traditional explanation is that in dryland rivers there is a diminution in flow downstream because of transmission losses, and so the ratio of sediment to flow often increases downstream, leading to aggradation and the development of a convex profile. It is also possible that in these rivers there is a less clear diminution in grain size of sediment downstream in comparison with humid climate rivers. Certainly, studies of drainage basins in the high plains of the USA (Zaprowski et al. 2005), indicate that in tectonically stable settings, areas with higher intensity rainfall and greater mean annual precipitation, have increasingly concave long profiles. However, a strong case has been made that uplift rate histories explain many of the main characteristics of the long profiles of those rivers that drain the tectonic swells of Africa (Roberts and White 2010).

5 The Coastline

The coastline of Namibia stretches 1,570 km from the mouth of the perennial Kunene river in the north to that of the perennial Orange river in the south (Bird et al. 2010). In between, all the rivers that flow into the Atlantic are ephemeral. The coastal environments of Namibia are shaped by the dynamic interplay between sediments coming from on land (transported by rivers and the wind) and the waves, tides and currents which characterise the eastern margins of the Atlantic Ocean . Namibia is located within a swell wave environment, and experiences near constant, and often large, waves coming from the SW which have travelled huge distances across the Atlantic. Namibia’s tidal regime is categorised as microtidal, with tidal ranges at Walvis Bay in the order of 1 m. Mean annual tidal range is roughly equivalent to the mean significant wave heights, meaning that both wave and tidal processes are crucial to shaping the coastal environment. The Namibian coast also experiences the effects of the Benguela Current, a strong surface current which comes from the Southern Atlantic and brings cold waters up the coast to around the mouth of the Kunene River . It is driven by Southern Atlantic anticyclonic winds which are strongest in winter. Counter currents flow south closer to the shore, and at depth below the Benguela Current. Along the Namibia Coast there are also coastal areas of surface upwelling (the largest of which is at Lüderitz) where cold water ascends from depth. These upwelling areas support important fisheries. The Namibian coast is fronted by continental shelf—which extends some 100 km off most of the coastline , narrowing to 35 km wide north of the Walvis Ridge . Details of the coastal sediments are provided by Rogers and Rau (2006), and the distribution of the main coastal types is discussed by Harris et al. (2012), who note that the predominant coastal form is the sandy beach, making up about 68.5 % of the total. Other important coastal landscapes are rocky coasts, deltas and complex spit and lagoon systems.

The Skeleton Coast southwards from the Kunene mouth (which is obstructed by a bar), is backed by a series of major salt pans or coastal sabkhas, which have been mapped by Schneider and Genis (1992). Their origin is uncertain and deserves further attention. At Cape Fria low hills of Cretaceous basalt end in cliffs (Noli and Avery 1987). Between there and Möwe the shore is sandy with fringing rock reefs at False Cape Fria . Rocky Point is another prominent basalt headland, whereas Möwe is a low foreland of Damara metasediments. From Möwe to Palgrave Point the low lying coast is backed by scattered salt pans and small shifting dunes . Damara metasediments and Cretaceous basalts also outcrop. With varying success a series of rivers cut through the dune cordon and the beach, but some are ponded up (see Chap. 8). From Palgrave Point to Cape Cross there are alternations of rocky cliffs, again in Damaran metasediments and Cretaceous basalts, and there are many salt pans . Cape Cross itself projects out about 5 km into the Atlantic and is the site of a large, smelly and noisy seal colony. The coast from Cape Cross southwards is generally rocky until c 50 km north of Walvis Bay . There are marine terraces at c 2–17 m above storm tide level (Wieneke and Rust 1973) and some extending upwards to a maximum of 30 m (Davies 1973). At Swakopmund the mouth of the Swakop River has from time to time built outwards as a result of flood flows, which have, as we have already seen, also demolished the old bridge between Swakopmund and Walvis Bay . The next important river is the Kuiseb, which forms a delta just to the south of Walvis Bay (Huntley 1985). The town is an important lagoonal harbour, protected by the Walvis Peninsula and Pelican Point. Tides along this coastline are regular and semi-diurnal. There is a mean spring tide range of 1.42 m (0.27–1.69 m) and a mean neap tidal range of 0.62 m (0.67–1.29 m). These semi-diurnal tides flush Walvis Bay twice daily.

Buffetted as it is by large waves, strong winds, and rapid long-shore drift of sediment, the southern coast of Namibia is a highly dynamic environment with ever-changing spits and lagoons (Watson and Lemon 1985). One of these spit and lagoon systems is that at Walvis Bay . Strong longshore sediment transport from the south drives the Walvis Peninsula spit northwards at a rapid rate (Elfrink et al. 2003). The dominant wave direction is between 225° and 270° (Hughes et al. 1992) and wave data are given in detail in www.gecko.na/documents/ffd_o5a.pdf (accessed January 19th 2013). Analysis of old maps and photographs showed that its tip, Pelican Point, grew by an average of 17.4 m per year between 1885 and 1980. Between 1980 and 1996 Pelican Point prograded over a distance of 340 m, an average of 22.6 m per year (Schoonees et al. 1998). In all, the spit extended northwards by some 760 m between 1973 and 2010. From time to time, as in 1900 (Waldron 1900), 1959 and 2000, ephemeral mud islands have developed offshore and have been associated with gas eruptions (Emeis et al. 2004).

Box 2: The shifting sands of Sandwich Harbour

South of Walvis Bay is Sandwich Harbour (Fig. 1.15) (often called Sandvis and before that Porto D’Ilheo), which lies astride the Tropic of Capricorn. This convex sandplain is about 15 km long and protects a 9 km long lagoon, at the south end of which is a damp salt pan (Wilkinson et al. 1989). At the north end there is a highly unstable spit. In the nineteenth century Sandwich was a much used anchorage and port, visited by British and American whalers. Early maps of the harbour for 1880, 1889 and 1905 are presented in Schultze (1907). The first two indicate that the entrance to the harbour may have been 9 m deep, but the last shows that it had been displaced southwards and had shallowed to only c 3 m (Fig. 1.16). Bar growth and silting since the late nineteenth century mean that it is no longer usable (Kensley and Penrith 1977). Note the rapid changes that took place in just a few decades as recorded by the satellite images in Fig. 1.17. Sandwich Harbour may mark a former mouth of the Kuiseb, and the other headlands of this type may also be ancient river mouths produced by streams that formerly flowed across the Namib Sand Sea , but which now terminate in inland sumps like Sossus Vlei and Tsondab Vlei .

Fig. 1.15

Sandwich Harbour, August 2010

Fig. 1.16

Sandwich Harbour in 1905 (from Schultze 1907)

Fig. 1.17

A sequence of Landsat images from 1973 to 2000, showing changes in Sandwich Harbour (courtesy of Dr Frank Eckardt)

The third and fourth spit and lagoon systems are Conception and Meob Bays . At Conception Bay (Fig. 1.18) the degree of change is shown by the rusting remains of a German steamship, the Eduard Bohlen, which ran aground in 1909 (Harris et al. 2012). By 1973 the wreck was about 400 m inland from the shore. Here the spit has closed the lagoon completely to form a coastal pan, and it is possible that this will in due course be the ultimate destiny of the Sandwich and Walvis spits as well. Meob Bay has a small lagoonal pan at its northern end, but its main feature is a suite of ancient recurved spits at its southern end.

Fig. 1.18

Google Earth image of Conception Bay . Scale bar 2 km. Note the small enclosed pan at the northern end and the many old ridges at the southern end (© 2012 Digital Globe, TerraMetrics, Google)

From Walvis Bay to Lüderitz the coastline is backed by the Namib Sand Sea and large dunes often come right down to the shoreline. South of Lüderitz, where there are bays cut into basement gneisses and schists of pre-Damara age, is the so-called Diamond Coast . South of Elizabeth Bay are beautifully developed barchan dunes , while at Bogenfels there are cliffs and a classic arch, c 55 m high, developed in Gariep Complex dolomites . There is also a Holocene raised beach (Compton 2006). North of Chameis Bay the coastline is roughly linear and has many rocky headlands and small, north-facing sandy bays, which are known as ‘J-bays’ because of their shape. The rocks that form these J-bays and other south-facing re-entrant bays belong to the Gariep Group. From Chameis Bay to the Orange, the straightish coastline has been intensively mined for diamonds . The coastline of the southern Namib also has many small offshore islands . Finally, at the border with South Africa is the mouth of the Orange River , a source of sediment for the coastline to its north (Bluck et al. 2007).