Geomorphology of Desert Environments

Previous reviews have suggested that the rocircle of vegetation has often been given scant regard in the understanding of dryland geomorphology (Francis, 1994; Bullard, 1997). Bullard (1997) emphasized the landmark collections of papers in Viles (1988), Thornes (1990a) and Millington and Pye (1994) as reflecting a turning point in geomorphological perspectives, that is further emphasized by the 118 papers recorded in the ISI database since 1990 (but none before) up to mid-2007 which deal explicitly with the topic in some way. While it is untrue to suggest that work on the subject was not carried out before the 1980s – for example, Bryan (1928), Cooke and Reeves (1976), Hadley (1961), Huntington (1914), Melton (1965), Rempel (1936) and White (1969) – what has changed is the framework in which such research is carried out in dryland environments. This change is two-fold. First, geomorphologists have more explicitly recognized the need to incorporate a consideration of vegetation and, more broadly, ecosystems into their research designs. Secondly, ecologists have equally perceived the need for a more explicit evaluation of geomorphic and related hydrologic processes in order to be able to understand vegetation and ecosystem patterning.

Previous reviews of weathering in deserts (e.g. Cooke et al. 1993, Goudie 1997) have been excellent at identifying the mechanisms considered to operate and the landforms with which they are generally associated. Invariably, however, such reviews – especially if orientated towards students – deal primarily with perceived certainties. In reality, weathering studies continue to be characterized more by uncertainties and gaps in knowledge – especially in deserts. This chapter will therefore attempt to concentrate upon the ongoing development of ideas. The aim is not to be exhaustive or comprehensive, but by focusing on a limited number of underlying themes it hopefully questions some traditionally held views and could stimulate future research.

Pedogenic and geomorphic processes operating in deserts are inextricably linked. These linkages are particularly well expressed in the development of patterned ground and desert pavement. In addition, the nature and efficacy of hydraulic, gravitational, and aeolian processes on desert surfaces are strongly influenced by the physical and chemical characteristics of the underlying soils. As a result, the evolution of a diversity of desert landforms is either directly or indirectly linked to pedogenic processes.

Warm desert environments are characterized by the occurrence of a variety of surface and near-surface, chemically precipitated crusts, including calcrete, silcrete, and gypcrete, and their intergrades (Dixon 1994). Collectively, these are referred to as duricrusts (Woolnough 1930). This chapter focuses on the morphology, chemistry, mineralogy, and origin of the major types of crusts found in dryland environments. Particular attention is given to recent research from Australia, the United States, the Middle East and Africa.

Desert landforms are characterized by an abundance of ‘bare’ rock and mineral surfaces. Mountains host widespread exposures of bedrock. Gravel desert pavements cap alluvial terraces and fans. Even sand dunes are themselves composed of rock fragments exposed to the atmosphere without substantive plant cover. This chapter focuses on an irony, that the supposed fundamental bare-rock nature of desert landforms stretches the truth.

The very slow chemical and physical weathering rates in desert areas coupled with a relatively high efficiency of wash processes, due to the general sparseness of vegetation, results in more widespread occurrence of slopes with little or no regolith than in areas with humid climates. This chapter outlines the processes and landforms occurring on desert slopes that are either massive bedrock or are scarps and cuestas in layered rocks dominated by outcropping resistant rock layers.

Desert hillslopes below the angle of repose are dominated by the weathering characteristics of the underlying lithology, and specifically by the rate of production of fine material compared to the rate of removal. The previous chapter considered hillslopes underlain by massive rocks, or those in layered rocks dominated by outcropping resistant layers. These lithologies are weathering limited, and give rise to hillslopes where a surficial layer of weathered material is thin or absent. On more readily weathered lithologies a more-or-less continuous layer of debris is found. This layer of debris is subject to pedogenic processes. This chapter deals with such hillslopes.

Badlands have fascinated geomorphologists for the same reasons that they inhibit agricultural use: lack of vegetation, steep slopes, high drainage density, shallow to non-existent regolith, and rapid erosion rates. Badlands appear to offer in a miniature spatial scale and a shortened temporal scale many of the processes and landforms exhibited by more normal fluvial landscapes, including a variety of slope forms, bedrock or alluvium-floored rills and washes, and flat alluvial expanses similar to large-scale pediments. Although badlands evoke an arid image, they can develop in nearly any climate in soft sediments where vegetation is absent or disturbed.

Solar radiation, wind, and water are the driving agents of desert landscapes. Water has four major roles to play. First, in sustaining any life forms that exist; secondly, as a chemical substance which interacts with other chemical substances, notably salts; third, as a medium of transport of mass; and, fourth, as a direct source of energy. The last role, though small by comparison with the roles of solar and wind energy, may none the less be critical in determining the threshold of operation of runoff, through its impact on infiltration. Since infiltration is one of the major thresholds in dryland morphological development, the factors which control it have a role out of all proportion to the energy involved.

Dryland alluvial rivers vary considerably in character. In terms of processes, high energy, sediment-laden flash floods in upland rivers contrast dramatically with the low sediment loads and languid flows of their lowland counterparts while from a form perspective, the unstable wide, shallow and sandy braid plains of piedmont rivers are quite different from the relatively stable, narrow, deep and muddy channels of anastomosing systems (Nanson et al. 2002). It is also apparent that few, if any, morphological features are unique to dryland rivers. The variety of dryland river forms and the absence of a set of defining dryland river characteristics makes it difficult to generalise about dryland rivers and raises questions about whether it is necessary (or even desirable) to consider dryland river systems separately from those in other climatic zones.

Pediments, gently sloping erosional surfaces of low relief developed on bedrock, occur in a wide variety of lithologic, neotectonic, and climatic settings. Reported on six continents, their distribution spans the range of subpolar latitudes from the Arctic to the Antarctic and the range of climate from hyperarid to humid tropical (Whitaker 1979). On the Cape York Peninsula in tropical north-east Queensland, a tectonically quiescent region of Precambrian granitic and metamorphic rocks overlain by gently dipping Mesozoic and Cenozoic sediments, pediments are a dominant landscape component occurring as fringing piedmonts and strath valleys of the crystalline rocks and as broad erosional plains on the sedimentary rocks (Smart et al. 1980).

Alluvial fans are a conspicuous conical landform commonly developed where a channel emerges from a mountainous catchment to an adjoining valley (Figs. 14.1 and 14.2). Although present in perhaps all global climates, fans in deserts have been the most studied due to their excellent exposure and ease of access. Drew (1873), working in the upper reaches of the Indus River valley in the western Himalaya of India, provided the earliest illustrations and scientific description of desert alluvial fans (pp. 445–447):

Lakes and other mappable bodies of standing water exist at the atmosphere-lithosphere interface. Over shorter time intervals, water body configurations (hydrography) respond directly to atmospheric (hydroclimatic) forcing. Over longer intervals, hydrography also reflects tectonic and volcanic forcing from the lithosphere. Hydrographic patterns in lake basins, in turn, strongly influence and even control many geomorphic and stratigraphic patterns (e.g. Mabbutt 1977). These linked patterns (hydroclimatic + tectonic → hydrographic → geomorphic + stratigraphic) make lakes and kindred water bodies superb instruments for gauging environmental change and recording palaeoenvironmental history.

Hemiarid (‘half arid’) lake basins are drainage basins that have arid lowlands, nonarid highlands, and topographic and hydrologic closure. As a result of these characteristics, hemiarid lake basins contain or have contained nonoutlet lakes in their lowest reaches. A rich body of lacustrine and palaeolake evidence is stored in many hemiarid lake basins, providing the basis for reconstructing basin hydrography and the underlying hydrology, hydroclimate, and tectonics. Much of this lacustrine evidence occurs in regional and local patterns of geomorphology, sedimentology, and stratigraphy, which in lacustrine geoscience, with an emphasis on the depositional record, differ more in etymology than substance. Regional and local geomorphic patterns are equally important in effecting the hydrographic and related reconstructions from hemiarid lake basins. Regional geomorphic patterns constitute the predictive basis for making local geomorphic observations, and local patterns are the observational basis for building and testing regional models.

Aeolian processes, involving the entrainment, transport, and deposition of sediment by the wind, are important geomorphic processes operating in arid regions. This chapter, in association with Chapters 18, 19, and 20 form an integrated unit that discusses the fundamentals of aeolian sediment entrainment and transport, dune morphology and dynamics, wind erosion processes and aeolian landforms, and the significance of dust transport.

Sand dunes form part of a hierarchical self-organized system of aeolian bedforms which comprises: (i) wind ripples (spacing 0.1–1 m), (ii) individual simple dunes or superimposed dunes on mega dunes (also called draa or compound and complex dunes) (spacing 50–500 m), and (iii) mega dunes (spacing < 500 m). Most dunes occur in contiguous areas of aeolian deposits called ergs or sand seas (with an area of < 100 km²). Smaller areas of dunes are called dune fields. Major sand seas occur in the old world deserts of the Sahara, Arabia, central Asia, Australia, and southern Africa, where sand seas cover between 20 and 45% of the area classified as arid (Fig. 18.1). In North and South America there are no large sand seas, and dunes cover less than 1% of the arid zone. The majority of dunes are composed of quartz and feldspar grains of sand size, although dunes composed of gypsum, carbonate, and volcaniclastic sand, as well as clay pellets, also occur.

Aeolian erosion develops through two principal processes: deflation (removal of loosened material and its transport as fine grains in atmospheric suspension) and abrasion (mechanical wear of coherent material). In a vegetation-free environment, the relative significance of each of these processes is a function of surface material properties, the availability of abrasive particles, and climate. The resulting landforms include ventifacts, ridge and swale systems, yardangs, desert depressions (pans), and inverted relief. Dust is an important by-product of some forms of erosional activity.

In aeolian systems there is a fundamental difference between the behaviour of coarse sediments (sands) and fine sediments (dusts). Sand-sized material (63–2000 μm) travels predominantly by saltation, reptation and creep within the lowest levels of the atmospheric boundary layer (<3 m above the surface) and travels short distances. In contrast, dust-sized material, generally defined as <63 μm, is transported in suspension at a wide range of heights above the surface and can rapidly travel considerable distances. Dust plumes disperse as they travel away from source diffusing the concentration of sediment. Consequently, although there are clearly definable sources for dust emissions, these finer particles can be transported around the globe and their deposits can be both far removed from their origins and extensive. Every year up to three billion tonnes of dust are released into the atmosphere from the Earth’s surface. The spatial and temporal patterns of these dust emissions are often closely controlled by desert geomorphology, and in turn have an impact both directly and indirectly on the desert landscape and further afield.

The dusky brown to black coating of rock varnish dominates bare rock surfaces of many desert landforms (Oberlander, 1994). Thicknesses less than even 0.020 mm (or 20 micrometres, μ m) are enough to darken light-colored rock types (Fig. 21.1). The gradual pace of change on many desert landforms permits the slow accretion of rock varnish at rates of a few micrometres per thousand years (Dorn, 1998; Liu and Broecker, 2000). Just about any rock type will accumulate varnish, in so long as rock-surface erosion is slow enough to permit varnish accretion.

Geomorphic systems disclose great differences in their sensitivity to climatic change, and the various relief units carry the imprints of past processes to dissimilar degrees. Fluvial systems are highly susceptible to climatically induced changes in process. Hillslopes, on the other hand, are generally regarded as being rather resistant to such changes. In addition to the sensitivity of relief units to climatic change, another point of crucial geomorphic interest is how long the legacies of past processes are preserved in the form elements. Unfortunately, sensitivity to change and the length of time of preservation of past changes are usually inversely correlated. This means that we have either a detailed record of short-term changes for a limited period of time or a relatively inaccurate record of only major changes for a longer timespan. Where a detailed sedimentary record of past climatic changes has survived, related landform records may not be complete, because climatic changes are generally more frequent than landform changes.

Significant changes in the style of superimposed fluvial deposits have long been used by sedimentologists as an indication of broad changes in climate. Certainly, in traversing the globe from one environmental extreme to another – from humid to arid regions – it is possible to discern substantial differences in the character of rivers, both from the point of view of the forms that they adopt and the sediments that they carry (Schumm 1977; Wolman and Gerson 1978). However, rivers react to a number of large-scale stimuli, and it is often difficult to determine whether a change in character reflects tectonic or climatic influences, or both (Frostick and Reid 1989a). Attempts to attribute cause have inevitably and ingeniously simplified the setting by choosing systems in areas where most factors are presumed to have remained more-or-less constant whilst the putative controlling factor has varied monotonically. So, for example, location on a stable craton may allow an assessment of the impacts of climate change on river systems without the additional complications that arise from tectonic instability (Schumm 1968; Rust and Nanson 1986). However, while this approach may be useful in deducing the response of rivers to recent shifts in climate – say, those of the late Pleistocene and Holocene – and may be instructive in indicating the direction and magnitude of likely changes to be expected in river systems during periods of environmental change, the sedimentary legacies are often too similar to those assumed to have arisen from tectonic influences to ascribe changes in edimentary style confidently to either set of causes when stepping further back in time (Frostick et al. 1992).

Alluvial fans develop at the base of drainages where feeder channels release their solid load (Blair and McPherson, 2009; Leeder et al., 1998; Harvey et al., 2005). A classic fan-shape forms where there is a well-defined topographic apex. Multiple feeder channels, however, often blur the fan-shape resulting in a merged bajada. Alluvial fans can be found in almost all terrestrial settings. These include alpine (Beaudoin and King, 1994), humid tropical (Iriondo, 1994; Thomas, 2003), humid mid-latitude (Bettis, 2003; Mills, 2005), Mediterranean (Robustelli et al., 2005; Thorndrycraft and Benito, 2006), periglacial (Lehmkuhl and Haselein, 2000), and different paraglacial settings (Ballantyne, 2002). The geographical focus of this chapter, however, rests on alluvial fans in regions that are currently deserts or that experienced episodes of aridity in the Quaternary.

Lakes have long been recognized as being rich storehouses of environmental information. A lake basin collects water, but also sediment, much of which has been weathered and transported via fluvial processes from the near and far reaches of its drainage basin. The amount of water held in a lake is recorded on the landscape in coastal erosional and depositional landforms created at the water’s edge. The sediments deposited on the bottom of the lake can be clastic, geochemical, or biogenic, and include materials derived within the standing water body itself, such as through coastal erosion, chemical precipitation, or biogenic concentration, as well as those delivered to the lake from the surrounding drainage basin. In most cases only a small percentage of a lake’s sediment load is delivered from outside of the drainage basin as aeolian fallout. Because, under natural conditions, climate is the main determinant of the amount of water in a lake and because it influences some important characteristics of the lacustrine sediments and biota, changing climatic conditions are represented in the suites of abandoned shorelines and accumulations of sediments left by the lake over time (Fig. 25.1). This archival property makes the geomorphic and sedimentologic evidence of present and past lakes valuable as environmental and palaeoenvironmental indicators. Such evidence from late Quaternary palaeolakes, in fact, ranks among of the most complete and accessible sources of palaeoclimatic proxy data currently available for the late Pleistocene and Holocene.

During the late Quaternary, the world’s major deserts experienced dramatic changes in the nature and frequency of aeolian processes (Fig. 26.1). Sand seas (ergs) cover 5% of the global land surface and reveal evidence of repeated phases of dune formation (Thomas et al. 2005). This paper presents a review of dune-building episodes during late Quaternary time and their palaeoclimatic significance. The emphasis of the paper is on African and North American sand seas. Although beyond the scope of this paper, a more detailed synthesis and chronologies of global sand seas is presented by Tchakerian (1999), Goudie (2002), Munyikwa (2005) and Lancaster (2007).

Geoarchaeology is a particular focus of archaeology; some academics argue that it is a subdivision of the natural sciences (e.g. Rapp and Hill 1998); others suggest it is a combination of the sciences and humanities, using the same logic, principles and dating techniques (Huckleberry 2000). If amalgamation is the case, Gladfelter (1981) argues that geoarchaeology should be defined in a generalised and hence inclusive way. So according to Gladfelter (1981: p. 343), geoarchaeology ‘deals with earth history within the time frame of human history’ or more specifically applies the theory and methods used in Earth Sciences in the study of the human past (Gifford and Rapp 1985; Rapp and Hill 1998). Hence, geoarchaeologists are concerned with reconstructing past environments, using a range of different proxies, in areas of archaeological interest. Geoarchaeologists apply skills from the fields of geomorphology, pedology, sedimentology, hydrology, biology, geochemistry, remote sensing, geographical information systems and geology to study the evidence of past human-landscape interactions. Such studies involve a range of field and laboratory skills which include surveying techniques, the use of sedimentological, botanical, palaeontological, palaeogeomorphological, pedological, and geochemical techniques for palaeoenvironmental reconstructions, documentation of the site in terms of natural and cultural processes and activities, as well as developing chronological frameworks. Geoarchaeology therefore uses multi-disciplinary (Yerkes 2006) and multi-proxy approaches. As a consequence, geoarchaeologists do not have to be archaeologists, indeed much geoarchaeological research involves people from a variety of disciplines. This diversity of backgrounds has led to a wide assortment of definitions of geoarchaeology (e.g. see Butzer 1971, 1982; Dincauze 2000; Renfrew 1976, 1983; Taylor 2003; Yerkes 2006).

Deserts are superb repositories of geological, geomorphic and archaeological evidence. The very aridity to which they owe their existence has enabled them to preserve a remarkably good record of past depositional and erosional events. The fossil river valleys of the Sahara, the great salt lakes of Australia, China, and Patagonia, the dissected volcanic mountains of the Arabian peninsula and the Afar Desert – all are legacies of former tectonic, volcanic, and climatic episodes which ultimately gave rise to the deserts we see today. Each desert reflects its own individual geological inheritance and geomorphic history; each is unique in its assemblage of landforms; each ideally deserves detailed and separate study in its own right (Pesce 1968, McKee 1978, Rognon 1989).