No deserts are completely rainless although parts of the Libyan and Chilean deserts approach complete aridity. In such places erosion is extremely slow, although occasional showers can have sudden catastrophic effects. Trujillo in Peru received only 1.4 inches of rain between 1918 and 1925, but during March 1925 it received 15.5 inches, of which 8.9 inches fell in the three days 7-9 March. Such events apart, the present work of landscape development are controlled by the wind. (Outside the true deserts, vegetation prevents wind from being a significant agent or erosion), although it can carry enormous quantities of dust far beyond the deserts and can move sand and dust into characteristic depositional forms.
It is certain that wind cannot be responsible for most of the eroded landforms of present day deserts, and it is also certain that many of the landscape features of the African and Asian deserts were produced in times of wetter climate. In the Sahara, for instance, there are numerous old lakebeds, which have been dry for a great length of time. From mountains like the Ahaggar radiate systems of valleys which could have been cut only by running water, but which are now completely dry and choked by invading sand-dunes. There is considerable biological evidence of ‘relict faunas’ like tropical fish and small crocodiles in Saharan waterholes, and rock carvings indicating that big game of the savanna type once existed in areas now completely sterile. It is reasonable to conclude that many deserts are fossil landscapes formed under processes now not active. (All that is now happening at present is extremely slow weathering, and minor modification of the surface by wind).
Weathering in the very dry deserts is so slow as to be hardly perceptible. In its 3500 years in Egypt Cleopatra’s Needle, a stone obelisk, suffered no observable damage from weathering , but in 100 years in London’s atmosphere most of the surface has been severely rotted. (The main weathering processes at work in deserts were once thought to be mechanical disintegration of rocks by alternating expansion and contraction caused by frequent and violent temperature changes. It is probable, however that far more effective than this is the chemical weathering made possible by absorption by rocks of moisture in the form of dew, thus permitting a deeper rotting of rocks like granite, and the production of gravels and sand).
The most detailed studies of wind action were made by Bagnold (who concluded that wind itself has no direct erosive action, and that blown sand can seldom abrade rocks which are more than 18 inches above the ground and that sand, as opposed to dust, is seldom lifted more than 6 feet above ground level). The erosive effect of sand blasting is, therefore extremely limited. Sand blasting is thought to be responsible for differentially eroding rock surfaces so that resistant rocks are left standing as small undercut residual pedestals while the surrounding weaker rocks have been removed by sand blasting. Hollows left by wind vary in size although most seem to be rather small. Two of the largest hollows attributed to deflation by wind are the P’ang Kiang hollow in Mongolia which is 5 miles wide and 200 to 400 feet deep, and the Big Hollow in Wyoming, USA, which is 9 miles long, 3 miles wide, and 150 feet deep. It is probable that, although there is no general base-level of erosion for wind deflation, wind cannot be effective below the level of the water table so there is a local limit to the depth of deflation hollows.
The main action of wind is not erosive, but in transport and deposition. Wind picks up dry dust and can carry it high into the atmosphere. Nearly a quarter of a million square miles of northern-eastern China are covered with Loess, or wind-borne dust, which has been carried out of the Gobi Desert by winds associated with the winter high pressure system. It was deposited in areas of the increased rainfall and bound by the growing steppe grasses. The grains of loess are often cemented together with calcium carbonate, and being very porous loess preserves steep slopes in the face of many cliffs where cave habitations have been cut. The loess of much of North America, Europe, and New Zealand is derived not from a warm desert but from wind blown dust of bare river beds and the glacial moraines left at the end of the Ice Age. Loess is always very fertile and is extremely important to agriculture in much of Asia, North America, Europe, and the South Island of New Zealand.
Sand may be moved by surface creep or by saltation. Saltation, or the hopping of grains of sand is caused by turbulent air lifting the sand almost vertically and then moving it forwards. As it lands the sand grain may cause other grains to move and be lifted. The smallest features caused by saltation are ripples but these are of little geomorphic significance. Larger accumulations vary from flat sheets like those of the Libyan Desert, 3000 square miles in extent, to irregular accumulations greatly varying in size, of which the dunes are the best-known types. The most common dune is the barchan or crescent shaped dune with ‘wings’ pointing away from the direction of the wind. Barchans never develop on perfectly flat sand sheets but where there is some slight irregularity on which sand can accumulate. The barchan is not stationary but migrates in the direction of the dominant wind. The sand grains on the windward side move up the back of the dune and fall over the crest onto the leeward side, which is kept steep by wind, eddies.
Small barchans may be as little as 12 feet high and large ones as much as 100 feet high. In all cases barchans are approximately ten times as wide as they are high. The largest dunes are stationary but as size decreases the speed with which they travel increases. Large groups of barchans are often found so close together that they lose some of their perfection of form and they may then be called transverse dunes.
The other type of dune is the seif dune. Seif dunes are long sand ridges of irregular shape and variable length. Many seifs in the Egyptian desert are up to 300 feet and 1800 feet wide; some are over 200 miles long. While these figures may represent the largest known, they do, at least indicate that seif dunes are very large features, especially when compared with the barchans. Seifs seem to be developed where there are dominant winds from a different direction from the prevailing ones, which have accumulated the sand. The figure below, shows the development of a barchan into a small seif by the action of wind from two directions.
Both seifs and barchans tend to form in families, but the two types are seldom found in association with each other.
In contrast with mobile barchans and seifs Bagnold recognised several other types of sand accumulation all of which are stationary. Whalebacks, or flat topped sand ridges extending parallel to the wind direction, occur particularly in Egypt where they may be 100 miles long, 2 miles wide and 150 feet high. Bagnold regarded whalebacks as remnants of seif dunes. Much smaller features are the sand-shadows, which accumulate in the lee of obstructions and are built up by wind is funnelled round them, as at the exits of Wadis. Both shadows and drifts are destroyed if they move downwind (see below).