en zh es ja ko pt

Volume 40, Number 3May/June 1989

In This Issue

Back to Table of Contents

Lakes of the Rub' al-Khali

Written by Arthur Clark
Illustrated by Michael Grimsdale

The lakes of the Rub' al-Khali may be nature's ultimate mirage. That they once slaked the thirst of man and beast - including hippopotamus, water buffalo and long-horned cattle - seems certain. The long-ago presence of hippos is attested by finds of their fossilized teeth, so pristine they might have been lost just yesterday. The fossilized bones of water buffalo and long-horned cattle, as well as of wild asses, wild goats or sheep, oryx, gazelle, and possibly camels and hartebeest, have all also been found in the petrified lake mud. Clam shells are in evidence, too, hundreds of kilometers from the nearest modern coast. Chipped-stone tools are scattered in the vicinity.

Today, giant dunes as high as 250 meters (800 feet), not pools of water, are the hallmark of the world's largest continuous body of sand. The Rub' al-Khali, or Empty Quarter, covers some 650,000 square kilometers (250,000 square miles), more than the combined land areas of Holland, Belgium and France.

All that remains of the lakes are their eroded basins, standing up in forlorn little buttes a meter or so high, in stark white or gray against the Rub' al-Khali's infinity of rusty-red sand. A visitor new to the region might miss them entirely. But retired Aramco geologist Hal McClure says, "Youget to where you can spot them immediately. They break the tan monotony of the terrain."

McClure's 1984 doctoral thesis at London University reported in detail on the lakes, filling a blank in the world's ancient climate map. Strikingly, the new date for the virtually trackless Rub' al-Khali showed broad matches with the climate histories of a whole belt of deserts that includes the Simpson in Australia, the Rajasthan in India and the Kalahari and Sahara in Africa.

The great Rub' al-Khali desert began to form more than two million years ago, according to McClure, while the lakes were formed just yesterday in geological terms. In fact, evidence indicates that lakes formed twice: once from roughly 37,000 to 17,000 years ago, and then again from around 10,000 to 5,000 years ago. In the interim, "hyper-arid," period, as today, rain was very rare.

McClure theorizes that the lakes were created by cataclysmic rainfall, like that seen in the summer monsoons which today water the Indian subcontinent and, on the Arabian Peninsula, extreme southern Oman. He speculates that the summer monsoon moved to the north twice in recent geological history, most likely creating lakes in what he calls "one-time fill-up incidents." "It would rain like hell one monsoon season and then not rain in a particular area for the next 10 or 100 years," he says. The lakes had no links with rivers, above or below ground, or any other source of continuous replenishment, and their bed sediments present no evidence of regular refilling.

Rub' al-Khali lakes "weren't enormous lakes like in East Africa or like Lake Superior," explains McClure. They probably ranged in depth from two to 10 meters (six to 32 feet), he says, though some were only "ephemeral puddles." A few may have lasted several years, but most existed only "a few months to a few years."

The monsoon's movement to the north, which is also strongly suggested by the climate histories of the deserts in Africa, India and Australia, may have been due to a slight wobbling of the earth in its orbit around the sun. In a 1981 study, John E. Kutzbach of the University of Wisconsin's Center for Climatic Research argued that such shifts would have resulted in hotter summers in the northern hemisphere, in turn creating a low-pressure corridor through the Arabian Gulf and present-day Iraq which provided a passageway for the inrush of moisture-bearing air from farther south.

"Evidence is accumulating that variations of the earth's orbital parameters ... played a major role [in the climate fluctuations] of the past several hundred thousand years," he wrote.

McClure admits that the idea that lakes dotted the Empty Quarter in recent prehistory "might have been met with skepticism" as late as the 1970's. But there were "a thousand or more" lakes, claims the geologist, who followed up clues uncovered by explorer Harry St. John Philby in the heart of the desert more than 50 years ago, along with later discoveries there by Aramco prospectors, to develop his data about the area's ancient environment.

It was Philby, an Englishman, who first reported freshwater shells, along with late Stone Age implements, at a number of isolated spots in the desert. His objective in traversing the Empty Quarter in 1932 was to find Wabar (See Aramco World, November-December 1986), the legendary city he thought might be the seat of an ancient civilization comparable with those of Egypt and Mesopotamia. He was disappointed in his search for Wabar, but delighted with the shells and stone tools.

"Apart from Wabar itself this perhaps the most interesting and dramatic discovery of the whole journey," he wrote in his book The Empty Quarter. The unexpected finds "set me off imagining ancient man in occupation of the banks of an old river or lake, or visiting them to hunt with spears and arrows the beasts that came down to drink."

He continued, on a slightly wistful note, "The onset of desert conditions in the Empty Quarter must have developed contemporaneously with the early beginnings of serious civilization in such countries as Egypt and Mesopotamia. Such a civilization might indeed have made a beginning in these parts, only to be stifled by Nature's hostility...."

Philby's belief that Neolithic man hunted at the water's edge was probably on target, but lacking today's more sophisticated research tools "he didn't fully grasp [the lakes'] significance or extent as actual lakes," says McClure.

In the 1950's, Aramco geologists mapping the Empty Quarter also noted the existence of lake beds, but did not study them in detail. McClure's own discoveries came in 20 years of exploring in the Rub' al-Khali as a member of drilling and geological parties, and during research trips on his own time through the dunes. (See story, page 34.) Today, he is pursuing his research in affiliation with the British Museum (Natural History), London.

The lake beds he charted "are distributed down the middle length of the Rub' al-Khali" - a distance of some 1,200 kilometers (750 miles). What remains of them are marl terraces and patches of hardened crust a few meters wide, distributed in thin, "shoestring" forms, a kilometer or more long, between the dunes. Now, however, instead of lying below the level of the desert, the lake beds stand up in relief, the sand that once surrounded them having been blown away by thousands of years of scouring winds.

McClure's conclusion that the Rub' al-Khali was freckled with ponds is based on the discovery of hundreds of lake basins; for his thesis, he picked out 19 in the southwestern and north-central parts of the desert for detailed stratigraphic examination. Radiocarbon dating was used to calculate the ages of mollusk shells and lake-bed marl, and thus determine the age of the lakes themselves.

McClure believes lakes formed between the dunes and in hollows on top of them when torrential rainfall spilled down dune slopes. The runoff brought with it clay and silt particles from the dune sides, and clay or silt pans formed where the water pooled. The pans impeded water penetration down into the sand and held any new precipitation; the result was a lake.

All the lakes that have been dated to the initial period - between 37,000 and 17,000 years ago - have been discovered between present-day dunes, meaning that, despite winds working on their surfaces, the sand mountains probably have undergone very little net movement in thousands of years. And - unlike today's high, sharply crested dunes - they were probably uniformly longitudinal and rounded, so that any water pooled between them.

By contrast, many basins dated to the second lake-forming period - between 10,000 and 5,000 years ago - are perched high up in the dunes. In general, the later lakes appear to have been smaller and shorter-lived; McClure blames the hot, dry intermediate weather for the change.

He suggests that extremely arid winds between 17,000 and 10,000 years ago whipped the dunes into "chopped-up and crested" shapes similar to those of today. As a result, catchment areas shrank and, when the rains returned, water pooled in hollows on the dunes themselves.

The geologist believes that "light to medium rainfall" was probably distributed generally over the Rub' al-Khali during both lake periods. This allowed "a grassland terrain to develop very much all over, not just limited to the sand slopes around lake basins." Such a weather regime would have been ideal for the mammals represented by fossils that have been discovered in the lake beds.

Despite finds of hippo teeth in lake beds in the southwestern part of the Empty Quarter, dated to the first lake period, McClure cautions against imagining that large herds of the beasts "splashed around in lush lakes throughout the Rub' al-Khali." He thinks the hippos, along with water buffalo, probably moved into the region from a lusher, bordering habitat, put died out when the lakes dried up.

Cattle, which require less standing water than hippos and water buffalo, are represented by fossil discoveries in lake beds from both periods. "They very likely ranged throughout the Rub' al-Khali" during the two periods, writes McClure. He adds that rock carvings in western Saudi Arabia depict domesticated animals with long, curved horns; these may well be renditions of cattle that descended from the wild Bos primigenius that fed on the lakeside grasslands.

Pools that existed in the Mundafan area, located on the slopes of the Tuwaiq Escarpment in the southwestern Rub' al-Khali, seem to be the exception to the rule of short-lived lakes. Dates determined for the lowest and uppermost sediment levels of a major lake bed there show an estimated 800-year span for the latest lake period. Mundafan pools had a much larger catchment area than their interdune cousins, and were fed by runoff from the escarpment - an important geological feature - as well as from surrounding dunes, says McClure, probably accounting for their longevity.

Clam and snail shells, and remains of plants, offer another perspective on the fossil lakes.

The presence of grass, shrubs and herbs around the lakes may be indicated by what McClure calls "tubule scree." These thin, tubular fragments of knobby, rock-like material are found scattered in the sand around the lake basins. They were likely formed when calcium carbonate dissolved in the lake water precipitated onto the stems or roots of plants as the lake gradually evaporated.

Micron-sized quartz or silica bodies called opal phytoliths, which are constituents of all plants and which give grass blades their sharp edges, for example, have also turned up in large numbers in samples from the lake beds. "Wherever you get a lot of phytoliths, it is very likely there were a lot of grasses," says McClure. The fact that "a lot of grass grazers [are represented by fossils found] in the vicinity makes it even more likely that a lot of grass was present."

The scene at the ancient lakes was much like that around man-made desert ponds today (See Ammco World, November-December 1986). The evidence shows they were probably fringed with the tall, tufted reed Phragmites, cattails and "the odd tamarisk tree," the geologist says.

The lakes were also home to abundant algae, including Chora species, some of which are commonly used to decorate fishbowls today. Calcium carbonate from the lake also precipitated out around the algae, and when a mat of algae died "a pile of calcium carbonate was deposited" on the lake floor, notes McClure. Together with clay, it formed marl, a soft, impure form of limestone.

Lake beds of pure calcium carbonate and gypsum stand out stark white in the desert. There are also beds of clay, of sand cemented by calcium carbonate, and of clay-cemented sand.

Because most lakes were rarely refilled, they became too alkaline too quickly to develop fish populations. Only one surveyed by McClure had any evidence of fish life. Snails and tiny bivalve crustaceans called ostracods are the most common fossils, probably "seeded" by birds or by animals moving from lake to lake.

Colonies of freshwater clams also flourished so long as the lakes had sandy bottoms and were not too alkaline; their shells are always found in the sandy levels of lake beds, beneath later marl buildups. Shells up to six centimeters long (2.3 inches) have been collected.

Notably, while the early lakes lie on white sand, lakes dated to the second phase all sit on foundations of rust-colored sand, indicating that the famous reddish color of the Rub' al-Khali dates from between the two periods of lake formation.

"Despite intense aridity, there was enough moisture, in the form of very light rainfall or dew, to leach iron from minerals contained within the sand," explains McClure. "Hyper-aridity then evaporated the moisture and caused the deposition of the iron in the form of iron oxide."

Man in the Rub' al-Khali is more of an enigma. Like Philby McClure thinks that Neolithic man hunted on lake shores.

On the basis of his discovery of stone tools and a thin layer of ash at an important lake site in the Mundafan region, the area with the greatest extent of lakes, McClure believes prehistoric man may have burned the reeds in and around the water to flush game. "I very much suspect fires were deliberately set to run birds or animals," he says.

No fossil remains of man, "absolutely none," have been found at the lake sites, says McClure, but finds by early modern explorers of flint tools, including knives, scrapers, borers and arrowheads, place humans at the lakes in the period from 10,000 to 5,000 years ago. Much earlier evidence of manjn the shape of stone hand axes dating back 100,000 years or more, has been collected on the fringes of the Empty Quarter, but these tools are not associated with lakes. No stone implements were discovered in association with lakes of the earlier period documented by McClure, meaning that man was apparently not present during that time.

"Human populations flooded into this part of the world about 10,000 years ago, a fact generally recognized by pre-historians - but from where they came no one knows," says the geologist.

No one is certain, either, where the humans went when the rains petered out. But McClure is willing to bet that they migrated to the north. There, he thinks, possibly in what today is northern Iraq, Rub' al-Khali man provided the base for the Semitic population of the modern Middle East.

Could the monsoons return in our time, to dot the Empty Quarter with reedy, living lakes once again?

McClure thinks so.

In July 1977, he notes, three weeks of "freak" monsoon-like rainfall occurred in the northeastern Rub' al-Khali. No dune-slope runoff or ponding resulted at the time, "but it's intriguing to think that should the same thing happen this summer, or next summer - and be intense enough to again form lakes - it could herald a genuine return of the monsoon to the Rub' al-Khali," he says.

Such apparently infrequent occurrences of heavy rain might not seem significant, or sufficient to tip the climatic balance. But they could, in fact - as some experts have said of the 1988 drought in America -herald a true climatic shift.

"In our short human time scale," notes McClure, "we really can't see the most momentous changes in the weather cycles."

Arthur Clark is an Aramco writer based in Dhahran and a frequent contributor to Aramco World.

This article appeared on pages 28-33 of the May/June 1989 print edition of Saudi Aramco World.


Check the Public Affairs Digital Image Archive for May/June 1989 images.