UAE Sabhka



Introduction to coastal sabkhas
The mainland "sabkha" coastal plain was formed from a combination of Holocene accretion of supratidal, intertidal and lagoonal sediments and an earlier accumulation of the underlying sediment wedge from the Late Pleistocene. This mix of evaporite and carbonate facies has been interpreted by Kirkham (1997) to have been initiated during the Flandarian transgression. He suggests that at this time these sediments and the associated evaporite minerals accumulated as a series of discontinuous coastal lagoons just to the south and leaward of the cerithid rich beach ridges that parallel the modern shore. Locally some discontinuous cyanobacterial mats formed to the lea of these ridges though the commoner sediment sediments are carbonate and quartz sands.

Evans et al. (1964a, b) explained how this mainland sabkha stretches "from Ras Ghanada almost to the Qatar peninsula in the west, a total distance of almost 320 kilometers." The width of the sabkha paralleling the Khor al Bazam varies, reaching a width of as much as 32 km just south west of Abu Dhabi Island. It flanks intertidal flats and coastal terraces except where locally hills of Tertiary and Quaternary rocks jut out as peninsulas. In places the sabkha surface lies flush with eroded Quaternary rocks.

The surface of the sabkha takes several forms:

  • Old beach ridges
  • Sandy material between and to the lea of beach ridges
  • Salt encrusted algal flat
  • Salt encrusted sandy late Holocene lagoonal and tidal sands
  • Outwash fans extending out from Quaternary to Tertiary hills.

The most conspicuous features of the sabkha are the old beach ridges which can be identified by their distinctive appearance in the field, or directly from aerial photographs. Like "cheniers," they have linear shapes with a smooth seaward margin and an irregular landward outline (Byrne et al., 1959). The beach ridges normally consist of well sorted, coarse skeletal sands (often rich in cerithid gasteropods), and mark various stages of seaward accretion of the sabkha in the form of intertidal spits. They drape headlands (for example, Ras al Aish) and cross embayments (for example Al Mirfa and Khusaifa). It is likely that some of the beach ridges represent barrier beaches that formed seaward of now infilled lagoons on the lee of offshore banks. This beach ridge belt is some 2 3 km wide. It is thought to represent a shoreline formed some 4000 years B.P. (Kirkham, 1997) when the Khor al Bazam was more open and deeper. South of these ridges the sediment surface is very sandy and probably represents the final infill of lagoons. However, cross sections cut in some of the sandy sediment revealed cross bedding very similar to that shown by much of the miliolite, thus indicating that they may be of Quaternary age. Other sediments of the sandy surface show crude horizontal laminae, with well sorted horizons of foraminifera rich sand that probably accumulated during storm flooding. These beds are commonly contorted in response to the crystallization of halite or the rotting of sargassum like seaweed washed inland with the foraminifera.


Qanatir Sabkha Traverse
One of the most convenient places for the geologist visiting the UAE to view the coastal sabkha evaporites is just a short distance east of the western margin of the Dhabaiya peninnusula on the so called "Al Qanatir" and/or "Rafiq" traverse. Al Qanatir is a barrier island marking the eastern end of the Khor Al Bazam. Here, along a road just to the south of this island a traverse can be made from the earliest Holocene beach ridges to the present shoreline.

You can reach this by driving west from Abu Dhabi on the main road to Jebel Dhanna. After passing the sign to "Dhabiya" (note phonetic spelling) look for a raised dirt road that rises above the sabkha and takes off north over it. This road breaks the palm tree hedge to the right of the main road. This line of date palms screens the road from the sabkha. The Qantir or Rafiq road can be identified by the power line that flanks it and in the distance close to this line a wooden raised guard tower is visible. The turn is marked by a non descript sign in arabic to Rafiq (you can see this on a linked photo). After passing the Dhabiya turn drive slowly along this section of main road or you may miss the turn onto the dirt road to the right. Watch out for the large trucks! If you miss this turn you will be forced to cross the median and probably return to the Dhabiaya turn and come back again.


At Qanatir where the upper supratidal zone has been trenched, the vertical shoaling upward sediments reflect the prograding character of the sabkha seaward. Notable exceptions, however, do occur where the slope of the sabkha adjacent to beach ridges is disrupted by tidal channel depressions and by elevated hard ground tepee structures. Traced from seaward to landward (North to South) the surface sediments on the Qanatir traverse are:

  • Lower intertidal to shallow lagoon carbonate
  • Lower intertidal flat hard or firm grounds cemented by carbonates into megapolygons
  • Series of distinct cyanobacterial mat surfaces (upper intertidal),
    moist carbonate mud, and gypsum mush on the lower salt flat (lower supratidal)
  • halite crust on the lower salt flat (mid supratidal)
  • Stranded beach ridges rich in cerithids
  • Thin (3 mm) polygonally cracked halite surface on the upper salt flat (upper supratidal)

As with the Abu Dhabi sabkha (now beneath an urbanized cover) this variation in surface sediment type and morphology has been determined to reflect the local topography and a relationship to the tidal range (Butler, 1969; Butler et al., 1982; Kendall et al., 1998; and Kirkham, 1998).


Shallow subtidal to lower Intertidal sand and muddy sediment
At Qantir seaward of the cyanobacterial mats subtidal to intertidal sand and mud flats are composed of foraminiferal gastropod sands and/or muds that represent the shoaling eastern termination of the Khor al Bazam. These flats locally may have a gradient of approximately 1:5,000 and vary in width from 0.5 km to 8 km. In the subtidal they may have a varying thicknesses of unconsolidated sand and mud which in turn may be underlain and locally overlain by alternating layers of poorly cemented limestone (1 to 3 cm thick) and unconsolidated carbonate sand (3 to 6 cm thick). The cemented layers have flat upper surfaces and irregular lower surfaces. Some are of different composition to the loose sand above them. These particular horizons often extend from 2 m below the low water mark where in the eastern Khor al Bazam just to the west of the traverse they form megapolygons tens of meters in diameter (Assereto and Kendall, 1977) just below the low water mark. These surface sediments are probably cemented by calcium carbonate precipitated by blue green algae and bacteria (Nesteroff 1956).

The surface of the subtidal to intertidal sand flats are ripple marked and/or covered by growing seaweed. There are often subtidal spits that may be fixed in position by subtidal to intertidal beach rock cementation. Local break point bars, and runnels and bars occur, all of which are controlled in their distribution by the direction of wave approach. The landward edge of the cyanobacterial mats is sometimes delimited by intertidal spits (some built by by huge death assemblages), beaches or occasional mangrove clumps. These eventually pass into intertidal cemented hardground and then mid to high intertidal cyanobacterial mats.

The surface may be divided on the basis of fauna into Cerithium flats on the lower area and crab flats on the upper (Kinsman, 1964a). Scopimera sp., a crab that characterizes the latter zone, produces radial patterns of feeding balls. The sediments of the flats are extensively burrowed by crabs, worms, and molluscs, so the primary are destroyed. Ginsburg (1957) reported the same phenomenon in carbonate sediments of Florida.

Hardground & Below

Lower Intertidal Sand/Mud Flats and Hardgrounds
In the more elevated portions of the lower intertial flats, just seaward of and close to the cyanobacterial flats, the sediment is composed of carbonate sands where evaporation of capillary water during low tide (Ginsburg 1953b) appears to be driving cementation of the surface sediments. This latter process is probably especially active during the height of summer. Before additional sediment accumulates, the upper surface of the beachrock is often truncated by storm action. Sugden (1963) and Purser and Loreau (1973) observed similar crusts at the western end of Abu Dhabi.These layers of contemporaneous beachrock are forming extensive laterally continuous hardgrounds from the unconsolidated sediment of the sand flat. Often these cemented layers are buckled into megapolygons 2 to 3 m in diameter. The edges of these megapolygons protrude above the adjacent sands and are coated by a blackened surface of cyanobacteria. These crumpled margins are filled by several generations of sediment and cement. We believe the initial cracking is the product of thermal expansion, but subsequent fracturing and fill are related to generations of cementation and sedimentation within the cracks, forcing further expansion.

Local dissolution of the hardground surface appears to be driven by infestation of cyanobacteria. The best examples occur where all the hardground but crumpled margins have dissolved. Commonly occurring in association with the hardgrounds are ubiquitous colonies of mitolids and cerithids. Additionally, occasional single mangroves colonize small areas of the sand and mud flats. Seaward of the cyanobacterial flat and hardgrounds is the undulating surface of carbonates sands and muds described in the previous section. Locally, in the lower lying areas, tidal creeks dissect this surface and are incised into carbonate mud. Along the tidal channels mud cracks are common.


These hardgrounds are discontinuous but extend landward below the prograding cynobacterial mats and sabkha. Thus trenches in the landward sabkha show the hardgrounds capped by either "peat like" layers of cyano bacterial flat sediment adjacent to open lagoons or are capped by intertidal burrowed muds adjacent to the protected lagoons. Both these peats and burrowed muds are capped landward by supratidal evaporites.

Beneath the sabkha the hardgrounds are cemented by a calcium carbonate and/or locally by later gypsum. This cemented layer tends to form a seal between the underlying more marine sediments with their marine waters and the overlying supratidal salt flats with their saline brines. This hardground is not present everywhere and seldom overlies the algal peat layer. This latter occurs where the stranded shoals and coastal spits lie almost perpendicular to the shore and are elevated above the adjacent tidal flats. Here the hardground extends above the peat layer to be on lapped by a layer of gypsum mush with a surface of algal mat.

Cyanobacterial Flat Facies
As indicated earlier, the inner shelf is capped by islands that protect coastal lagoons enhancing the development of wide tidal flats whose landward margins are rimmed by cyanobacterial flats. Traced from the sea to the ancient beach sediments to the present shoreline, the surface sediment varies from a series of distinct cyanobacterial mat surface morphologies, moist carbonate mud gypsum mush, convolute halite crust to a thin polygonally cracked halite surface. These variations in surface sediment type and morphology reflect the topography and its relationship to tidal range levels. The slope of the flats and adjacent salt flats or sabkha are interupted by beach ridges, tidal channel depressions and by elevated hardground teeped structures.

One of the areas where extensive flats of laminated algal mat are forming is on the protected intertidal and supratidal flats that flank the Khor al Bazam (Kendall and Skipwith, 1966) and these are well displayed on a north south traverse at Qanatir. These cyanobacterial mat accumulations have an average width of approximately 2 km and can reach a thickness of at least 30 cm. At the east end of the lagoon (to the lea of Abu Al Abyad), the largest flat is 42 km long, while to the west, another flat (at Khusaifa) is 9 km long. In some areas the flats extend landward in the subsurface more than 2 km beneath a thin cover of evaporites and wind blown sediments. Smaller flats occur in the shelter of islands, headlands and swash bars.

The algal mats (or microbial mats, Golubic 1991) of Abu Dhabi have been described by and subdivided into groups based upon gross surface morphology of the mats (e.g. Kendall and Skipwith, 1968; Golubic, 1991). The morphologic variations represent differences in microbial species which form the mats, or modifications due to environmental conditions, primarily a result of the relative amounts of exposure to air. Thus, mat types can be directly correlated to the topographic level of the flat. Several types and modifications are evident in the Al Qanatir samples.

The larger algal flats are divided by surface morphology into four geographical belts. From the low water mark moving landward these are:

  • Cinder Zone:
    The mamillate (Golubic, 1991) or cinder mat (Kendall and Skipwith, 1968) inhabits the lowermost intertidal zone and is the first mat type in the vertical and lateral sequence of mat types in the mat facies. The warty black algal surface of this zone, the color and size of the raised bumps resemble a weathered volcanic cinder layer. These bumps are shaped like small pustules, 2 to 3 cm in diameter and contain an unlaminated algal and sediment peat.
  • Polygonal zone:
    In the mid intertidal zone, a smooth, continuous, and generally polygonally desiccated, flat mat occupies several sub levels and displays several morphologic modifications. This mat is layered and commonly has a beige to pink surface color although can be dark greenish black. Within tidal creeks the flat mat accumulates to thicknesses of 20 cm and desiccates at the surface into large (1 meter) polygons. Polygonal zone algal mat separated into desiccation polygons a few cm to 2 m in diameter which cover laminated algal peat. Sediment fills the cracks between the polygons.The edges of the polygons upturn, and on a small scale, provide a niche for pinnacles of algal mat. The pinnacles of the pinnacle mat are also common in areas with low tide exposure and occur with several environmental modifications (e.g. ripples caused by waves) near the shore.
  • Crinkle zone:
    This zone is chararacterized by a leathery algal surface forming a blistered skin forms in the upper intertidal zone, a setting flooded during the middle to high tide. This dominant cover of crenulated or crinkled mat (Kendall and Skipwith, 1968), or convoluted mat (Golubic 1991) extends over a layer or mush of small gypsum crystals (.5 mm diam or so) mixed with a carbonate mud. The crenulated mat forms a leathery, wrinkled mat, which is black on air exposed upper surfaces of the folds, but commonly retains the pinkish beige color on the lower less exposed surfaces. The folds trap air and gases which are expelled when walked upon. A common morphological variation to the texture of these mats includes the pinnacled "tufts" (seen in other settings seaward) on the upper more aerated surface of the crenulated mat (usually in the mid intertidal zone) along transitions between the polygonal mats with the crenulated mat. In the upper most intertidal zone, the crenulated mat can locally completely desiccate and shrivel to a dried crust during low tide. There is often no significant accumulation of microbial material in this zone.
  • Flat zone:
    This setting is chararacterized by firm, smooth algal mat with no topographic relief, overlying quartz rich carbonate sand and gypsum mush. Smaller algal flats have one or more of these zonal belts.

The algal growth and structures appear to be determined by the frequency and duration of subaerial exposure and the salinity of the tidal waters. They are only related to wave energy where they are limited by wave and tidal scour at the edge of the Cinder zone and along ebb channels.

Preservation of algal facies
Field observations indicate that as the mat becomes more deeply buried further inland, gross morphologic structures are less well preserved. For example, below the gypsum mush facies polygonal and mamillated mats are easily distinguishable in buried sections; whereas beneath the lower supratidal facies (further inland) only the polygonal mat was identified by field inspection. The crenulated algal mat subfacies was not evident in the subsurface at all. This mat may not be preserved because 1) it forms more inland in a higher and more oxidizing depositional setting and/or 2) it forms in a setting conducive to precipitation of post depositional, displacive gypsum.

Chemical Results
These analyses are for mats collected by our students and colleagues (Kendall, Alsharhan, & Cohen, 2002) within the Qantatir area and are presented here for comparison to studies by Cardoso et al. (1978) and Kenig et al. (1990) in other areas of Abu Dhabi. Results of the geochemical analyses for the mats studied petrographically in this paper will be discussed in a subsequent paper.

Total organic carbon (TOC) of the sediment in the study area shows a wide range of organic enrichment (0.46 8.40% TOC). Whole rock pyrolysis yielded moderately high Hydrogen Indices (Hl) of 389 597, which are typical of marine Type 11 kerogens. These values vary slightly from the values ( 0.5 2.7% TOC and Hl 510 to 675) published by Kenig et al. (1990).

Elemental composition (C, H, N, 0, S) of isolated solid organic material (kerogen precursors) showed atomic hydrogen to carbon (H/C) ratios (1.20 1.54) that fall directly on the Type 11 evolutionary pathway as shown in a modified van Krevelen diagram (Tissot and Welte, 1984) Stable carbon isotope ratios ranged from 8.41 to 10.78 %.

Previous studies of Abu Dhabi algal mats assessed relative "preservation potential" of mat types based on the preservation of gross structures (Park, 1977), the subsurface extent (thickness and distribution) of the preserved mat, and the assumption that the mat will not be altered physically beyond easy recognition of the gross structures. The first part of this assessment may not be correct since the distribution and thicknesses of buried mat types cannot be unequivocally assumed based on the present extent of living equivalents. For example, if conditions 500 years ago were not conducive to mamillate mat growth in this area, it cannot be assumed that lack of mamillate mat in sediment 500 year old means that the mat was not preserved.

The second assumption is that the buried mat should "look like" it's living counterpart on the surface. In this regrad, geometric associations of the mat with inorganic sediment, such as the well formed laminae and the desiccation polygons of the flat mat, are easily recognized; thus the flat mat can, without a doubt, be recognized in the subsurface. The mamillate mat, on the other hand, does not have such distinctive features and, additionally, is prone to compression. Consequently, the "occurrence", or rather recognition, of mamillate mat in the subsurface may not be as common as that of flat mat, but this is a function of the initial extent, condition, and composition of the mat, and of the criteria used for identification, not a result of poor preservation potential.

Petrographic analysis provides an alternate means of recognizing mat types in the subsurface. Additionally, study of microstructures can delineate subtle diagenetic changes in the mat as well as the degree of preservation of the cyanobacterial constituents of the mat. Thus, the following observations of cellular and amorphous organic components of the mats are offered:

  1. Coccoid species (or at least the cell walls of the these organisms), are relatively well preserved regardless of the mat type in which they are found. Literature citing fossil examples of cyanobacteria emphasize this point, since many of the examples are coccoid species similar (or perhaps identical) to those presented in this study (Golubic and Barghoom, 1977; Awramik, 1984; Golubic and Yun, 1985).
  2. The amorphous organic material (sheath material and probably cell fluids) may degrade in a manner that destroys depositional microstructure. If or when this material is preserved, it will probably be preserved in an amorphous form with origins difficult or impossible to decipher.
  3. The dissolution of inorganic components (e.g. pods of micritic carbonates) and subsequent compaction of the mats may alter the depositional structure of the mat.

The marine groundwaters of the sabkha show a progressive landward increase in salinity as a result of evaporation (Kinsman 1964b, Butler, 1965). This produces four parallel gradational belts of distinct evaporite mineral assemblages and associated structures.

  • Upper intertidal: gypsum and celestite crystals and dolomitized calcium carbonate within the capillary zone.
  • High water: calcium sulfate hemihydrate (Skipwith, 1966), anhydrite nodules and dolomite accompanying solution of gypsum in the capillary zone; dolomite and large "sand crystals" of gypsum below the water table; halite precipitated by the evaporation of stranded tidal waters and capillary water at the air sediment interface.
  • Above high water: anhydrite polygons and diapirs within the capillary zone; gypsum and dolomite below the water table.
  • Adjacent to outwash fans: anhydrite converted to gypsum by the influx of less saline ground water.
The occurrence of these mineralogic belts is variable and one or more is often absent. The typical sabkha cycle is shown in a figure linked to the thumb nail images of the adjacent column. The surface of the sabkha is modified by both marine and aeolian erosion. For instance, on beach ridges, wind carries away only the finer grades of sand leaving behind a lag deposit of gastropod shells. Aeolian erosion is limited to where sands are dampened by capillary water from the water table, but marine erosion can extend deeper. In parts of the sabkha where it has been particularly effective, marine erosion exposes the water table. Such areas are usually covered by a thin, dry, halite crust, gypsum protrudes as vertical sand crystals, and the polygonal forms of anhydrite may be exposed. Marine flooding also breaches old beach lines transporting their sediment onto the sabkha behind. The outgoing water re breaches the beach ridges to produce small deltas (for example, west of Khusaifa and west of Quala). The sheets of flood water are driven about the sabkha by strong winds.
Salt encrusted algal flats occur between beach ridges, and to the north and seaward of them. They are covered by storm washover sediments as on the sabkha, south of the large eastern algal flat. Trenches dug at least 2 km inland from the present highwater mark reveal that algal laminae are still preserved above lagoonal sediments and a beachrock crust.

gypsum Mush Facies
gypsum actively precipitates at a depth of a centimeter or less beneath the surface sediment in the lower supratidal zone, and beneath crenulated mats of algae in the upper intertidal zone. The gypsum may be interlayered with storm washover carbonate sediment. The sediment is moist, preventing formation of anhydrite, except where dewatering of the sediment occurs, especially where footsteps and vehicle tracks have compressed the sediment. The surface carbonate washover sediment may contain halite, but it does not form an obvious surface. The crystals in the gypsum layer are variable in size, ranging from a few millimeters to a centimeter in diameter. The crystals overlie a cyanobacterial peat, commonly forming a transition zone of organic rich gypsum layers. The cyanobacterial peat is commonly laminated and displays desiccation polygons on the bedding plane. The cyanobacterial peats overlie ubiquitous hardgrounds that are cemented by carbonates or gypsum. They are common at the top of unconsolidated intertidal to subtidal carbonate sands and muds but are not always present.

Lower Salt Flat Facies Seaward of Beach Ridges
The seaward margin of the lower salt flat facies is primarily defined by a thin initially horizontal to updip convoluted crust of halite that overlies the layer of gypsum mush that locally and landward is being replaced by anhydrite. Updip and landward towards the stranded beach the upper sequence of anhydrite replaces the interBedded carbonate mud and gypsum mush completely. The anhydrite thickens often creating a mottled texture, particularly where the gypsum alters to anhydrite through dewatering. Locally in the storm washover sediments it forms nodules. The gypsum and/or anhydrite may be interlayered with storm washover carbonate sediment, while overlying a thicker layer of gypsum, a well developed cyanobacterial peat, a carbonate hardground, or a thick sequence of carbonate sand. The cyanobacterial peat is commonly laminated and displays desiccation polygons on the bedding plane. The cyanobacterial peats overlie hardgrounds that are cemented by carbonates or gypsum. They are common at the top of the carbonate sand, but are not always present.

This anhydrite may locally form the sediment surface or lie just beneath a polygonally cracked surface crust of halite. Underlying the anhydrite or halite surface is, in descending order, a sequence of storm washover carbonates, anhydrite, minor gypsum, and cyanobacterial peat all. Lying beneath the peat is a 2 5 cm layer of unconsolidated carbonate sand. Locally, a burrowed carbonate mud may replace this. Beneath this layer is a cemented 2 5 cm thick layer of medium to very coarse grained carbonate sand (an intertidal hardground) whose composition is identical to the overlying unconsolidated sand. Below the hardground is carbonate sand that is often over a meter thick. Locally, all but the carbonate sand facies may be absent. Large (up to 6 cm) gypsum crystals are common within or immediately below the lowest cyanobacterial peat layer, often preserving the original cyanobacterial laminations.

Stranded Beach Ridge Facies
Three to four kilometers from the intertidal cyanobacterial flats to the north are a series of stranded beach ridges that form topographic highs about a meter above the adjacent sabkha. These features are visible on aerial and remotely sensed images and are used by travelers to avoid the adjacent soft sabkha. Their most striking component is cerithid tests, but other gastropod and pelecypod tests are common too, along with encrusting bryozoan (that once coated long since decayed bladder wrack), cuttle fish bones, and foraminifera. This beach ridge belt is some two to three kilometers wide. It is thought to represent a shoreline formed some 3000 4000 years B.P. when the Khor al Bazam was more open and deeper (Kirkham 1998). No gypsum or anhydrite have been found within these ridges, probably because the groundwater trapped within them moves freely through them and is constantly replaced by marine waters. carbonate cementation, however, is quite common.

Upper Salt Flat Facies Landward of Beach Ridges
The upper salt flat facies south and landward of the cerithid beach ridges is capped by a widespread polygonally cracked surface crust of halite. Locally, the halite may be crumpled into convoluted 5 cm high sand rich ridges. The halite crust overlies a sand dominantly composed of carbonate with some quartz. Both the carbonate and the quartz were transported to their location by a combination of wind and storm wash over from the tidal flats to the north, and from the Tertiary and Quaternary sediments to the south by the wind and flash floods. Caught up in the sand, and traced from the surface down, are dispersed anhydrite nodules (1 to 5 cm in diameter), which downward are locally replaced by diapiric anhydrite layers, 1 to 10 cm thick. Beneath this anhydrite zone, with it's predominantly carbonate sand matrix, is the water table. Close to the water table at about a meter in depth, gypsum rosettes become common. As one approaches the landward margin of the coastal sabkha the anhydrite layers invert to gypsum in response the influx of continental water from the south and east.

Friday, March 29, 2013
Tulsa Web Design    Tulsa Graphic Design     Tulsa SEO    Tulsa Search Engine Optimization